Members: Login / Logout
Abstracts

Abstracts: Poster Abstracts

Please see below abstracts for posters to be presented at Focus 2017. Click the heading to jump to the relevant abstract.

Focus 2017: Poster Abstracts

  • P001 Serological testing for HIV, hepatitis B, hepatitis C and syphilis in dried blood spot samples on the Abbott Architect platform

    Briony Johnson (Heart of England NHS Foundation Trust, Birmingham, United Kingdom), Steven Wilson (Public Health Laboratory Birmingham, Birmingham, United Kingdom)

    Background: Dried blood spot (DBS) samples are particularly useful for outreach initiatives, home testing, hard-to-reach patient populations, and those who are difficult to bleed or reluctant to undergo standard venous phlebotomy. Although there is a reduction in test sensitivity relative to standard blood samples, DBS testing is still extremely valuable, since these groups are likely to be missed by standard serological screening programmes. The aim of the project was to assess the suitability of the Abbott Architect platform for routine serology testing for HIV, hepatitis B (HBV), hepatitis C (HCV) and syphilis in DBS samples.

    Methods: A panel of 90 DBS elutes, including 19 HIV-positives, 25 HBV-positives, 29 HCV-positives, 12 negatives and 5 dual-positive samples were used to compare the DiaSorin Liaison XL and Abbott Architect analysers. Perkin Elmer DBS cards were used, with 4 spots from each sample eluted in 1.5 mL 0.05% Tween-20 in PBS. In addition, a cohort of syphilis-positive patients was recruited to provide anonymised paired clotted blood and DBS samples for syphilis testing on the Abbott Architect. WHO international standards (NIBSC) were used to create dilution series for assessing analytical sensitivity for HIV, HBV and syphilis.

    Results: There were no observed discrepancies between the qualitative results generated on the DiaSorin Liaison and Abbott Architect platforms for any of the samples tested (100% sensitivity and specificity); however, the Architect platform showed a significant improvement in analytical sensitivity for all targets and all specimen types over the current Liaison test.

    Conclusion: This study indicates the Abbott Architect platform showed excellent clinical agreement with the DiaSorin Liaison and improved analytical sensitivity. Transferring the tests onto the Abbott platform will also improve our laboratory workflow and turnaround times.

  • P002 An audit of laboratory monitoring trace element status in patients receiving long-term parenteral nutrition in a District General Hospital

    Kate Fenna (Royal Surrey County Hospital, Guildford, United Kingdom), Callum Livingstone (Royal Surrey County Hospital, Guildford, United Kingdom)

    Background: Patients are treated with parenteral nutrition (PN) when nutrition by the oral and enteral is not feasible. Some patients require PN for >4 weeks. This is called termed long term PN (LTPN). In patients treated with PN laboratory monitoring is important both for early detection of biochemical complications and to ensure adequate nutrition. NICE guidelines state that in patients treated with LTPN, trace elements (TE) should be measured at baseline and monthly thereafter, namely zinc (Zn), copper (Cu), selenium (Se) and manganese (Mn). Levels should be monitored sooner in patients with severe gastrointestinal (GI) losses.

    Methods: This audit aimed to determine compliance with TE testing before and during LTPN at the Royal Surrey County Hospital (RSCH). All patients treated with LTPN under the care of the nutritional support team (NST) between Jan 2014 and Nov 2016 were included. Data on these patients was obtained from pharmacy PN records and from the LIMS.

    Results: Of the 383 patients that received PN, 45 patients (23 male, 22 female) received LTPN. Compliance with baseline testing was 36% (16/45). Compliance with monitoring was 22% (10/45). Analysis of requests indicated the most common reasons for non-compliance were incomplete TE requesting and wrong specimen type.

    Conclusions: Overall the audit suggested that there was poor compliance with both baseline testing and monitoring of TE concentrations in patients treated with LTPN. Recommendations made in order to improve practice included education of staff, updating of the specimen directory for correct specimen types required, an agreed test profile request code to reflex testing of Zn, Cu, Se and Mn (assuming correct specimens received) and stocking TE specimen collection tubes to minimise trace element contamination of specimens.

  • P003 A regional audit on the provision of faecal calprotectin testing in Northern Ireland

    Elinor V Hanna (Northern Health and Social Care Trust, Antrim, United Kingdom), Jennifer S Hamilton (Belfast Health and Social Care Trust, Belfast, United Kingdom), Gareth C McKeeman (Belfast Health and Social Care Trust, Belfast, United Kingdom)

    Background and Aim: NICE guidance DG 11 recommends faecal calprotectin (FC) to differentiate between inflammatory bowel disease (IBD) and irritable bowel syndrome (IBS). The aim was to establish the current service provision and demand for faecal calprotectin testing in NI.

    Method: A questionnaire was sent to Biochemistry laboratories in Northern Ireland across the 5 HSC Trusts.

    Results: A response was received from all 5 Trusts. All offered FC but only one Trust performed the test in house. Four Trusts referred samples to England or Wales (3 to same lab) with variation in cost. The laboratory performing FC in house was cheapest. There was variation in the cut-off values quoted for faecal calprotectin depending on the method used. FC testing was available in secondary care for all Trusts but only available to primary care in 4 Trusts (2 of these via GI consultant). One Trust had restricted the availability until a care pathway and funding were in place for primary care. Results were available electronically in 3 Trusts. One laboratory had a 2-fold increase in requests from 1203 (2014/15) to 2471 (2015/16). There were variations in the pattern of primary care to secondary care requesting. There were a significant number of patients who had repeated FC testing performed (between 2-5 times in a year) and often this was for monitoring of known IBD.

    Conclusions: Regional recommendations have been developed. These include the need to establish funding and a regional diagnostic pathway to guide use of faecal calprotectin in primary care with respect to NICE DG 11. There should be standardisation of a referral lab, cut-offs used, and electronic reporting. In addition, there are cost/efficiency benefits of one laboratory in NI providing analysis of FC for the region.

  • P004 The effect of different sampling techniques on the faecal immunochemical test haemoglobin concentration

    Carolyn J Piggott (NHS Bowel Cancer Screening Programme Southern Hub, Guildford, United Kingdom), Cerin John (NHS Bowel Cancer Screening Programme Southern Hub, Guildford, United Kingdom), Sally Benton (Surrey Pathology Services, Guildford, United Kingdom), Helen Bruce (Surrey Pathology Services, Guildford, United Kingdom)

    Faecal immunochemical Test (FIT) testing is an analytically superior method for detecting haemoglobin (Hb) in faecal samples compared to the traditional guaiac method. It is used in many screening programmes across the world and is being indicated as a suitable test for ruling out colorectal cancers in patients with intermediate risk symptoms.

    Faecal samples are typically collected into FIT collection devices by the patient before being sent to a laboratory for analysis. Clear instructions are provided by manufacturers to guide patients on how to collect the sample however, the nature of the collection method means that there is likely to be high variability in the amount of sample provided hence impacting the final result. Manufacturers claim the sample quantity in collection devices are standardised by an internal collar to control excess faeces.

    The aim of this study was to investigate whether different sample quantities loaded onto the collection sticks impact Hb concentrations.

    Five homogenous Hb-positive faecal samples were prepared and loaded, with increasing amounts of faeces on the stick, using collection devices supplied by Alfresa Pharma Corp, Eiken Chemical Co, Kyowa Medex Co Ltd, and Sentinel Diagnostics.

    Devices were inverted, stored for 24 hours at 23°C in the dark, then mixed before analysis on the NS-Prime (Alfresa), OC-SENSOR DIANA (Eiken), HM-JACKarc (Kyowa Medex), and SentiFIT 270 (Sentinel).

    For the NS-Prime the range of concentrations for the five samples with the least amount of sample added was 14-29 µg/g and these changed by 96-137% with increased faeces; likewise OC-SENSOR DIANA 12-41 µg/g changed by 99-135%, HM-JACKarc 11-44 µ/g changed by 91-162%; and SentiFIT 270 24-67 µg/g changed by 100-134%.

    The results demonstrate that with increased sample loaded onto the collection stick, there can be increased Hb detected, and this may impact the clinical patient pathway if a single cut-off for referral is used.

  • P005 Comparison of four faecal immunochemical test haemoglobin concentrations

    Cerin John (NHS Bowel Cancer Screening Programme Southern Hub, Guildford, United Kingdom), Carolyn Piggott (NHS Bowel Cancer Screening Programme Southern Hub, Guildford, United Kingdom), Sally Benton (Surrey Pathology Services, Guildford, United Kingdom)

    Faecal immunochemical Test (FIT) testing is an analytically superior method for detecting haemoglobin (Hb) in faecal samples compared to the traditional guaiac method. However, there is no international standardisation and each method is likely to give different results depending on the antibody used to the globin moiety. The aim of this study was to compare a range of Hb concentrations in naturally positive faecal samples on four analytical systems, the HM-JACKarc (Kyowa Medex Co Ltd), NS-Prime (Alfresa Pharma Corp), OC-SENSOR DIANA (Eiken Chemical Co.) and SentiFIT 270 (Sentinel Diagnostics). Nine homogenous Hb-positive faecal samples were loaded into each of the devices supplied by the manufacturers. The devices were inverted and stored for 24 hours at 23°C in the dark and mixed before analysis on the respective analyser. The results (ng Hb/mL buffer) were converted to µg Hb/g faeces using the manufacturers’ stated conversion and compared using regression analysis. The OC-SENSOR (sample result range 5-377 µg/g) and HM-JACKarc (11-479 µg/g) gave similar results to each other (y=1.0089x-3.5135). The NS-Prime (10-318 µg/g) had a negative bias relative to the OC-SENSOR (y=0.7923x-0.583) and HM-JACKarc (y=0.734x+6.1362). The SentiFIT (23-1104 µg/g) had a positive bias relative to OC-SENSOR (y=2.8629x-30.911) and HM-JACKarc (y=2.6716x-8.1282). The largest difference in Hb concentration was for the samples analysed on the SentiFIT and NS-Prime (y=3.5454x-24.529). The results show there are differences between the FIT analytical systems for Hb measurement, and some systems are closer than others. Depending on the analytical system, samples with the same Hb concentration could be positive on one system and negative on another. It is important to be aware of this if considering a single cut-off for symptomatic and screening populations.

  • P006 Introduction of symptomatic patient screening for faecal occult blood

    Natalie Hunt (Lancashire Teaching Hospitals NHS Foundation Trust, Preston, United Kingdom), Carol Wignall (Lancashire Teaching Hospitals NHS Foundation Trust, Preston, United Kingdom), Susan Wareing (Lancashire Teaching Hospitals NHS Foundation Trust, Preston, United Kingdom), Martin A Myers (Lancashire Teaching Hospitals NHS Foundation Trust, Preston, United Kingdom)

    NICE updated guidance NG12 on Suspected cancer: recognition and referral (June 2015), recommends faecal occult blood (FOB) testing in patients suspected of having colorectal cancer. We chose faecal immunochemical testing (FIT) as it is a more sensitive and specific method of detecting human haemoglobin than guaiac acid methods.

    Lancashire Teaching Hospitals released local guidance in June 2016 to mirror NICE guidance on using FOB to triage low risk patients for colorectal cancer without red flag symptoms. We send our FOB requests for FIT analysis to a referral laboratory and we have carried out a FIT method evaluation with the aim of measuring it locally.

    There is debate about which cut-off to use in symptomatic patients and we use a cut-off of 10 µg/g.

    Method: Two faecal collection devices were sent to each patient with an FOB request between Sept-Dec 2016 (n=60). One device was sent to the referral laboratory for analysis using HM-JACK arc and the other was processed in-house using the iO OC-Sensor. Analyse-it was used for statistical analysis.

    Results: Concordance between both methods was seen for 95% of patients. Three patients (5%) had discrepant results but all of these were close to the cut-off, this will need to be explored but may be due to sampling from different parts of the faeces with separate devices. Overall, 17% patient samples were positive using HM-JACK compared with 18% with OC-Sensor. We are confident of the OC-Sensor method precision to levels lower than the recommended cut-off of 10 µg/g making the assay fit for the purpose of symptomatic patient screening.

    Conclusion: HM-JACK and OC-Sensor methods compared very well, both analytically and clinically, in this patient comparison study. Whichever method is used, over 80% of low risk patients could avoid invasive colonoscopy by introducing FIT testing.

  • P008 Use of second generation Mitra micro-sampling device for blood lead measurements

    Arshad N Mahmood (Viapath LLP, London, United Kingdom), James Rudge (Neoteryx Europe, RD Geleen, Netherlands), Kishor Raja (Viapath LLP, London, United Kingdom)

    Lead can have detrimental effects on human health and wellbeing even at low concentrations; it is thus considered a serious global problem. Despite its removal from petrol, risk of lead exposure exists from old paints, pipes, industrial waste and medicines (ayurvedic). The current process of venous blood sampling for lead measurement has various constraining factors: firstly, the possibility of contamination from the sampling site, needle or vacutainer. Appropriate packing and postage is also required for transportation and there is risk of inadequate volume obtained from young patients or patients difficult to bleed (e.g. elderly) or are needle-phobic. The Mitra® devices can accurately collect 10 µL whole blood (from a finger prick), and could help overcome some of the aforementioned issues. The method developed for analysis entailed extraction of sample from pre-dried Mitra tip (left covered for 60 minutes post exposure to surplus patient venous blood sample) in 2% tetra-methyl ammonium hydroxide (TMAH) and with brief sonication (5 sec). Vortexing instead of sonication yielded considerably higher lead concentrations. The extracted sample was centrifuged prior to analysis by inductively coupled plasma-mass spectrometry (ICP-MS) using standard mode of operation. Intensities (as counts per second) of extracts obtained from Mitra tips exposed to various blood lead concentrations (0.15-2.45 µmol/L) compared well with those obtained using our standard wet method. The Mitra tip (+ sampler body) by itself did not introduce any lead contamination. Intra-assay variability was <15% at clinically significant and high blood lead concentrations when using Mitra tip as the sampling device. There was good correlation for blood lead between samples assayed as per normal (wet method) against samples extracted from the Mitra tip (r2=0.92, n=32, concentrations ranged from 0.20-2.50 µmol/L). The Mitra tips offer a potential solution for collection and transport of small volumes of whole blood for lead measurement.

  • P009 Acute hyperammonaemic encephalopathy in a patient with arginine deficiency

    Mark Lum (Aberdeen Royal Infirmary, Aberdeen, United Kingdom), Fiona M Brandie (Aberdeen Royal Infirmary, Aberdeen, United Kingdom), Helen Regan (Aberdeen Royal Infirmary, Aberdeen, United Kingdom), Kevin Deans (Aberdeen Royal Infirmary, Aberdeen, United Kingdom)

    Acute hyperammonaemic encephalopathy is an important diagnosis to consider in the adult patient with reduced consciousness, as if not treated promptly it has significant morbidity and mortality.

    A 49 year-old gentleman was admitted to the High Dependency Unit with reduced conscious level. Arterial blood gases showed respiratory alkalosis. Previous history included protein losing enteropathy of unknown aetiology. Due to symptomatology he had a significantly self-restricted diet. He was treated for presumed sepsis with IV antibiotics, and then empirically with antivirals as there was no improvement. Brain imaging did not demonstrate any acute changes. Metabolic encephalopathy was suggested by neurology as a potential cause, and they advised that plasma ammonia be checked. This was elevated at 148 μmol/L (reference interval <50 μmol/L).

    Plasma amino acid, urine organic acid and blood spot acylcarnitine analysis revealed no evidence of an inherited metabolic disorder. Plasma amino acid analysis revealed significantly increased glutamine of 2338 µmol/L (300-900), consistent with the elevated ammonia concentration, and arginine towards the lower end of the reference interval at 24 µmol/L (20-140). There was no evidence of significant liver dysfunction, nor was the patient receiving any medication known to cause hyperammonaemia. Blood, stool, urine and CSF cultures were negative. The patient was treated with intravenous sodium benzoate, sodium phenylbutyrate and arginine, leading to the patient’s ammonia and glutamine decreasing to within the reference intervals, and conscious level returning to normal.

    This case demonstrates the need for ammonia to be checked urgently in cases of unexplained reduced conscious level and/or respiratory alkalosis. When hyperammonaemia is identified, treatment must be instituted promptly. There is a role for laboratory specialists in urgently communicating elevated ammonia results and directing clinicians to sources of specialist metabolic advice.

  • P010 An evaluation of the AST:ALT ratio in identifying patients in primary care for appropriate referral to gastroenterology

    Raphael Buttigieg (NHS Lothian, Edinburgh, United Kingdom), Sara Jenks (NHS Lothian, Edinburgh, United Kingdom)

    Introduction: In NHS Lothian we developed a guideline for the investigation and management of patients with asymptomatic abnormal LFTs in primary care in 2014. This guideline recommended using the AST:ALT ratio as part of a screen for patients developing fibrosis and cirrhosis. The ratio was only calculated if the ALT result was above the reference range (>50 U/L). An AST:ALT ratio of >1.0 was considered to be abnormal, and an automated comment was attached to all abnormal results recommending referral to gastroenterology. The aims of this audit were to evaluate the impact of this ratio on patient management in primary care.

    Methods: We retrospectively reviewed all AST:ALT ratios measured in primary care between 01/03/2015 and 31/08/2015. We audited whether patients were referred to gastroenterology and the outcome of their gastroenterology review. This was done for all 56 patients with an abnormal AST:ALT ratio and a randomly selected subgroup of 47 patients with an AST:ALT ratio just below the referral threshold of between 0.8-1.0.

    Results: There were 742 AST:ALT ratio requests received from primary care within the six month audit period. Two hundred and eighty eight ratios were measured and the remainder were cancelled due to ALT being <50 U/L. Of the 288 AST:ALT ratios measured, 56 (19%) were abnormal. Of these 56 patients, 82% had all the recommended liver screen investigations completed and 73% were appropriately referred to gastroenterology. Of those patients referred 93% had further investigations arranged and/or were kept under gastroenterology follow-up. Of the patients with a ratio just below the referral threshold of 0.8-1.0 and an unremarkable liver screen only 6% patients were referred, potentially inappropriately, to gastroenterology.

    Conclusions: The AST:ALT ratio has been successfully used with NHS Lothian to identify appropriate patients for referral to gastroenterology with almost all of those referred being kept under review by gastroenterology for further investigation. Further research on risk stratifying patients, and comparison to FIB4 and APRI is currently underway.

  • P011 An audit of the first nine months of a primary care faecal calprotectin service

    Amy R Frank (Derriford Combined Laboratory, Plymouth, United Kingdom), Jenny Lowles (Performance and Management Information Department, Plymouth, United Kingdom), Jinny Jeffery (Derriford Combined Laboratory, Plymouth, United Kingdom), Sean J Costelloe (Derriford Combined Laboratory, Plymouth, United Kingdom)

    Background: Faecal calprotectin testing was introduced for general practice in the region in April 2016. Prior to this GPs referred patients to gastroenterology when they presented with symptoms of inflammatory bowel disease (IBD). Local guidelines were issued to GPs when the service was introduced. This audit was performed to determine whether these guidelines are being followed appropriately and whether introduction of the service has affected the number of referrals for colonoscopy.

    Method: Data for a nine-month period from April 2016 was obtained by the Trust Performance and Management Information Department. Data including calprotectin results, colonoscopy requests and diagnosis was extracted from multiple electronic systems in use within the Trust.

    Results: During the data period, 699 patients aged 18-45 years had a first request for faecal calprotectin from primary care. Of these 41 had a colonoscopy within 6 months of a calprotectin test and 29 of these had a new diagnosis of IBD. Categorising these patients by calprotectin result (µg/g) showed 10/15 with a result <45, 5/7 with a result 45-100, 3/3 with a result 101-250, 9/13 with a result >250 and 2/3 patients with no numerical result had a diagnosis of IBD following colonoscopy.

    Patients >45 years of age accounted for 29% of requests, despite local guidelines stating that calprotectin is not recommended in this age group. This figure did not change significantly after the provision of information on the guidelines for GPs requesting tests on the electronic ordering system.

    Conclusion: Patients with calprotectin results >250 were ten times more likely to have a colonoscopy than those with results <45 (43% vs 4%), but of those patients who received a colonoscopy 66-70% patients were given a diagnosis of IBD irrespective of their calprotectin concentration. In the 101-250 µg/g group the data was insufficient to evaluate the percentage.

  • P012 Development of an External Quality Assessment Scheme for quantitative faecal haemoglobin

    Samantha Jones (Cardiff and Vale UHB, Cardiff, United Kingdom), Gareth Davies (WEQAS, Cardiff, United Kingdom), Mary A Thomas (WEQAS, Cardiff, United Kingdom), Daniel Hopkins (Cardiff and Vale UHB, Cardiff, United Kingdom)

    Introduction: In 2016, the WEQAS Faecal Haemoglobin (FOB/FIT) programme was further developed for quantitative FIT.

    Aims: To develop and validate material for quantitative FIT and assess its stability and commutability. To assess the process of loading material onto the analyser sample pickers.

    Method: Organic material from the current qualitative FIT programme was shipped to 2 Screening Centres to assess its stability in transit. The material closely mirrors the basic constituents of human faeces and was spiked with human whole blood (Hb) to cover the pathological range. Samples with no preservative in universal containers and pre-loaded into the sample pickers with stabilizing buffer were assessed.

    Further matrices were developed and assessed for recovery, stability and consistency of sample loading on the OC-Sensor Diana and HM JACKarc; original organic matrix; original organic matrix with improved biocide agent; original organic matrix with alternate source of mucoid; new in house organic matrix; third party supplied organic matrix. Comparability of the matrices was assessed from results across 5 different FIT platforms. Ten sites across UK, Europe and Asia were recruited to take part in the study. Each site was sent 3 samples per month.

    Results and Conclusion: No significant differences were observed between the samples in the universal container and buffered tubes in one screening centre, however, a decrease in results was observed for the universal container compared with the buffered tube in the other centre.

    Good recovery was observed between spiked haemoglobin and the matrix with alternative mucoid.

    There were variable responses to sample loading dependent on the type of sample picker. Insufficient sample was taken up by some pickers; this was more pronounced with less smooth consistencies of matrix. It is foreseeable that this would also be the case for patient samples. All systems had mechanisms in place to prevent overloading.

  • P013 Stability of thiamine diphosphate, flavin adenine dinucleotide and pyridoxal-5’-phosphate in whole EDTA blood samples

    Adaku B Ajaegbu (Viapath, London, United Kingdom), Agata Sobczyńska-Malefora (Viapath, London, United Kingdom; King’s College London, London, United Kingdom)

    Water soluble vitamins B1, B2 and B6 are essential for many metabolic processes including carbohydrate, lipid, homocysteine and protein metabolism. Coenzymes thiamine diphosphate (TDP), flavin adenine dinucleotide (FAD) and pyridoxal-5’-phosphate (PLP) are markers of vitamin B1, B2 and B6 status respectively. The stability of these compounds is a key preanalytical variable which needs to be accounted for before analysis. Assay manufacturers recommend that samples for these tests are frozen following blood collection. These requirements may preclude the analysis of samples from distant locations.

    We assessed the suitability of whole blood (WB) EDTA samples for analyses of TDP, FAD and PLP when the specified temperature and duration of storage conditions had been exceeded. Eight WB samples were collected and each divided into aliquots. A baseline aliquot was immediately stored at -20°C. Other aliquots were stored in the dark at 2-8°C, and at room temperature (RT) for 24 hours, 72 hours and 5 days before freezing at -20°C. The samples were analysed according to Chromsystems instructions using HPLC with fluorescence detection.

    The mean baseline, 24 hours, 72 hours, and 5 days (fridge, RT) concentrations were: TDP = 132, (126, 135), (126, 137), (120, 120); FAD = 230, (226, 228), (223, 232), (216, 223); PLP = 110, (107, 122), (108, 121), (110, 116) nmol/L respectively. Repeated measures ANOVA demonstrated no difference in TDP, FAD and PLP concentrations for samples kept at 2-8°C for up to 5 days: TDP (p=0.132), FAD (p=0.169) and PLP (p=0.244). FAD was stable for up to 5 days (p=0.328) and TDP for up to 72 hrs (p=0.262) at RT while PLP was unstable at RT.

    Samples for TDP, FAD and PLP are suitable for storage and transportation at 2-8°C. TDP and FAD are stable at RT for at least 3 days. Storage of samples at RT for PLP analysis is not recommended.

  • P014 Unusual cause of hypokalaemia in a patient on total parenteral nutrition

    Rajiv Chudasama (University Hospitals of Leicester, Leicester, United Kingdom), Melanie Baker (University Hospitals of Leicester, Leicester, United Kingdom), James Stewart (University Hospitals of Leicester, Leicester, United Kingdom), Faizanur Rahman (University Hospitals of Leicester, Leicester, United Kingdom)

    A 44 year-old woman with severe Crohn’s disease, multiple resection and complex fistulae on home TPN presented with unresolving hypokalaemia, persistent for a period of 6 months. MRI revealed an enterocutaneous fistula connecting proximal jejunal loop; she also had multiple entero-entero fistulas involving the proximal jejunum and colon. Her blood tests at the time showed sodium 135 mmol/L, potassium 3.0 mmol/L, urea 10.7 mmol/L and creatinine 61 µmol/L. Hypokalaemia was persisting despite her having potassium 100 mmol/24 hours in her TPN feed. She only weighed 64 kg and her fistula losses were minimal. Her urine electrolytes at the time revealed urine potassium 66 mmol/L, suggesting urinary potassium loss. On reviewing her acid-base status, she was found to be profoundly alkalotic with bicarbonate 39 mmol/L, suggesting metabolic alkalosis. Dehydration and alkalosis would result in hyperkaliuria due to renin-angiotensin-aldosterone activation. Her TPN feed was reviewed, the patient was getting 350 mmol of Na, most of it being given as acetate. Total acetate in the feed was 400 mmol. This lady may have had hyperchloraemic acidosis earlier on, but her prescription was not altered when her condition changed. Acetate gets converted to bicarbonate; therefore it was switched to chloride in the TPN feed. Decreasing acetate corrected her alkalosis and subsequently her potassium level. Acetate is a conjugate base of acetic acid which after cellular uptake forms acetyl-CoA and enters the citric acid cycle. This results in carbon dioxide and water which attain equilibrium with bicarbonate through carbonic anhydrase. Gastrointestinal diseases can give rise to complex acid-base disturbances. These are dictated by the condition affecting it, the bowel segments involved and any surgeries done, some of which can result in stomas or fistulas. Good understanding of these factors and interpretation of blood and urine biochemistry is the key in managing these disorders.

  • P015 A one-year retrospective audit on calprotectin: how well is primary care adhering to the pathway for inflammatory bowel disease?

    Benjamin T Palmer (Betsi Cadwaladr University Health Board, Rhyl, United Kingdom), Steven J McCann (Stockport NHS Foundation Trust, Stockport, United Kingdom)

    Following the recommendations of NICE guidance 11 2013, faecal calprotectin testing was incorporated into the pathology services at Stockport NHS Foundation Trust hospital on 01 November 2014, with the aims to improve patient experience, reduce the number of unnecessary endoscopies and release endoscopy capacity. An audit was therefore carried out to specifically look at how well primary care is adhering to the clinical pathway 1 year on. A questionnaire was designed and sent out to all primary care surgeries (n=587) in the catchment area that had requested faecal calprotectin testing. After reviewing the 234 (40%) questionnaires returned, 216 (37%) responses were suitable for assessment. Results have demonstrated that the worst non-conformance to the IBD pathway was length of time of presenting signs/symptoms with 69% not compliant, followed by exclusion of gastrointestinal infection (63%), age compliance (48%) and exclusion of red flag signs/symptoms (35%). Both withdrawal of non-steroidal anti-inflammatory drugs (NSAID) and antibiotics prior to testing, however, achieved 94% compliance. Overall it would appear that the adherence of primary care to the IBD clinical pathway at Stockport NHS Foundation Trust hospital could be improved and that they may benefit from re-education and training. Particular areas to focus on would be age compliance, immediate referral of any patient with red flag signs/symptoms and the exclusion of gastrointestinal infection by clinical examination and patient history before requesting the faecal calprotectin test. Early investigations into clinical outcomes would suggest that the current cut-off level of calprotectin (50 μg/g) is set at an appropriate level for stratifying patients for endoscopy. However, this observation is only based on 3 new confirmed cases of IBD: at the time of writing this audit 34 requests were awaiting diagnostic confirmation and therefore the results of these requests will be of interest to further substantiate this finding.

  • P016 Verification of faecal calprotectin using the BÜHLMANN fCAL® turbo assay

    Emma L Evans (Heart of England NHS Foundation Trust, Birmingham, United Kingdom), Helen Wiggins (Heart of England NHS Foundation Trust, Birmingham, United Kingdom), Craig Webster (Heart of England NHS Foundation Trust, Birmingham, United Kingdom)

    Introduction: Faecal calprotectin (fCAL) is an indicator of inflammatory processes and has an important role in the diagnosis and monitoring of inflammatory bowel disease (IBD). NICE (DC11) recommends fCAL testing in primary care to aid in the differential diagnosis of IBD and irritable bowel syndrome (IBS) in adults where specialist assessment is being considered.

    As a consequence of this guidance we have seen the workload for fCAL dramatically increase. Due to the nature and inherent inhomogeneity of this specimen and the increased workload, our laboratory evaluated the user-friendly CALEX® Cap Extraction device using the using the BÜHLMANN fCAL® turbo assay on the Abbott ARCHITECT platform.

    Methods: Extracts of 35 faecal patient samples were analysed with The BÜHLMANN fCAL® turbo and compared to results generated with the BÜHLMANN fCAL® ELISA method. The stability of fCAL both in stool and extracted samples was assessed at various temperatures.

    Results: The assay has been tested to be linear in the range from 21-2058 µg/g calprotectin in stool. The between batch assay precision (CV) was ≤4.13%. Passing and Bablok regression was y=1.44+1.26 and Altman-Bland difference plot demonstrated a +7.9% positive bias against the ELISA method. Stool samples were only stable if stored in the freezer (below -18°C) for up to 7 days. Extracted samples were stable in the fridge (2-8°C) or in the freezer (below -18°C) for up to 7 days (p>0.05).

    Conclusions: Here we demonstrate the BÜHLMANN fCAL® turbo assay on the Abbott ARCHITECT platform is fit for purpose and we have recently received UKAS accreditation for this assay. Close collaboration with our gastroenterology team has enabled us to provide a new IBS local pathway for primary care to follow.

  • P017 Deletion of IL-6 induces impairment in UCP1 production in response to high fat diet

    Benjamin Zucker (Imperial College London, London, United Kingdom), Sam Virtue (University of Cambridge, Institute of Metabolic Science, Cambridge, United Kingdom), Jeffrey W Dalley (University of Cambridge, Department of Psychology, Cambridge, United Kingdom), Antonio Vidal-Puig (University of Cambridge, Institute of Metabolic Science, Cambridge, United Kingdom)

    Previous research has shown that transitions from chow diet to high fat feeding causes a brown adipose tissue (BAT) mediated thermogenic response, a phenomenon termed diet-induced thermogenesis (DIT). DIT is considered to be an adaptive response to increase energy output in cases of high energy input. Exogenous IL-6 administration has been shown to elicit a thermogenic response in mice and its absence in IL-6-/- mice impairs other BAT-dependent forms of thermogenesis, such as cold-induced thermogenesis. This study aims to investigate whether IL-6 is the chemical mediator responsible for eliciting BAT-dependent thermogenesis in response to high-fat diet (HFD). Energy expenditure was measured by indirect calorimetry while WT and IL-6-/- mice ate chow or HFD over 48 hours in each condition at thermoneutrality. RT-qPCR was used to measure expression of candidate genes, including uncoupling protein 1 (UCP1). No significant difference was found upon comparison of energy expenditure between IL-6-/-and WT mice on HFD; however, IL-6-/- mice did show a significant reduction in UCP1 expression when compared to WT mice in response to HFD (p=0.009). Furthermore, corollary evidence indicated alterations in the expression of other genes involved in BAT-dependent thermogenesis, such as PPARg2 and PGC1a. These results implicate IL-6 as a potential mediator in the pathway that effects DIT in response to HFD. However, the absence of a significant difference in thermogenic response between IL-6-/- and WT mice indicates that IL-6 is unlikely to be the sole mediator of this response.

  • P018 Relationship between faecal calprotectin levels and histology obtained at colonoscopy

    Lataben Seyani (Imperial College Healthcare NHS Trust, London, United Kingdom), Jamshid Alaghband-Zadeh (Imperial College Healthcare NHS Trust, London, United Kingdom), Alan P Courtney (Imperial College Healthcare NHS Trust, London, United Kingdom)

    Background: Faecal calprotectin correlates with the number of neutrophil granulocytes in the intestinal lumen. It has been demonstrated to be a useful marker of disease activity in inflammatory bowel diseases (IBD), such as Crohn’s disease and ulcerative colitis. The aim of this study was to compare histology findings from multiple biopsies obtained from patients at colonoscopy with measured faecal calprotectin levels.

    Methods: Histology and faecal calprotectin results were obtained from 102 patients attending Imperial College Healthcare NHS Trust between January and June 2016. For the purposes of the study, only patients with histology reports from the lower GI tract (including the anus, rectum, colon, and caecum) were included. Histology reports were scored according to Rubin et al (2013) which records increasing severity from on a scale from 0-5 (where, 0 = normal, and 5 = severe active inflammation). Faecal calprotectin measurement was performed using the Calpro2 immunoassay on the Phadia 250 from Thermo.

    Results: A significant increase in calprotectin results was observed in line with increasing histological inflammation (p<0.001). It was noted that some patients demonstrated negative histology, but had raised faecal calprotectin results. Equally other patients showed positive histology, but with lower than expected faecal calprotectin results.

    Conclusions: A significant association between increasing levels of faecal calprotectin and increasing severity of histological inflammatory activity was observed. This suggests that the magnitude of measured calprotectin could be used to predict histological severity. Further, the wide range of faecal calprotectin concentrations observed in each histological inflammatory category suggests that faecal calprotectin could be detecting inflammation not visualised on histology. Further investigation is needed to explain these findings.

  • P019 Developing ALT reference ranges: are age and gender important?

    Emma Bodenham (Imperial College NHS Trust, London, United Kingdom), Jaimini Cegla (Imperial Healthcare NHS Trust, London, United Kingdom), Tricia Tan (Imperial College NHS Trust, London, United Kingdom)

    Aim: Gender-specific reference ranges for alanine aminotransferase (ALT) are used by some laboratories, based on the observation that levels are higher in men versus women in some studies. We set out to determine whether a healthy volunteer population from the UK would also show a difference in ALT levels based on gender. We also hypothesised that ALT may increase with age, and investigated whether there was a difference in ALT levels between age groups.

    Methods: The results from total of 241 healthy volunteers (120 female and 121 male), recruited into clinical studies between 2007-2016 were reviewed. Participants were between 18-68 years old. Analysis was performed by gender and by age.

    Results: The median (2.5th, 97.5th percentile) ALT for the all-ages female group was 19.5 (9.0, 67.8) IU/L, and in the all-ages male group this was 29.0 (12.1, 88.4) IU/L (p<0.0001, Mann-Whitney U test). When analysed by age (including both genders), the median ALT (2.5th, 97.5th percentile) for the 18-25 year old subgroup was 19.0 (9.3, 85.7) IU/L, rising to 27.0 (9.2, 75.4) IU/L in the 26-39 year old subgroup, 24.0 (110, 101.6) IU/L in the 40-54 year old subgroup and 27.0 (13.0, 68.0) IU/L in the ≥55 year old subgroup (p=0.023, Kruskal-Wallis test; p=0.0344 for difference between 18-25 year old and ≥55 year old subgroups). Similar increases in ALT with age were seen with gender-specific analysis.

    Conclusions: This data shows ALT values were generally higher in males and in older populations. However, in this study although all participants were labelled ‘healthy volunteers’ we cannot exclude that the increase in ALT seen with age may be due to other patient factors (e.g. body mass index and patient comorbidities such as fatty liver disease), as these were not recorded. As participants were trial volunteers, participants generally were relatively young. Future studies will be required to further investigate the effect of age. However, our initial findings indicate a gender difference in ALT and support the use of gender-specific reference ranges.

  • P020 Clinical utility of PIVKA-II in the diagnosis of hepatocellular carcinoma

    Volha Klimovich (Viapath, London, United Kingdom), Kieran Voong (Viapath King’s College Hospital, London), Roy Sherwood (Viapath King’s College Hospital, London), Dominic J Harrington (Human Nutristasis Unit, Viapath, St Thomas’ Hospital)

    Alpha-fetoprotein (AFP), the marker that is currently in widespread use in diagnostic modalities in the UK for hepatocellular carcinoma (HCC), lacks sensitivity and specificity. Under-carboxylated forms of the vitamin K-dependent protein prothrombin (Protein Induced by Vitamin K Absence, PIVKA-II), is a tumour marker which has demonstrated higher rates of sensitivity and specificity in the diagnosis of HCC than those of AFP. Furthermore, PIVKA-II can be successfully used as a prognostic marker, which widens the field of the potential application of this tumour marker. However, utilisation of PIVKA-II for this purpose is currently not employed in the UK.

    Our aim was to determine the clinical utility of PIVKA-II for the diagnosis of HCC. The pilot project was based in the Nutristasis Unit of Viapath at St Thomas’ Hospital (London). In the course of the project an automated chemiluminescent microparticle immunoassay was verified and used for quantitation of PIVKA-II in 87 patient samples received from the Gassiott Gastroenterology Clinic (St Thomas’ Hospital, London) and the Hepatocellular Carcinoma Clinic in the Institute of Liver Studies (King’s College Hospital, London). All samples found to be PIVKA-II positive were screened for the presence of vitamin K antagonists (warfarin) to eliminate possible interference.

    In our study PIVKA-II exhibited higher sensitivity and specificity (79.4% and 96.6%) comparing to those of AFP (70.6% and 82.4%). These preliminary data suggest that PIVKA-II is a promising tumour marker for HCC diagnosis.

  • P021 Evaluation of the BÜHLMANN fCAL® turbo calprotectin method on the Roche Cobas 6000 (c501)

    Pamela Bowe (North Cumbria University Hospitals NHS Trust, Carlisle, United Kingdom), Francis Leslie (North Cumbria University Hospitals NHS Trust, Carlisle, United Kingdom), Sally Slack (North Cumbria University Hospitals NHS Trust, Carlisle, United Kingdom)

    Background: Faecal calprotectin results are used to assist clinicians with the differential diagnosis of inflammatory bowel disease (IBD) and irritable bowel syndrome (IBS). Since the NICE diagnostic guideline (DG11) was published there has been a significant increase in demand for this test. Many laboratories use ELISA technology to analyse faecal calprotectin. We investigated the performance of the new BÜHLMANN fCAL® turbo method which is CE marked for use on a number of mainline chemistry analysers.

    Method: The BÜHLMANN fCAL® turbo particle enhanced turbidimetric immunoassay (PETIA) method on the Roche Cobas 6000 (c501) was compared to the BUHLMANN calprotectin ELISA method on the Dynex DS2. Samples were extracted using the BÜHLMANN CALEX® extraction device prior to analysis on both methods. Regression analysis and Bland-Altman plots were used to compare results on 58 patient samples. Intra-assay precision was determined using 10 replicates of patient samples and inter-assay precision was calculated using 17 replicates of internal quality control material. NEQAS samples were analysed and bias relative to the all laboratory trimmed mean (ALTM) was assessed.

    Results: Comparison of patient results showed good correlation (r2=0.97). Regression analysis produced the following equation: fCAL® turbo = 1.14 DS2 result -23.42. The fCAL® turbo method demonstrated a negative bias at lower concentration and a positive bias at high concentration when compared with the ELISA method. Intra-assay precision (%CV) was 3.1% and 1.3% at concentrations of 48 µg/g and 247 µg/g respectively. Inter-assay precision was 3.3% at 73 µg/g and 1.1% at 247 µg/g. When compared to the NEQAS ALTM the fCAL® turbo method demonstrates a positive bias at concentrations above 100 µg/g.

    Conclusion: The BÜHLMANN fCAL® turbo method demonstrates acceptable performance. A commutable reference material for calprotectin is required to define analytical accuracy in the future.

  • P023 The utility of CSF ACE as a biomarker of neurosarcoidosis

    Alan P Courtney (Imperial College Healthcare NHS Trust, London, United Kingdom)

    Background: Sarcoidosis is a multisystem disease of unknown cause resulting in accumulation of noncaseating granulomas (NCGs) in affected organ tissues. Studies suggest that ~5-15% of sarcoidosis patients develop neurosarcoidosis (NS), but of these approximately half are subclinical. NS can be a severe/life threating disorder and diagnosis is notoriously difficult, which often relies upon the exclusion of other causes, such as multiple sclerosis (MS). CSF abnormalities are common in NS (80% at presentation). Here we demonstrate the utility of CSF ACE as a biomarker of NS.

    Methods: Fifteen NS and 34 MS patient results were randomly collated from internal patient samples collected at Imperial College Healthcare NHS Trust. Measurement of CSF ACE used an in-house colorimetric assay-based on the serum method (FAPGG conversion to FAP and GG, absorbance change at 340 nm) and optimised for use in CSF.

    Results: General CSF abnormalities were observed in both NS and MS groups of patients. However, these were generally non-specific in NS patients. Approximately half of NS patients had a moderately raised CSF ACE of ≥1.5 IU/L compared to less than 10% of MS patients. Thirteen percent of NS patients had CSF ACE that was ≥2.0 IU/L, compared with only 3% of MS patients. No MS patients had results greater than 2.4 IU/L.

    Conclusions: These results suggest that, while there is no ideal single biomarker for NS, CSF ACE can be useful as a valuable tool to aid diagnosis of neurosarcoidosis. Specifically CSF ACE may be useful in identifying some patients with NS who have CSF oligoclonal patterns that are consistent with MS.

  • P024 Mutational analysis of mucolipidosis IV and determination of mutation age in Omani population

    Badriya Alawi Jr. (Tallaght Hospital, Dublin, Ireland), Fahad Al-Zadjali (Sultan Qaboose University, Muscat, Oman), Khalid Al Thihli (Sultan Qaboose University, Muscat, Oman)

    Background: Type IV Mucolipidosis (MLIV) is caused by deficiency of the mucolipin enzyme encoded by MCOLN1. It is a slowly progressive neurodegenerative disorder causing severe psychomotor developmental delay, vision impairment and achlorhydria. It usually presents during the first year of life and is characterised by the presence of lysosomal inclusions in many cell types in the affected patients. Although approximately 70-80% of patients with MLIV are identified in Ashkenazi Jewish, MLIV is a pan-ethnic disorder. Recently, two Omani families were clinically diagnosed as MLIV, based on their clinical phenotype. The two families are independent and unrelated, though coming from the same region in Oman, yet the same causative mutation was identified in affected individuals.

    Objectives: To carry mutation analysis of MCOLN1 gene in the two families of MLIV and to screen for the identified mutation in the Omani general population coming from the same region of the identified families, in order to determine carrier frequency.

    Method: Patients and families were ascertained clinically through the Genetic and Developmental Medicine Clinic at the SQUH. Patients underwent clinical evaluation and laboratory investigations. DNA samples were extracted from peripheral blood samples of the patients, their parents and siblings. MCOLN1 gene exons were sequenced by direct DNA sequencing. Mutation screening was conducted on 1280 subjects from the same geographical regions as the patients origination using high resolution melting.

    Results: A novel mutation was identified; c.237+5G>A, in all affected individuals, even though they are unrelated. The novel mutation was not detected through screening 1280 individuals from the same geographical regions of the two families.

    Conclusion: A novel MLIV causative mutation was found in the Omani population. The absence of this variant among 1280 healthy individuals from the same region support this variant being pathogenic. Inability to detect the mutation amongst screened individuals from the same region does eliminate the possibility of a founder effect, given the rarity of this disorder, yet with the same mutation implication in both unrelated families.

  • P025 Implementation and evaluation of a clinical decision-support tool to reduce inappropriate requests for genetic thrombophilia screening

    Lewis Green (University Hospitals of North Midlands Department of Clinical Biochemistry, Stoke-on-Trent, United Kingdom), Christopher J Duff (Department of Clinical Biochemisty, Stoke-on-Trent, United Kingdom)

    Background: Estimates suggest 30-40% of pathology requests are unnecessary. With increasing requirements for cost-effectiveness, there is a need for efficacy and demand management. At our hospital we provide a genetic screening service to determine if patients have an increased inherited risk of developing venous thromboembolism or its related events. This requires the genotyping of two genetic risk factors, Factor V and prothrombin, for the p.R506Q and g.G>A20210 variants, respectively.

    Aims: Our service evaluation aims to determine the number of unnecessary inherited thrombophilia screen tests requested and to use relevant guidelines to implement an IT-based clinical decision-making algorithm intervention to reduce this number.

    Methods: Thrombophilia requests were retrospectively analysed to determine the proportion of unnecessary requests. Following this, the local order-communications system for primary care was reconfigured to promote more appropriate requesting. This included a computer-based intervention in the form of a bespoke clinical decision-support algorithm, developed in line with current national and international guidelines for inherited thrombophilia testing.

    Results: Over a 2-year period (Nov 2014-Nov 2016) 732 thrombophilia screens were requested, of which 100 (14%) tests were inappropriately requested. In the 2 months following implementation, 19 tests were requested translating to a 64% reduction in requests compared to the 2 months before and a 68% reduction compared to the same time period 12 months previously. However the proportion of inappropriate requests post intervention remained the same at 15% (4 tests).

    Conclusions: The findings confirm that a high proportion of genetic thrombophilia tests are inappropriately requested. Although the intervention did not significantly reduce the proportion of inappropriate requests it did reduce the total number of requests. Thus reducing the total number of inappropriate requests. There is the potential for greater effect to be observed with time.

  • P026 Investigating an unusual case of dual positive skin antibodies

    Kirsty A Swallow (Sheffield Teaching Hospitals, Sheffield, United Kingdom), Kristina Emsell (Sheffield Teaching Hospitals, Sheffield, United Kingdom), Graeme Wild (Sheffield Teaching Hospitals, Sheffield, United Kingdom), Ravishankar Sargur (Sheffield Teaching Hospitals, Sheffield, United Kingdom)

    Introduction: It is very rare to see positive skin antibodies to both desmosome and the basement membrane antigens. We observed apparent dual positive results using our routine test. The sample was from a lady of 100 years of age, who had symptoms of pemphigoid. We carried out further investigations to determine if there was false positivity or if this was a true dual positive result.

    Methods: Routine testing of skin antibodies was performed by indirect immunofluorescence (IIF) on monkey oesophagus tissue. Further investigations included testing the sample by IIF on monkey oesophagus tissue after ABO blocking, testing on split skin substrate and identification of the specific skin antibodies, BP230, BP180, desmoglein-1 and desmoglein-3, using transfected cell lines and ELISAs.

    Results: Results post ABO blocking indicated that antibodies to basement membrane were still positive, the presence of antibodies to desmosome was questionable. Testing on split skin substrate confirmed the presence of basement membrane antibodies, as did the BP180 antibody assays by IIF and ELISA. Desmosome antibodies were shown to be negative for desmoglein-1 and desmoglein-3 antibodies using all the additional methods. The results show that the patient was positive for basement membrane antibodies only and the initial result for desmosome antibodies was false positive.

    Conclusions: Some samples may produce ‘odd’ results for skin antibodies which need further investigations to determine if a clinically relevant antibody is present. Additional investigations such as ABO blocking can be carried out to clarify atypical results. However, ABO blocking does not exclude all possible causes of false positive desmosome antibodies. Further investigations to confirm specificity of BP230, BP180, desmoglein-1 and desmoglein-3 skin antibodies are important in providing a definitive result in cases that require further confirmation.

  • P027 Development of an in-house ImmunoCAP coating technique for the detection of pneumococcal serotype antibodies

    Kristina Emsell (Sheffield Teaching Hospitals, Sheffield, United Kingdom), Graeme Wild (Sheffield Teaching Hospitals, Sheffield, United Kingdom), Kirsty A Swallow (Sheffield Teaching Hospitals, Sheffield, United Kingdom), Ravishankar Sargur (Sheffield Teaching Hospitals, Sheffield, United Kingdom)

    Introduction: Serotype specific pneumococcal antibodies are currently measured using ELISA and Luminax technologies. We report developing an in-house method to coat ImmunoCAPs with specific pneumococcal polysaccharide to quantify pneumococcal serotype antibodies in patient sera.

    Method: Purified polysaccharide from pneumococcal serotypes 14 and 19F was used for method development. These were biotinylated using a solution of dimethyl sulfoxide (DMSO) and D-biotinoyl-є-aminocaproic acid-N-hydroxysuccinimide ester. The biotinylated polysaccharides were applied to streptavidin-coated ImmunoCAPs (Thermofisher) and incubated to allow the coupling of the polysaccharide to the solid phase. Ten replicates of the WHO international standard preparation 007sp were analysed on the coated CAPs to ascertain inter-assay precision. The samples were measured in 3 sets. Set 1 had no pre-treatment, set 2 was pre-absorbed with CWPS and set 3 was pre-adsorbed with CWPS Multi (CWPS + 22F) prior to testing. A small patient comparison study was performed to demonstrate the ability of the CAPs to distinguish between positive and negative responses.

    Results: Initial results showed that the assay could measure pneumococcal serotype-specific (14 and 19F) antibodies. The assay was able to discriminate between the positive and negative cohorts. The inter-assay %CV for the 007sp reference preparation across all 3 sets was <10%.

    Conclusions: We were successfully able to coat pneumococcal polysaccharide onto CAPs using an in-house method. Results showed that variation between each CAP was comparable with other pneumococcal antibody ELISAs. Further work is in progress to optimise the method. This includes stability work on CAP storage and further analysis into batch to batch variation. We will also need to do this for the remaining serotypes. This method could be expanded into other allergens and antigens although assay validation will need to be considered with regard to UKAS accreditation of the tests.

  • P028 Tolerating allergy testing: an audit of patterns of electronic requesting for specific IgE testing in the investigation of allergy by GPs in the West Midlands

    Abigail J Clark (Sandwell and West Birmingham Hospitals NHS Trust, West Bromwich, United Kingdom), Sadia Noorani (Sandwell and West Birmingham Hospitals NHS Trust, West Bromwich, United Kingdom), Helen Sandy (Sandwell and West Birmingham Hospitals NHS Trust, West Bromwich, United Kingdom)

    Specific IgE (sIgE) testing is a valuable tool in the investigation of allergy in primary care, providing useful information about a patient's sensitisation status to a particular allergen – but is it being misused? It’s widely acknowledged that there is no substitute for a detailed clinical history in the diagnosis of allergy, but just how far has practice shifted from targeted testing to use of “allergy screening” panels? A retrospective horizontal audit was carried out in the Immunology Department at Sandwell and West Birmingham Hospitals (SWBH) NHS Trust, following changes made to electronic requesting, to see how effectively an “other allergen” free-text option for specific IgE testing was being used by GPs in primary care. From a total of 387 specific IgE tests, with an average of 5 tests per request form (range 1-24), it was found that overall 23% of sIgE tests were positive (≥0.35 kUA/L), and 77% were negative (<0.35 kUA/L). Of the positive test results, 5% were classed as “borderline” (0.35-0.70 kUA/L), and 13% were “weak positive” (0.70-3.50 kUA/L). The three most requested sIgE tests were “house dust mite” with 35 requests, “wheat” with 28, and “peanut” with 26. This audit highlighted some key issues: the large numbers of negative results suggest sIgE testing is being used to “screen” for allergy, GPs need to be supported by the Immunology Department to provide guidance and advice about appropriate use of sIgE testing, and further changes to electronic requesting rules will be needed to be both flexible but structured to guide investigation of allergy.

  • P029 Coeliac screen in a “Tired All the Time” test profile: of any value?

    Angela Ballantyne (New Cross Hospital, Wolverhampton, United Kingdom), Anjali Ekbote (New Cross Hospital, Wolverhampton, United Kingdom), Clare Ford (New Cross Hospital, Wolverhampton, United Kingdom), Rousseau Gama (New Cross Hospital, Wolverhampton, United Kingdom)

    Coeliac disease (CD) is an autoimmune condition caused by intolerance to gluten present in wheat, rye and barley. NICE (NG20, 2015) recommends coeliac screening for patients presenting with prolonged fatigue. A coeliac screen, therefore, was added into a primary care ‘Tired All the Time’ (TATT) profile provided by Royal Wolverhampton Hospital Trust.

    Our coeliac screening involved an initial IgA anti-tissue transglutaminase (tTG). Positive IgA-tTG (≥7 U/mL) samples undergo IgA anti-endomysial antibody (EMA) testing. Undetectable IgA-tTG (0 U/mL) samples undergo IgG-tTG testing. Patients positive for IgA-tTG and IgG-tTG should be referred for small bowel biopsy (SBB).

    The TATT profile was audited 9 months after inclusion of the coeliac screen. There were 1468 TATT requests. Nineteen were IgA-tTG positive; 10 EMA positive (1 known CD; 9 referred for SBB), 3 EMA weak positive (2 referred for SBB) and 6 EMA negative (4 referred for SBB). One was positive for IgG-tTG/undectable IgA-tTG (no follow-up). It is possible that “no follow-up” patients may have been referred to other trusts for a SBB.

    Of those referred for SBB: IgA-tTG positive/EMA positive: 5 confirmed CD; 2 awaiting biopsy, 1 could not tolerate endoscopy, 1 private SBB result unavailable. IgA-tTG postive/EMA weak positive: 1 confirmed CD, 1 awaiting biopsy. IgA-tTG postive/EMA negative: 2 confirmed CD, 1 SBB negative, 1 could not tolerate endoscopy.

    Conclusions: 1) CD may present with TATT and coeliac screening in a “TATT” testing profile is of value as it is the only laboratory method of predicting those patients who should be referred for SBB. 2) Additional EMA test for all positive IgA-tTG was of no added value. 3) We have introduced improved automated interpretive comments to ensure appropriate referral for SBB and a business case has been submitted to enable alignment with NG20.

  • P030 In search of evidence-based action limits

    Jennifer S Johnstone (Ninewells Hospital and Medical School, Dundee, United Kingdom), Michael J Murphy (Ninewells Hospital and Medical School, Dundee, United Kingdom)

    Background: Communicating ‘critical’ results to the requesting clinician is an essential component of clinical authorisation in order to enable urgent action to be taken. RCPath guidance for out of hours (OOH) reporting of laboratory results to primary care acknowledges the absence of underpinning evidence. The 2010 guidelines advise that the action limits for potassium are results <2.5 mmol/L and >6.5 mmol/L. Locally, critical results from primary care that are authorised OOH are phoned to the GP NHS 24 regional hub in Dundee.

    Aim: For primary care potassium results authorised OOH, we asked the following questions: Are phoned results acted on? If so, when? What is the outcome of any action taken?

    Methods: A retrospective study from December 2015 to May 2016 inclusive was undertaken. Potassium requests from primary care were retrieved from LIMS. The potassium result was recorded along with other data retrieved from LIMS and OOH callsheets. Data were analysed for potassium results that were validated OOH (1800-0800).

    Results: Two hundred and twenty potassium results (<3.1 mmol/L and > 5.9 mmol/L) from primary care were validated OOH; 27 were phoned to the OOH hub due to the potassium result; 7 were phoned due to other results; 16 patients were referred to hospital OOH due to the potassium results, with all other results being acted on albeit not OOH. For the 16 patients referred to hospital OOH, 6 results confirmed the original abnormal result.

    Conclusion: Critical potassium results are phoned urgently and are acted on, although not always OOH. If a result is phoned, the most frequent action is to refer to hospital OOH. Different actions occur for the same potassium results indicating that the actions are dependent on the patient and the clinician. The total number of results phoned to OOH was too small for any evidence base to be established.

  • P031 Demand management initiatives to reduce follicle stimulating hormone requesting for suspected menopause in primary care

    Jane Armer (East Lancashire Hospitals NHS Trust, Blackburn, United Kingdom), Kathryn Brownbill (East Lancashire Hospitals NHS Trust, Blackburn, United Kingdom)

    Introduction: The recent NICE guidance on menopause (NG23) states that follicle stimulating hormone (FSH) testing should not be performed routinely in the diagnosis of menopause in women aged over 45 years. Instead, menopause should be diagnosed clinically based on vasomotor symptoms and the absence of periods for >12 months. This was included in the recommendations from Choose Wisely whose aim is to improve patient/clinician conversations to guide appropriate testing.

    Following publication of the NICE guidance, we produced a single page information sheet for GPs about the menopause and indications for FSH testing. This was agreed with the GP leads of our 2 local CCGs and distributed to all GPs during September 2016. In addition, the Duty Biochemist has been appending comments based on NG23 where appropriate to FSH results in women >45 years.

    Aim: To determine the clinical effectiveness and sustainability of the demand management initiatives on the number of FSH requests from primary care.

    Methods: The number of primary care FSH requests for all females and females >45 years during August 2016, October 2016 and December 2016 was extracted from the laboratory information system.

    Results: Using the number of requests from August as a baseline, the total number of FSH requests for females from primary care decreased by 5% and 29% in October and December respectively. For females >45 years, the number of FSH requests decreased by 15% and 41% in October and December respectively. If the December 2016 workload is sustained, the annual cost saving is over £3000.

    Conclusions: We have demonstrated a significant reduction in FSH testing by providing a clear, concise information sheet to GPs and appending clinical comments to FSH results. We have had excellent feedback from individual GPs. This reduction in FSH testing has been fed back to the local CCG leads and similar demand management initiatives are planned.

  • P032 Referrals to chemical pathology in a Belfast teaching hospital

    Paul Hamilton (Belfast Health and Social Care Trust, Belfast, United Kingdom), Giles Aldworth (Belfast Health and Social Care Trust, Belfast, United Kingdom), Brona Roberts (Belfast Health and Social Care Trust, Belfast, United Kingdom)

    The nature of referrals dealt with by chemical pathologists is likely to vary widely depending on local policies and the availability of other specialties within a hospital trust. A computerised database was used to capture all calls for clinical advice to chemical pathology in the Belfast Health and Social Care Trust between 1st August 2015 and 1st August 2016. The nature and source of the referrals were analysed. Five hundred and fifty one unique patients were referred in the 1 year period, many of whom were discussed on several occasions, usually with specialty registrars. Sixty-four patients were being treated in the community, while 487 were in hospital. The commonest reason for referral was a problem relating to serum sodium (63.3% of referrals), with hyponatraemia being much more common than hypernatraemia. Other reasons for referral were problems with phosphate (4.7%), multiple electrolytes (4.5%), acid-base (4.2%), calcium (4.2%), magnesium (4.0%), potassium (3.6%), liver enzymes (1.5%), lipids (1.3%), kidney function (0.7%), vitamin D (0.7%), inherited metabolic disorders (0.7%), and other problems (6.5%). Trauma/orthopaedics, haematology, oncology and medical wards were the most frequent sources of referrals. The use of a referrals database has been helpful in improving record-keeping and handover, and in the identification of units that might benefit from outreach teaching sessions. It also now forms the basis for a weekly consultant-led “virtual ward round” in which trainee decisions are reviewed.

  • P033 Validation of a serum creatinine based electronic (e)-alert system for acute kidney injury in a paediatric cohort

    Jennifer Holmes (Cwm Taf University Health Board, Cardiff, United Kingdom), Gethin Roberts (Hywel Dda University Health Board, Aberystwyth, United Kingdom), Kate May (Cwm Taf University Health Board, Cardiff, United Kingdom), Kay Tyerman (Leeds Teaching Hospital NHS Trust, Leeds, United Kingdom), John Geen (Prince Charles Hospital, Merthyr Tydfil, United Kingdom), John D Williams (Cardiff University, Cardiff, United Kingdom), Aled O Phillips (Cardiff University, Cardiff, United Kingdom)

    Background/Aims: Based on a presumption that early identification may help raise standards of care, an automated electronic (e)-alert system for acute kidney injury (AKI) based on KDIGO criteria has been implemented nationally across the National Health Service in Wales. The system uses patients’ historical biochemical results to define baseline serum creatinine (bSCr). This approach is limited for a paediatric cohort in which the absence of previous biochemical results is common. We aimed to validate alternative models of defining bSCr. We also modelled a modification of KDIGO criteria proposed to overcome concerns relating to ‘over-diagnosis’ of AKI in neonates.

    Methodology: Data was collected on 2,390 e-alerts. To compare alternative bSCr definitions with the algorithm method, we used estimated creatinine clearance criteria (eCCl120) and midpoint normative values (NormMid) for age and gender. We applied a modification to the e-alert system which involved suppressing alerts by neonates with a bSCr <0.5 mg/dL that did not rise to >0.5 mg/dL.

    Results: eCCl120 resulted in an estimated bSCr significantly lower than that obtained from patients’ historical results (non-neonates: 0.40 mg/dL vs. 0.52 mg/dL; neonates: 0.17 mg/dL vs. 0.43 mg/dL). NormMid more closely approximated bSCr (non-neonates: 0.63 mg/dL vs. 0.52 mg/dL; neonates: 0.38 mg/dL vs. 0.43 mg/dL). For non-neonates Bland-Altman analysis demonstrated equal unit bias indicating an equal degree of agreement between Algorithm-eCCl120, and Algorithm-NormMid. A positive percentage bias for Algorithm-eCCl120 confirmed that eCCl120 underestimates SCr. There was a stronger agreement between Algorithm-NormMid than Algorithm-eCCl120 for neonates. 69.6% of neonatal alerts were by a bSCr <0.5 mg/dL. Of these 52.4% represented rises to >0.5 mg/dL and 47.6% (33.1% of all neonatal alerts) rises to ≤0.5 mg/dL, for which there was no association with mortality or renal impairment.

    Conclusion: When historical biochemical results are unavailable, using the NormMid method to define bSCr is favoured over the eCCl120 method when diagnosing paediatric AKI. In neonates, rises in SCr to ≤0.5 mg/dL from values <0.5 mg/dL may represent ‘over-diagnosis’ of AKI.

  • P034 Baseline biochemistry in recurrent renal stone patients undergoing screening for distal renal tubular acidosis: who should we test?

    Joseph M Taylor (Royal Liverpool University Hospital, Liverpool, United Kingdom), Suzannah G Phillips (Royal Liverpool University Hospital, Liverpool, United Kingdom), Vinita Mishra (Royal Liverpool University Hospital, Liverpool, United Kingdom)

    Background: Kidney stones are a common complaint with highest incidence in Caucasian males between 40 and 60 years. Without intervention recurrence can be 50% within 10 years. Disturbances in urinary pH, for example distal renal tubular acidosis (dRTA), predispose patients to stone formation and may alter management.

    We aimed to investigate if patients were being appropriately tested at baseline, identify patients with possible dRTA and compare their biochemistry to non-dRTA patients.

    Methods: Clinic referrals for recurrent renal stones were investigated according to in-house protocols including baseline biochemical investigation and furosemide acidification testing to examine stone risk factors. Results from these investigations were collected and analysed. Data was separated into possible dRTA (failure to acidify urine pH <5.5) and no dRTA for comparison.

    Results: Data was collected on 34 patients (mean age 57 years, 86% male); 44% of patients were investigated using all tests in the protocol. Discounting a recently added test, 86% of patients were investigated fully. Six patients had possible dRTA based on furosemide test results. No statistically significant differences between groups for the 18 baseline biochemical parameters were seen apart from urine pH (p=0.03) and serum bicarbonate (p=0.02). Mean urine pH was higher and mean serum bicarbonate lower in possible dRTA cases.

    Conclusion: Despite statistical differences in baseline urine pH and serum bicarbonate between groups, no parameters analysed in this study are useful in deciding which patients should undergo urine acidification testing. As dRTA is a risk factor for recurrent stone formation and some data suggests patients may require different management, it is the recommendation all patients presenting with recurrent renal stones should undergo this test. Further study will involve follow up of all patients initiated on pharmacotherapy with potassium citrate to compare outcomes and changes in biochemistry between possible dRTA and the no dRTA groups.

  • P035 Accuracy and reproducibility of different urine collection methods to assess albuminuria and the optimal time to collect spot samples

    Daniel Chapman (University of Exeter Medical School, Exeter, United Kingdom), Kim Gooding (University of Exeter Medical School, Exeter, United Kingdom), Timothy McDonald (Royal Devon and Exeter NHS Foundation Trust, Exeter, United Kingdom), Angela Shore (University of Exeter Medical School, Exeter, United Kingdom)

    Introduction: Increased urinary albumin excretion (UAE) is an early indicator of kidney complications in diabetes and cardiovascular disease in healthy individuals. The relationship between UAE and vascular dysfunction is continuous and commences within the clinically defined “normal” albuminuria range. To investigate these relationships the variability of UAE needs to be understood in the normal range.

    The “gold-standard” measure of albuminuria is albumin excretion rate (AER) from a timed overnight collection, but for convenience spot (untimed) urine collections are often used that report the albumin to creatinine ratio (ACR). This work investigates the accuracy and reproducibility of UAE in different collections to determine what time of collection of spot samples is the best surrogate for timed collections in the normal range; first morning void (FMV), second morning void (SMV) or random.

    Methods: Seventeen healthy participants completed three study visits. On each visit they collected all of their urine in different containers over a 36 hour period. For each collection method the mean of three repeats was calculated for each participant, the median value [interquartile range] among participants was reported. Median intra-individual reproducibility of collection methods (coefficient of variation calculated from three collections) and Spearman’s rank correlation of ACR in spot samples with timed overnight AER were determined.

    Results: Median AER was 4.5 µg/min (2.7-6.9) during the day and 3.2 µg/min (1.2-4.0) overnight (p=0.105). Median ACR in FMV, SMV and random samples was 3.1 mg/g (2.3-4.5), 4.8 mg/g (2.8-6.6) and 4.8 mg/g (2.8-7.2) respectively. Intra-individual reproducibility of ACR in FMV, SMV and random spot samples was 15.9%, 22.1% and 22.6% respectively. Correlation of ACR in FMV, SMV and random spot samples with timed overnight AER was 0.88, 0.54 and 0.59 respectively.

    Conclusion: FMV has the best reproducibility and strongest correlation with the overnight AER. A FMV sample is the best untimed surrogate of an overnight collection when investigating the “normal” albuminuria range.

  • P036 The impact of exercise on albuminuria

    Daniel Chapman (University of Exeter Medical School, Exeter, United Kingdom), Kim Gooding (University of Exeter Medical School, Exeter, United Kingdom), Timothy McDonald (Royal Devon and Exeter NHS Foundation Trust, Exeter, United Kingdom), Angela Shore (University of Exeter Medical School, Exeter, United Kingdom)

    Introduction: An increase in urinary albumin to creatinine ratio (ACR) is an early indicator of kidney complications in diabetes and cardiovascular disease in healthy individuals. A transient increase in ACR after exercise has been identified in diabetic patients, but it is not known if it occurs in healthy individuals. This work investigates at what exercise threshold ACR increases, and how long it takes to return to normal in healthy individuals.

    Methods: Eleven participants completed four study visits of different exercise intensities; mild, moderate, heavy and none (resting). Exercise intensity was randomised for each visit. To ensure all participants were tested at the same intensity, relative to their capacity for exercise, personalised exercise thresholds were calculated using an individual’s VO2 max and gas exchange threshold results, determined from an incremental ramp test. During each visit participants collected every void of urine in separate containers over a 36 hour period. Exercise was carried out on a cycle ergometer for a maximum of 30 minutes or until exhaustion.

    Results: Heavy exercise resulted in a significant increase in ACR (inter quartile range) from 0.85 mg/mmol (0.62-1.49) at rest to 3.55 mg/mmol (1.44-7.00) (p=0.003). After heavy exercise, 64% of participants had an increased ACR in the clinically defined microalbuminuria range. The maximum duration of increased ACR after heavy exercise was 304 minutes. Heavy exercise was compared to cycling up a steep hill, all participants reached exhaustion after a mean (±SD) duration of 17 (±7) minutes. There was no significant difference in the change in ACR after mild (p=0.80) and moderate (p=0.13) exercise when compared to resting.

    Conclusion: In healthy individuals heavy exercise increases ACR sufficiently enough to place 64% of participants into the microalbuminuria range. Thus healthy individuals should not provide a urine sample for ACR measurement if they have undertaken heavy exercise within the previous 5 hours.

  • P037 Measurement of hepcidin 25, 22 and 20 by LC-MS/MS: generation of reference ranges in healthy individuals and in patients with chronic kidney disease requiring dialysis

    Kamaljit Chatha (University Hospitals Birmingham NHS Foundation Trust, Birmingham, United Kingdom), Alexander Lawson (Heart of England Foundation Trust, Birmingham, United Kingdom), Briony Johnson (Heart of England NHS Foundation Trust, Birmingham, United Kingdom), Craig Webster (Heart of England NHS Foundation Trust, Birmingham, United Kingdom)

    Background: Hepcidin is a 25-amino acid peptide hormone which has emerged as the major regulator of dietary iron absorption and cellular iron release. The measurement of hepcidin levels may therefore be a valuable diagnostic tool in the investigation and management of a number of disease states including haemochromatosis and anaemia. Here we present the first LC-MS/MS method able to measure hepcidin-25, 22 and 20 in serum and the generation of reference ranges in healthy individuals and in patients with chronic kidney disease (CKD) requiring dialysis.

    Methods: Serum samples were prepared by solid phase extraction using [13C18,15N3]-hepcidin as an internal standard. LC-MS/MS analyses were performed on a Shimadzu HPLC system interfaced to a API4000 mass spectrometer. Chromatographic separation was achieved using a Kinetex C-18 100 Å column (2.6 µm, 30 x 3.0 mm). Reference ranges were generated by measurement of hepcidin isoforms in 114 men and 118 women and 34 CKD patients prior to dialysis.

    Results: All compounds were separated with no interference observed between fragments. Intra- and inter-batch imprecision for the hepcidin isoforms ranged from 4.1 to 11.2% at concentrations of 7, 70, and 150 nmol/L, while relative error ranged from -3.3 to 11.6%. Hepcidin 25 and 22 were linear from 1-200 nmol/L with hepcidin 20 linear from 2.5-200 nmol/L. Reference ranges for hepcidin 25 were 0.5-20.3 nmol/L in healthy women, 0.5-23.6 nmol/L in healthy men with levels significantly higher in dialysis patients (median 36.6 nmol/L, IQR 36.7, p<0.0001). Hepcidin 20 and 22 were only detectable in patients with CKD yielding median values of 2.1 (IQR 2.2) and 4.5 (IQR 5.7) nmol/L respectively.

    Conclusion: We have developed an accurate and sensitive method to measure hepcidin 25, 22 and 20 in human serum by LC-MS/MS, which has been fully validated and used to generate of reference ranges in healthy individuals and in patients with CKD requiring dialysis.

  • P038 Laboratory evaluation of intravascular haemolysis

    Jennifer S Johnstone (Ninewells Hospital and Medical School, Dundee, United Kingdom), Michael J Murphy (Ninewells Hospital and Medical School, Dundee, United Kingdom)

    Background: Intravascular haemolysis is an important clinical entity that occurs in a variety of clinical settings. Laboratory evaluation involves measurement of haemoglobin (Hb), lactate dehydrogenase (LDH), reticulocyte count (retics) and haptoglobin (Htg). It is not clear how often the suspicion of intravascular haemolysis is confirmed, or the concordance of the findings of different laboratory markers in its confirmation.

    Aim: The purpose of the current study was to establish how often the suspicion of intravascular haemolysis is confirmed, and to examine the frequency of abnormal results in haemolytic screens (including Hb, LDH, retics, Htg).

    Methods: Two hundred and forty full screens for intravascular haemolysis were performed over a two-year period. Hb and retics were measured on Advia 2120. Hb was determined by colorimetric method whilst retics were stained and detected using flow cytometry. LDH was analysed on Advia 2400 using Lactate/NAD principle and Htg was measured using a nephelometric method on Dimension Vista.

    Results: Anaemia was found in 196/240 (82%). One or more of raised LDH, raised retics or low Htg was present in 192/240 (80%). Of these three abnormalities, one was present in 90/240 (37.5%), two were present in 67/240 (28%), and all three in 35/240 (14.6%). Raised LDH was found in 128/240 (53%), raised retics in 126/240 (52%) and low Htg in 75/240 (31%).

    Conclusion: Most screens for intravascular haemolysis are performed in patients who are anaemic; this is consistent with anaemia being a key trigger for the request. There is at least some evidence of haemolysis (at least one abnormal finding) in a large majority, although only a minority display comprehensive evidence (all three findings abnormal). Low haptoglobin is detected less frequently than either raised LDH or raised retics.

  • P039 An underlying cause of iron overload in a bone marrow transplant patient

    Laura E Tooth (South West London Pathology, London, United Kingdom), Matthias Klammer (South West London Pathology, London, United Kingdom)

    A 25 year-old female presented to her local A&E department with blurred vision. On examination she was found to have severe anaemia (Hb 37 g/L), a raised white blood cell count (338x109/L) and lymphadenopathy. Her blood film showed a predominance of blasts and scanty cytoplasm.

    Peripheral blood immuno-phenotyping confirmed a diagnosis of B-acute lymphoblastic leukaemia, with a bone marrow aspirate showing 95% bone marrow infiltration. Cytogenetic studies confirmed a high risk recurrence translocation between chromosome 4 and 11. She responded well to initial chemotherapy, but a more intensive regime was started due to residual disease detected post treatment. The patient underwent a HLA matched sibling allogenic transplant, with full intensity chemotherapy and radiotherapy. She had normal liver function tests post-transplant, but following weaning off immuno-suppressants, she developed rapid onset transaminitis and hyperbilirubinaemia, which could have been caused by direct action from the drugs, graft-versus-host disease, or infection or macrophage activation. A liver biopsy confirmed graft-versus-host disease of the liver (stage 3), but also revealed evidence of iron overload, with a serum ferritin concentration measured as 14,526 µg/L and a transferrin saturation of 90%. Iron overload is a serious complication of bone marrow transplantations due the requirement for multiple transfusions, but the ferritin in this patient was disproportionately raised and other causes were considered. Genetics showed the patient was a carrier for the C282Y mutation in the HFE gene. As the blood sample was taken post-transplant, the HFE mutation was confirmed in pre-transplant DNA. A buccal swab failed in this patient, but would have been the most appropriate sample type. The patient was treated with iron chelation therapy to remove the excess iron from her body, because her haemoglobin was too unstable for regular venesections, and she was re-started on immuno-suppressants.

  • P040 The introduction of an anaemia screen: a successful example of demand management?

    Martin K Langan (Countess of Chester Hospital NHS Foundation Trust, Chester, United Kingdom), Arvind Pillai (Countess of Chester Hospital NHS Foundation Trust, Chester, United Kingdom), Shirley A Bowles (Countess of Chester Hospital NHS Foundation Trust, Chester, United Kingdom)

    Background: Serum Vitamin B12 and folate requests increased by 154% between 2009/10 and 2015/16. It was apparent that these tests had become first-line screening tests, requested alongside an initial full blood count (FBC) and, often, a serum ferritin. Less than 10% of the results reported were abnormally low. In March 2016, the option of an anaemia screen was added to the laboratory electronic requesting system. This requires collection of an EDTA specimen and a serum sample, but the need for B12/folate or ferritin measurement is determined by an algorithm embedded in the laboratory computer system, which uses the results of the FBC.

    Aims: 1) To determine the impact of the anaemia screen on workload. 2) To ensure that the algorithm is operating as envisaged.

    Results: The number of B12/folate and ferritin measurements has fallen by 25% and 20% respectively. Reviewing the results from one month: based on the FBC, appropriate reflex measurements were done in all cases. Where an anaemia screen was selected, 9.9% of females and 26.1% of males were anaemic. The algorithm reflex requested ferritins in 50 patients (27 females, 23 males): this was low in 56% of males and 43.5% of females, compared to 11.9% and 30.2% respectively in patients, where the anaemia screen had not been selected. For 19 patients, B12/folate measurements were added by the algorithm: 16% had a low B12 and 11% a low folate, compared to 14% and 2.4% in the control group.

    Conclusions: 1) The introduction of the anaemia screen has had an impact on workload and costs (reagent cost saving ~£25,000). There is scope for further savings, if more clinicians can be encouraged to choose this option, rather than block requesting FBC, B12/folate and ferritin. 2) In the context of suspected anaemia, serum B12/folate and ferritin measurements are being done in appropriate circumstances, with improved diagnostic yield. 3) The anaemia screen facilitates rational, stepwise testing, but without the need for a second consultation/re-bleeding of the patient.

  • P041 Reflex testing of serum electrophoresis: are we doing too many?

    Victoria Ware (Aneurin Bevan University Health Board, Abergavenny, United Kingdom), Anthony Jackson (Aneurin Bevan University Health Board, Newport, United Kingdom), Fiona Stratford (Aneurin Bevan University Health Board, Newport, United Kingdom)

    Background: Over 10,000 serum electrophoresis (“paraprotein screens”) are requested in our trust every year. The large majority are conducted on myeloma patients. In patients without myeloma we wanted to assess if additional discretionary testing by the duty biochemist was significantly adding to the number of paraprotein screens performed.

    Our aims were: 1) to identify how many paraprotein screens were added in a 6 month period, 2) to identify the outcome of those paraprotein screens and 3) to use the data to propose indications for reflex testing for electrophoresis.

    Methods: A retrospective audit of all protein electrophoresis tests requested between January and June 2016 was performed. Data was collected from the laboratory information management system. Paraprotein screens added by the duty biochemist were identified and further clinical information gathered.

    Results: Four thousand two hundred and twenty-five paraprotein screens were performed over a 6 month period in patients without a known paraprotein; 298 (7%) were additional tests added at the discretion of the duty biochemist. Over 6 months this averaged approximately 2 a day. Sixteen new paraproteins were identified, the remaining patients results included: polyclonal increase in gamma globulins, hypogammaglobulinaemia and no abnormality detected.

    New myeloma patients could not be identified purely on the basis of raised globulins and differentiated from the other causes of raised globulins. In patients with normal globulin levels, reflexing paraprotein screens due to small rises in immunoglobulins did not yield diagnosis of myeloma.

    Conclusions: Discretionary testing is a useful practice with a detection rate for new paraproteins (5.4%) similar to previous studies. To reduce the number of tests performed we suggest only reflexing serum electrophoresis if there is a discrete rise in either IgG (>15 g/L), IgA (>10 g/L) or IgM (>10 g/L). This approach should maximise the detection of clinically significant paraproteins.

  • P042 Current antenatal screening methods for iron deficiency in Leeds and nationally results in the condition being underdiagnosed

    Michael R O'Sullivan (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Etienne Ciantar (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Julian H Barth (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom)

    Leeds and Bradford are two adjacent cities in West Yorkshire and their approach to the diagnosis and management of iron deficiency appears different despite clear guidelines issued by the British Committee for Standards in Haematology. The two approaches were compared and the Leeds secondary care team were not following the UK guidelines. As an incidental consequence of the Leeds approach there was an identification of a small cohort of patients who had iron deficiency based on their ferritin results but did not have associated anaemia. The diagnosis of iron deficiency in this cohort would be missed by the current antenatal screening methods adopted by Bradford and those following the national guidelines. This study tested how common the problem of non-anaemic iron deficiency is in Leeds and the resulting data casts doubt on the reliability of the current screening methods for the diagnosis of iron deficiency in pregnancy. This study shows that 32.6% of 245 consecutive pregnancies in Leeds have ferritin <15 µg/L and are classed iron deficient. Conversely the same patients being screened using the UK guidelines which requires the measurement of haemoglobin, only 6.9% of these pregnancies are classed iron deficient. There is therefore a large cohort of patients with iron deficiency who are not being detected. The aim of the study was to confirm that Leeds was indulging in unnecessary testing of ferritin and in conclusion the opposite was apparent. The data suggests that all pregnancies should be tested for ferritin as an adjunct to the full blood count which is collected at booking and at 28 weeks gestation. Anaemia was found to be more common in Bradford (7.1%) than Leeds (3.1%). The normal range for platelet count during pregnancy was found to be 136-378x109/L, lower than the normal range for non-pregnancy.

  • P043 Hypogonadotropic hypogonadism with infertility in type 2b juvenile haemochromatosis: recovery with iron depletion

    Savinda Weerasinghe (New Cross Hospital, Wolverhampton, United Kingdom), Alan MacWhannell (New Cross Hospital, Wolverhampton, United Kingdom), San San Min (New Cross Hospital, Wolverhampton, United Kingdom), Rousseau Gama (New Cross Hospital, Wolverhampton, United Kingdom)

    Juvenile haemochromatosis is rare with haemojuvelin mutations (type 2a) and HAMP mutations encoding hepcidin (type 2b) accounting for 90% and 10% of cases respectively. People with these mutations are unable to make hepcidin and therefore cannot inhibit iron absorption even when iron replete. Excess iron is deposited in various tissues and organs leading to predominantly cardiomyopathy, liver disease and “irreversible” hypogonadotropic hypogonadism. We report a 19 year old male of Pakistani origin presenting with an asymptomatic left varicocoele and was found to have isolated hypogonadotrophic hypogonadism (testosterone 0.8 nmol/L, LH 0.2 IU/L, FSH 0.3 IU/L and azoospermia) and increased liver transaminase activity (ALT 435 IU/L). A pituitary MRI, chronic non-metabolic liver screen and abdominal ultrasound were normal. One year later, after further investigations initiated by laboratory personnel, a diagnosis of juvenile haemochromatosis (transferrin saturation 93%); ferritin 5686 ng/mL) due to a HAMP gene mutation was made, emphasising the importance of a proactive diagnostic laboratory role. Echocardiography and NT-proBNP were normal. On family screening, his mother, father and brother were heterozygotes for the mutation without evidence of iron overload. His sister has since been diagnosed with juvenile haemochromatosis due to the same mutation. The hypogonadism was managed with testosterone replacement and regular venesection over 9 years and led to normalisation of liver function tests and iron studies. He then married and was concerned about fertility. Following withdrawal of testosterone, reassessment of gonadal function showed Leydig cell recovery and partial Sertoli cell recovery (testosterone 10 nmol/L, LH 3.2 IU/L, FSH 3.8 IU/L and mild oligospermia). This has opened up the prospect, if required, of assisted conception with intracytoplasmic sperm injection. Reversal of hypogonadotropic hypogonadism in type 2b juvenile haemochromatosis following iron depletion has not, to our knowledge, been previously reported.

  • P044 Vitamin D binding protein: the missing piece of the puzzle?

    Hazel Borthwick (County Durham and Darlington NHS Trust, Darlington, United Kingdom)

    Vitamin D requesting within NHS laboratories in the UK has significantly increased following renewed interest in this hormone. Examining the relationships between in-house routine markers of bone metabolism including serum 25-hydroxyvitamin D and additional markers such as 1,25-hydroxyvitamin D, serum and urinary vitamin D binding protein may help to identify better markers of vitamin D deficiency and improve utilisation of resources.

    A vitamin D binding protein assay was validated as part of this study to facilitate the measurement of this marker in the serum and urine of 44 general practice and outpatient participants. Serum and urine vitamin D binding protein, serum 1,25-hydroxyvitamin D and serum 25-hydroxyvitamin D were then analysed alongside in-house routine markers of bone metabolism (calcium, parathyroid hormone, alkaline phosphatase and phosphate) in an attempt to understand why routine markers of bone metabolism can remain within range when serum 25-hydroxyvitamin D is abnormal and investigate if there is a more suitable or surrogate marker of deficiency.

    A commercially available enzyme-linked immunosorbent assay (ELISA) for the measurement of vitamin D binding protein in serum and urine was validated and found to be fit for purpose. No significant differences were observed for serum 25-hydroxyvitamin D, 1,25-hydroxyvitamin D, serum or urine vitamin D binding protein between those who had within range in-house analytes of calcium homeostasis and those with out of range results. No in-house marker was identified as a suitable surrogate for serum vitamin D analysis.

    The relationships between serum 25-hydroxyvitamin D, serum 1,25(OH)2D, serum and urinary vitamin D binding protein and in-house measures of calcium homeostasis were investigated but no associations observed that could explain low vitamin D concentrations in people whose other markers of calcium metabolism are within normal limits. Serum 25-hydroxyvitamin D remains the most suitable marker to assess deficiency.

  • P045 A case of treatment-resistant hypocalcaemia in a myeloma patient on acyclovir

    Kirsty J Flowerday (Liverpool Clinical Laboratories, Liverpool, United Kingdom), Anna Milan (Liverpool Clinical Laboratories, Royal Liverpool and Broadgreen University Hospitals, Liverpool, United Kingdom), Andrew Davison (Liverpool Clinical Laboratories, Liverpool, United Kingdom)

    Background: Hypocalcaemia can present clinically with tetany, paraesthesia, and cardiac arrhythmias. The most common causes include hypoparathyroidism, vitamin D deficiency or abnormalities in vitamin D metabolism. Derangement of calcium homeostasis is common in myeloma however patients typically present with hypercalcaemia due to increased bone resorption.

    Case Description: A 55 year-old male with myeloma presented with an adjusted calcium of 1.89 mmol/L. Previous calcium had been normal since 2013 (AdjCa = 2.31 mmol/L). Magnesium concentration at presentation was 0.60 mmol/L so the patient was commenced on magnaspartate and AdCal. Calcium measured the following day was 1.71 mmol/L so he was administered 60 mL of 10% calcium gluconate and 16 mmols of i.v magnesium. Nonetheless calcium dropped further to 1.62 mmol/L. 25(OH) vitamin D was measured and found to be 25 nmol/L. The patient received a further 60 mL of calcium gluconate and was started on Adcal-D3 (2 tablets TDS) and 250 ng of calcitriol. Calcium dropped further still to 1.42 mmol/L and the patient exhibited ECG changes.

    Calcitriol was increased to 500 ng and, over 7 days, the calcium rose to 2.22 mmol/L. Further investigation showed the patient was prescribed prophylactic acyclovir, an anti-viral drug used to prevent shingles in patients taking Velcade. Case reports in the literature describe an increased incidence of vitamin D deficiency in patients taking antiretrovirals. The exact mechanism is unclear but the induction of 24-hydroxylase and inhibition of 25- and 1-α hydroxylases have been proposed.

    Conclusion: This case of treatment-resistant severe symptomatic hypocalcaemia highlights the importance of measuring vitamin D when investigating hypocalcaemia. As the calcium continued to fall until the patient was started on calcitriol, this suggests an abnormality in vitamin D metabolism as the cause of the hypocalcaemia. The acyclovir could have been a contributing factor and is an important consideration in patients taking anti-viral medication.

  • P046 An equation for albumin-adjusted total serum calcium from a Singaporean hospital-based Asian population

    Chin-Pin Yeo (Singapore General Hospital, Singapore, Republic of Singapore), Stephanie Fook-Chong (Singapore General Hospital, Singapore, Republic of Singapore), Homathy Sivakumar (Singapore General Hospital, Singapore, Republic of Singapore), Chong Jin Gea (Singapore General Hospital, Singapore, Republic of Singapore), Sau Yeng Wong (Singapore General Hospital, Singapore, Republic of Singapore)

    Introduction: Published equations for albumin-adjusted total serum calcium vary depending on the assay methodology for albumin and calcium and the population from which the equation is derived. It is generally recommended that laboratories looking to report adjusted calcium should derive local equations or validate existing ones to ensure fit for use. Our study aimed to derive and validate an equation based on our local hospital-based Asian population, using arsenazo (calcium) and bromocresol purple (albumin) methodologies.

    Methods: Paired albumin and calcium results, produced by our laboratory’s Beckman-Coulter AU5800 analyser, were extracted from the laboratory information system over a 3-month period for subjects with nil or limited evidence of disturbances of calcium homeostasis; based on clinical service source and blood indices. The dataset was split into derivation (n=1743) and validation (n=582) subsets. Albumin-adjusted equation was derived via linear regression analysis. The equation was validated via a bootstrapping procedure; and cross-validated by applying the derived equation to the validation subset and comparing the difference in r2 of the two subsets. We also compared the performance of our derived equation against the commonly used “text-book” equation, adjusted calcium = (total calcium) + 0.02[40 – (albumin)], on a separate cohort (n=2230).

    Results: Linear regression analysis yielded equation for adjusted calcium = (total calcium) +0.018[39.6 – (albumin)]. Bootstrap analysis showed good correlation with the derived equation. Cross-validation study revealed a shrinkage value of 0.0202 (optimal: <0.1), indicating the derived equation was stable. The derived and “text-book” equations categorised calcium status of patients with albumin 30-40 g/L to an agreement of 97% and patients with albumin <30 g/L to an agreement of 92%.

    Conclusions: Implementation of this locally derived albumin-adjusted total serum calcium equation hospital will avail our clinicians, who are currently using the “text-book” equation, to albumin-adjusted total serum calcium results that are specific for our laboratory’s test methodologies and clinical setting.

  • P047 Adjusted calcium: local application of observations from big data

    Emma L Dewar (Aberdeen Royal Infirmary, Aberdeen, United Kingdom), Douglas Robertson (Aberdeen Royal Infirmary, Aberdeen, United Kingdom), Ian Rothnie (Aberdeen Royal Infirmary, Aberdeen, United Kingdom), Helen Regan (Aberdeen Royal Infirmary, Aberdeen, United Kingdom), Katie Booth (Aberdeen Royal Infirmary, Aberdeen, United Kingdom), Mark Lum (Aberdeen Royal Infirmary, Aberdeen, United Kingdom), Fiona M Brandie (Aberdeen Royal Infirmary, Aberdeen, United Kingdom), William Simpson (Aberdeen Royal Infirmary, Aberdeen, United Kingdom), Bernard Croal (Aberdeen Royal Infirmary, Aberdeen, United Kingdom)

    Background: Large variation currently exists across pathology services as to the method and choice of equation for adjusted calcium calculations. Recent observations from big data have demonstrated clear age and sex related differences in serum albumin levels which would have significance for choice of adjusted calcium equation in these population sub groups.

    Aim: To use guidance on adjusted calcium equation generation along with knowledge from big data to construct a multi-equation approach to this problem.

    Methods: LIMS data was collected over a four-month period (14 April – 24 August 2015). Primary care results with a measured serum albumin and calcium were pulled using the following exclusion criteria (where available) of age (<18 years), calcium (<2.0 or >2.7 mmol/L), albumin (<20 or >50 g/L), creatinine (>200 µmol/L), urea (>15 µmol/L), ALT (>55 U/L), vitamin D (<25) and PTH (<1.7 or >9 pmol/L). Results, divided into gender and 10-year age categories (from 18 to 70 years), were used to derive suitable adjusted calcium equations for each of these groups. A comparison of the distribution of adjusted calcium using this multi-equation approach was made with the existing single adjusted calcium equation technique.

    Results: The existing single corrected calcium equation technique demonstrated significant misalignment with the target reference interval of 2.2-2.6 mmol/L – with over diagnosis of hypocalcaemia being a particular concern. The multi-equation approach using 12 distinct sex/age related equations allowed much better alignment with all 95% C.I. distributions fitting within the target reference range.

    Conclusions: The multi-equation approach for adjusted calcium calculation more closely aligns to targeted reference intervals and minimises inappropriate classification of calcium status. This strategy should replace the current single equation approach in order to better optimise direct and consequential costs to patient care and NHS finances.

  • P048 Bone health and stress fractures in elite female ballet dancers

    Shilpa R Patel (Imperial College Healthcare NHS Trust, London, United Kingdom), Richard L Abel (Imperial College London, London, United Kingdom), Alan P Courtney (Imperial College Healthcare NHS Trust, London, United Kingdom)

    Background: Elite female athletes, such as classical ballet dancers, have a higher incidence of stress fracture compared to the general population. Understanding the processes that regulate bone turnover in classical ballet dancers may lead to the development of new biomarker profiles that can be used to improve bone health in these individuals and thus reduce the incidence of stress fractures.

    Methods: Eleven female age matched subjects; 6 healthy controls, 5 ballet dancers. Urine samples, training load measurements were taken at baseline, 2 weeks and 4 weeks. Daily activity and dietary records were completed. The bone marker N-telopeptide of type 1 collagen (NTx) and oestrogen metabolites were measured in urine.

    Results: Differences in training load and diet between controls and ballet dancers were observed. No significant difference was observed in bone turnover, although an increased NTx result identified one dancer who had suffered a stress fracture. Ballet dancers were seen to have increased 2-hydroxyoestrogen (2OHE) compared to controls (p=0.0198), however 16a-hydroxyoestrogen (16aOHE) and the 2:16 metabolite ratio did not show any significant difference.

    Conclusions: NTx may be a useful marker to identify individuals with stress fractures, although it is surprising that the ballet dancers did not have a reduced bone turnover compared to controls (as with other elite athletes). Lifestyle factors can modulate oestrogen metabolism. Females with predominance of the 2OHE over the 16aOHE oestrogen metabolism pathway have been reported to be associated with increased risk of osteoporosis, suggesting a possible reduction of circulating oestrogen in individuals, such as these ballet dancers, and thus detrimental to bone health. Future work should focus on recruitment of a larger sample size, examining the effects of nutrition, mechanical loading and monitoring of dancer urine/serum biomarker measurements, to fully examine the relationship between metabolites, bone remodelling and stress fracture injuries, to optimise bone health.

  • P049 Are there South Asian-white differences in their intact parathyroid hormone responses to 25 hydroxyvitamin D?

    Deon Coley-Grant (New Cross Hospital, Wolverhampton, United Kingdom), Mohammed Jawad (New Cross Hospital, Wolverhampton, United Kingdom), Helen L Ashby (New Cross Hospital, Wolverhampton, United Kingdom), SI Handa (Leicester Street Medical Centre, Wolverhampton, United Kingdom), Michael P Cornes (New Cross Hospital, Wolverhampton, United Kingdom), B Kumar (New Cross Hospital, Wolverhampton, United Kingdom), M Hallin (New Cross Hospital, Wolverhampton, United Kingdom), PG Nightingale (Wellcome Trust Clinical Research Facility, University Hospital Birmingham, Birmingham, United Kingdom), Clare Ford (New Cross Hospital, Wolverhampton, United Kingdom), Rousseau Gama (New Cross Hospital, Wolverhampton, United Kingdom)

    Background: Intact parathyroid hormone (iPTH) differences at similar 25(OH)D levels have been reported between US resident whites, blacks and Hispanics, suggesting ethnic differences in vitamin D-PTH homeostasis. It is also possible that there are ethnic differences between South Asians and whites in their iPTH responses to 25(OH)D but this has not been studied. We, therefore, investigated possible South Asian-white differences in their intact parathyroid hormone (iPTH) responses to 25 hydroxyvitamin D [25(OH)D].

    Methods: Demographic and laboratory data on patients resident in Wolverhampton, UK were analysed. Linear regression models measured the association between iPTH and ethnicity before and after adjustment for a range of biochemical and demographic factors.

    Results: Eight hundred and fourteen patients consisting of 334 white subjects (215 women) and 480 South Asian subjects (346 women) were studied. Mean (SD) 25(OH)D concentrations were lower (p<0.0001) in South Asians [26.5 (17.9) nmol/L] than whites [55.2 (17.5) nmol/L]; but iPTH concentrations in South Asians [2.73 (1.85) pmol/L] and whites [2.49 (1.49) pmol/L] were similar (p=0.2172). Intact PTH was associated with 25(OH)D, age, gender, adjusted calcium, magnesium and phosphate but not with ethnicity even after adjustment for the factors shown to correlate with iPTH.

    Conclusion: These data support the concept of ethnic differences between UK resident South Asians and whites in their iPTH responses to total 25(OH)D.

  • P050 Hypoglycaemia in the over 75s: understanding the predisposing factors in Type 2 diabetes

    Adrian H Heald (Salford Royal Hospital, Salford, United Kingdom), Nagaraj Malipatil (Salford Royal Hospital, Salford, United Kingdom), Videlina Cholakova (Leighton Hospital, Crewe, United Kingdom), Marcos Narajos (Oxford University, Oxford, United Kingdom), Adnan Khan (Leighton Hospital, Crewe, United Kingdom), Gemma Donnahey (EMIS Health, Leeds, United Kingdom), Mark Livingston (Walsall Manor Hospital, Walsall, United Kingdom)

    Introduction: Hypoglycaemia has been recognised as a problem in the treatment for type 2 diabetes (T2DM). Here we describe how levels of HbA1C and treatment with a sulphonylurea or insulin relate to risk of significant hypoglycaemia.

    Methods: Incident hypoglycaemia as recorded for the previous 10 years was determined from the GP records for patients with T2DM aged 75 years or more.

    Results: The anonymised GP records of 5974 T2DM patients (2934 men and 3040 women) aged 75 years or more were analysed. Mean-age of the men was 81.0 (95% confidence interval (CI) 80.9-81.2) years and of the women was 82.2 (95% CI 82.0-82.4) years. Hypoglycaemic events of significance were recorded in 4.9% of men and 5.1% of women. The prevalence of hypoglycaemia was higher in those with a higher concurrent HbA1C. HbA1C for those people with a recorded significant hypoglycaemic attack(s) was 57.9 (95% CI 56.4-59.4) mmol/mol compared to those with no history of hypoglycaemic episodes at 51.6 (95%CI 51.3-52.0) mmol/mol (p<0.002). A total of 1376 patients were on sulphonylurea treatment, 846 on insulin and 95 on the combination of sulphonylurea and insulin. Even for those on sulphonylurea and/or insulin treatment, hypoglycaemia prevalence increased with HbA1C: for patients with an HbA1C of <48 mmol/mol, age and gender adjusted hypoglycaemia prevalence was 11.1%, for HbA1C of 48 to 57 mmol/mol, prevalence 9.9%, for HbA1C 58-67 mmol/mol prevalence, 13.2% and for HbA1C 68 mmol/mol or more, prevalence of hypoglycaemia was 16.1%. In multiple logistic regression analysis, treatment with a sulphonylurea or insulin very significantly increased the likelihood of a hypoglycaemic episode: odds ratio (OR) 8.94 (95% CI 6.45-12.42), p<0.001, independent of age, BMI, Townsend Index and gender.

    Conclusion: Prevalence of hypoglycaemia was greater in those individuals with higher HbA1C and in those on sulphonylurea/insulin treatment. Our findings suggest that it is variance in blood glucose rather than overall low blood glucose levels that predispose older people to hypoglycaemia.

  • P051 A case of hypercalcaemia

    Mayur V Patel (Oxford University Hospitals, Oxford, United Kingdom), Brian Shine (Oxford University Hospitals, Oxford, United Kingdom), Adam Darowski (Oxford University Hospitals, Oxford, United Kingdom)

    A 54-year old lady was admitted to hospital following a three-day history of abdominal pain and vomiting. She had several admissions for hypercalcaemia over the previous months, the cause for which remained under investigation. Her past medical history included hypothyroidism, chronic obstructive pulmonary disease, kidney stones, stage 3b chronic kidney disease (CKD), epilepsy and angina. Her medications included amitriptyline, carbamazepine, morphine, levothyroxine, solifenacin, macrogol and senna. The patient was an ex-smoker of less than 30 pack years.

    Clinical examination revealed a soft non-distended abdomen with tenderness in the left flank. Chest auscultation revealed a scattered expiratory wheeze. Biochemistry: adjusted calcium 3.58 mmol/L, Pho 1.81 mmol/L, PTH 0.6 pmol/L, Vit D 26.2 nmol/L, 1,25 OH Vit D 13 (43-144), PTH-rP <1.0 (<1.8), iCa2+ 1.71, HCO3- 31.1 mmol/L (24-30 mmol/L), 9 am cortisol 450 nmol/L, TSH 21.64 mU/L, FT4 8.5 pmol/L. Serum and urine protein electrophoresis were normal. Haematology: Hb 76 g/L, MCV 102.1. CT urinary tract showed a possible right ureteric calculus and several non-obstructing left renal calculi, faecally loaded colon with a foci of mural thickening in the descending colon and sigmoid. CT chest and liver showed no evidence of malignancy. Histology of biopsies taken from the antrum showed a reactive gastropathy and of the descending colon and distal sigmoid colon showed no evidence of malignancy.

    On further questioning the patient revealed that she was consuming two litres of milk and three small pots of yoghurt daily (3 g calcium) and three heaped tablespoons (84 g) of baking soda on alternate days for heartburn. A diagnosis of milk-alkali syndrome was made. The patient was asked to stop taking the baking soda and commenced on a proton pump inhibitor. Her plasma calcium levels improved. This case was difficult to diagnose as the patient’s chronic kidney disease masked the expected alkalosis on previous admissions; however, her plasma bicarbonate was higher than expected for her kidney disease.

  • P052 Point of care testing for drugs of abuse: does it do what it says on the tin?

    Rachel Dale (Colchester Hospital University NHS Foundation Trust, Colchester, United Kingdom), Catherine Street (Colchester Hospital University NHS Foundation Trust, Colchester, United Kingdom), Loretta Ford (City Hospital, Birmingham, United Kingdom), Jonathan Berg (City Hospital, Birmingham, United Kingdom)

    At Colchester Hospital point of care (POC) urine drug testing is performed with consent in pregnancy when patients disclose current or previous illicit drug use. Positive screening results are confirmed by ultra-performance liquid chromatography tandem mass spectrometry (LC-MS/MS). This study assessed the appropriateness of POC screening for illicit drug use using the Alere™ Drug Screen Urine Test Cup.

    The sensitivity and specificity of POC testing for opiates compared to LC-MS/MS was assessed by testing anonymised urine specimens from local antenatal clinics by both methods. The manufacturer’s cut-offs were validated using five drug-free urine matrices spiked with known concentrations of drug at, below and above the published cut-offs. The effect of variation in incubation time was assessed by reading results at 2 min, 3 min, 4 min, 5 min, 10 min, 60 min, 24 h and 48 h.

    The sensitivity and specificity of POC testing for opiates were 71.43% and 99.14% respectively. At the cut-off concentrations and the recommended 5 minute incubation time, 4/5 or 5/5 matrices tested positive for methadone, opiates, benzodiazepines and buprenorphine. Only 2/5 matrices tested positive for cocaine and no positive results were recorded for amphetamine at any concentration tested. At 60 minutes incubation the number of positive results declined for all drugs except opiates.

    These results suggest that POC antenatal drug screening is effective for identifying morphine, 6-monoacetylmorphine, buprenorphine and methadone. It is not suitable for identifying amphetamines present at concentrations up to 150% of the cut-off value. Cocaine, present at the cut-off concentration, was not always detected. There is variation in results with incubation time highlighting the importance of strict adherence to kit operator instructions.

  • P053 Illicit drug use in pregnancy: determining the scale of the problem

    Rachel Dale (Colchester Hospital University NHS Foundation Trust, Colchester, United Kingdom), Catherine Street (Colchester Hospital University NHS Foundation Trust, Colchester, United Kingdom), Loretta Ford (City Hospital, Birmingham, United Kingdom), Jonathan Berg (City Hospital, Birmingham, United Kingdom)

    Exposure to illicit drugs in utero can cause growth restriction, reduced birth weight and neonatal abstinence syndrome. This study sought to estimate the prevalence of illicit drug use in the antenatal population served by midwives at Colchester, Clacton and Harwich hospitals.

    Anonymised urine specimens were collected from all hospital and community antenatal clinics and maternity triage during one week in May 2016 and the gestation and clinic location were recorded. The specimens were screened for cocaine, amphetamine, opiates, methadone, buprenorphine and benzodiazepines using the Alere™ Drug Screen Urine Test Cup and for cannabis by automated immunoassay. Confirmatory testing for a panel of 26 drugs was performed by ultra-performance liquid chromatography tandem mass spectrometry and gas chromatography. Data summarising the disclosure of current drug use by women attending antenatal booking clinics between April 2015 and March 2016 were also examined.

    In total, 8.51% of specimens tested positive for one or more drugs which was considerably higher than the self-disclosure rate of 0.23 %. The most commonly used drugs were opiates (4.36%) followed by cannabis (4.12%) and cocaine (0.41%). The rate of cannabis or cocaine use decreased significantly with gestation (p<0.05) whereas the rate of opiate use increased with gestation.

    This study suggests that disclosure data underestimates the true prevalence of illicit drug use in pregnancy and routine screening may be warranted to identify pregnancies at risk from drug exposure. The decrease in non-opiate drug use with gestation suggests that women may be responsive to improved interventions during early pregnancy to aid cessation of illicit drug use.

  • P054 Putting Deacon’s Challenges in practice: calculation of ethylene glycol half-life during fomepizole treatment

    Timothy J Morris (University Hospital of South Manchester, Manchester, United Kingdom), Graham Horsman (University Hospital of South Manchester, Manchester, United Kingdom)

    Ethylene glycol is an important overdose with the potential to cause significant morbidity and mortality. Toxicity results not from the ethylene glycol itself, but from the metabolites glycolic acid and oxalic acid. Treatment is therefore aimed at preventing the formation of these toxic metabolites by inhibition of alcohol dehydrogenase. This can be achieved using either ethanol or fomepizole, which both work by competitive inhibition of this enzyme. This results in the half-life of ethylene glycol being increased by fomepizole from 3-8 hours to up to 20 hours. A similar prolongation of ethylene glycol half-life occurs with ethanol treatment. We present data showing the altered pharmacokinetic profile of ethylene glycol in a patient treated with fomepizole. An admission sample following ethylene glycol overdose showed ethylene glycol to be 2254 mg/L and ethanol 1256 mg/L. Ethylene glycol was measured daily over the following 4 days during fomepizole treatment, falling to an undetectable value (<25 mg/L) on day 5. The calculated half-life was 17.3 hours over this period. Renal function remained normal throughout, and the ethylene glycol half-life was noticeably constant from day to day. This data is broadly in agreement with that published for the increased half-life of ethylene glycol in patients treated with fomepizole, despite being confounded by presence of ethanol in the blood at the outset. This may have contributed to the ultimately favourable clinical outcome. Calculation of ethylene glycol half-life in this patient on a daily basis was helpful both in predicting the need for an expensive drug, and in overall management and discharge planning.

  • P055 Development of an External Quality Assessment Programme for whole blood immunosuppressants

    Gareth J Davies (WEQAS, Cardiff, United Kingdom), Julie Morris (WEQAS, Cardiff, United Kingdom), Mary A Thomas (WEQAS, Cardiff, United Kingdom)

    Aim: The main aims of the study were to: establish an External Quality Assessment (EQA) programme for whole blood sirolimus, tacrolimus and ciclosporin; validate suitable material for its stability and commutability; and to establish performance criteria for sirolimus, tacrolimus and ciclosporin.

    Methods: Whole blood was used as base material, with each drug gravimetrically added to a target concentration to cover an appropriate concentration range. Intermediate pools were prepared by mixing a high pool with the base to produce a linear panel. The ‘weighed-in’ value of the spiked drug and its known linear dilution was used to calculate the gravimetric value. Each laboratory was sent 3 samples per month and the returned results were compared with the gravimetric values. Short term stability was assessed on two pools stored at room temperature and 4°C over a period of 14 days with long-term stability assessed at -20°C using EQA data over a period of 12 months.

    Results: Sirolimus and tacrolimus appeared stable for 14 days at room temperature and 4°C at concentrations of approximately 2 and 8 µg/L. Ciclosporin appeared stable for 14 days at room temperature and 4°C at concentrations of approximately 100 and 300 µg/L. There was no significant decrease over 12 months at -20°C for the 3 immuosuppressants. For sirolimus recoveries of 77 – 110% were observed for immunoassay and LC-MS/MS methods respectively, however, lower recoveries (50%) were observed below 3 µg/L. For tacrolimus >100% recoveries for all methods were observed. For ciclosporin recoveries of 81-112% and 93-114% were observed for the immunoassay and LC-MS/MS methods respectively.

    Conclusions: We have established an EQA programme for sirolimus, tacrolimus and ciclosporin using liquid stable whole blood material with gravimetric values provided as an improved accuracy target.

  • P056 Method performance compared to gravimetric values on the WEQAS Therapeutic Drug Monitoring EQA Programme

    Gareth J Davies (WEQAS, Cardiff, United Kingdom), Julie Morris (WEQAS, Cardiff, United Kingdom), Mary A Thomas (WEQAS, Cardiff, United Kingdom)

    Introduction: The WEQAS Therapeutic Drug Monitoring (TDM) programme was introduced in 2013 covering 14 of the common drugs and became ISO 17043 accredited in 2015. Gravimetric values are provided for certain analytes as an improved accuracy target. This study assesses the performance of current methods against the gravimetric targets.

    Method: Sterile human serum was used as base material, with each drug gravimetrically added to a target concentration to cover an appropriate concentration range. Intermediate pools were prepared by mixing a high pool with the base to produce a linear panel. The ‘weighed-in’ value of the spiked drug and its known linear dilution was used to calculate the gravimetric value. Each laboratory is sent 3 samples per month and the returned results are compared with the gravimetric values.

    Results: The immunoassay method showed good agreement with gravimetric values for most analytes but there were variations in bias for some instruments. For carbamazepine there was good agreement between the Beckman AU analysers but the Roche Cobas C Module group showed a negative bias of between 10-19%. For phenytoin the Roche Cobas C Module group showed good agreement but the Beckman AU analysers showed a negative bias of between 3-15%. For vancomycin the Roche Cobas C Module group showed a negative bias of between 10-17%.

    Conclusion: The use of gravimetric ‘weighed in’ values as a performance target in EQA data provides a more stable, reliable target than the use of overall or method means, and can also assist in the assessment of traceability of methods.

  • P057 Stability of frozen urinary drugs of abuse samples

    Madeline Graham (Cwm Taf University Health Board, Merthyr Tydfil, United Kingdom; Cardiff Metropolitan University, Cardiff, United Kingdom), Valerie Lewis (Cwm Taf University Health Board, Merthyr Tydfil, United Kingdom), Brian P Tennant (Cwm Taf University Health Board, Merthyr Tydfil, United Kingdom), Alan Dodd (Cwm Taf University Health Board, Merthyr Tydfil, United Kingdom), Sally Thirkettle (Cwm Taf University Health Board, Merthyr Tydfil, United Kingdom), David Cassidy (Cwm Taf University Health Board, Merthyr Tydfil, United Kingdom), John Geen Lewis (Cwm Taf University Health Board, Prince Charles Hospital, Merthyr Tydfil, United Kingdom; University of South Wales, Pontypridd, United Kingdom)

    Introduction: Accurate repeat drugs of abuse (DOA) testing for result confirmation relies on stability during storage. Studies in the literature, predominantly using pooled ‘spiked’ urine or post-mortem samples, demonstrate amphetamine, benzodiazepine and cocaine stability for various temperatures and durations.

    Aims: To examine routine clinical urinary DOA sample stability, under current clinical storages practices, namely three months at -20°C and -80°C.

    Methods: Consecutive positive urine samples for amphetamine (n=17); benzodiazepine (n=20) and cocaine (n=10), measured by immunoassay (KIMS, Roche diagnostics) in our department, were split into two aliquots, stored at -20°C or -80°C for three months. Samples were thawed, vortexed for twenty seconds and centrifuged prior to repeat analysis. Additional integrity checks (creatinine, pH, osmolality or nitrate status) are routinely measured on DOA samples.

    Results: Mean drug concentrations decreased following storage, with comparable 95% confidence intervals (CI). Amphetamine concentrations (1995 ±476 ng/mL) decreased significantly when stored at -20°C (543 ±390 ng/mL) and -80°C (674 ±531 ng/mL) (both p=<0.001). The difference between storage temperatures was not significant (p=0.70). Benzodiazepines (1534 ±365 ng/mL) did not fall significantly when stored at -20°C (1306 ±331 ng/mL) or -80°C (1288 ±331 ng/mL). Similarly, cocaine (3203 ± 692 ng/mL) did not exhibit a significant difference when stored at -20°C (2503 ±708 ng/mL) or -80°C (2565 ±559 ng/mL).

    Conclusions: Current clinical storage practices of three months storage at -20°C or -80°C are fit for purpose for benzodiazepine and/or cocaine positive samples, however are unsuitable for amphetamine positive samples based on this study. Amphetamine degradation post storage has not been observed in any previous literature. These studies typically used GC/MS, which offers sensitivity and specificity superior to that of an immunoassay. We hypothesise that the decrease in measured amphetamine is due to interferences specific to the immunoassay method, potentially affecting antibody cross-reactivity with the epitope. Further studies are required to confirm and characterise this.

  • P058 Investigation into the lower limit of detection of the Cambridge Life Sciences paracetamol assay and evaluation of new serum-based calibrators

    Katie A Hadfield (Barnet Hospital, Royal Free London Hospital NHS Foundation Trust, London, United Kingdom), Penny Skinner (Barnet Hospital, Royal Free London Hospital NHS Foundation Trust, London, United Kingdom), Anne Trewick (Barnet Hospital, Royal Free London Hospital NHS Foundation Trust, London, United Kingdom)

    Instructions for use (IFU) for the Cambridge Life Sciences (CLS) paracetamol kit state that the lower limit of detection (LOD) for this assay is <2.0 mg/L. However, CLS have become aware that this LOD is not achievable on certain third party analysers. An optional serum-based calibrator, stated to improved accuracy and sensitivity of the assay, has been introduced. LOD for the assay calibrated with the serum-based calibrators is stated as <3.0 mg/L.

    We evaluated the LOD (concentration of blank + 2.7SD) for the CLS paracetamol assay calibrated using the routinely available aqueous calibrators and the optional serum-based calibrators on the Abbott Architect C16000 assay platform. A comparison of patient results as measured by each calibration method was also completed.

    LOD of the assay calibrated using aqueous calibrators was 6.3 mg/L. Recovery experiments demonstrated a constant over-recovery of 5.9 mg/L across the range of paracetamol concentrations measured (2.0-20 mg/L). LOD varied between calibrations, but was consistently >2.0 mg/L. LOD of the assay calibrated using serum-based calibrators was 1.8 mg/L. Twenty-three samples with paracetamol concentrations across the measuring range of the assay were analysed by both the aqueous- and serum-calibrated assays. Correlation between results was excellent (r2=1). However there was a constant absolute negative bias of 3 mg/L for results as measured by the serum-calibrated assay compared to the aqueous-calibrated assay.

    LOD as stated in the IFU for the CLS paracetamol assay is not achievable in our laboratory on the Abbott Architect C16000 assay platform. This appears to be the result of a matrix effect and will have resulted in the reporting of measurable paracetamol concentrations in paracetamol-negative patient samples. Calibration of the assay with the new serum-based calibrators mitigates this matrix effect, reducing the LOD to below that stated in the IFU.

  • P059 Evaluate the impact of delayed samples on the stability of lithium in whole blood

    Sumana Gidwani (Northern Health and Care Trust, Antrim, United Kingdom), David Sung (NHSCT, Coleroaine, United Kingdom), Ryan Dapatro (Northern Health and Care Trust, Antrim, United Kingdom), Lorraine Boyd (Northern Health and Care Trust, Antrim, United Kingdom)

    The stability data for most biochemical analytes has already been established. However there are limited studies that have evaluated the stability of lithium in whole blood. The Roche guidance suggests that the samples sent for lithium testing should undergo centrifugation within four hours. However, this can prove a challenge due to delays in transport from the primary health care (PHC), unless these samples were considered for centrifugation at source. This study aims to evaluate the impact of time delays on the stability of lithium in whole blood so as to facilitate defining acceptable pre-centrifugation delays in transport of samples from PHC. Twenty-two consecutive blood samples from PHC were analysed between 2nd-16th December 2016 (Roche Cobas c501 analyser). All samples were collected in SST gel tubes and were facilitated to reach the laboratory within two hours of collection and stored at room temperature. The serum from these samples were then analysed for lithium at 2, 4, 6, 8, 12, 18 and 24 hours (storage time) from the time of collection. Variations in concentrations were expressed as mean bias from baseline, using the analytical change limit (ACL). For each storage time, mean % change of concentration from baseline was determined. The mean bias was considered significant if the value of the bias was above the method ACL of 15%. One hundred and four readings were obtained from the above 22 samples. The mean change from baseline ranged from -3.8 to 4.7% (ACL<15%). There was no significant difference noted in the mean serum concentrations for lithium in these measurements. The mean serum concentration of lithium is not significantly affected if the samples are stored at room temperatures up to 24 hours. Therefore delays in these samples should not impact on rejection for analysis if processed within 24 hours and would obviate the need for considering centrifugation at source.

  • P060 Lead poisoning: an historical perspective

    Elizabeth Fox (St James's University Hospital, Leeds, United Kingdom), Carys Lippiatt (St James's University Hospital, Leeds, United Kingdom)

    The acute and chronic toxic effects of lead in humans have been recognised for many years. Numerous studies in the UK and beyond have demonstrated falling blood lead concentrations in the non-occupationally exposed general population over the latter half of the 20th century. This is reflected in a fall in the quoted reference ranges for blood lead worldwide. The reference interval published in Geigy Scientific Tables in 1962 was stated as 0.72-1.7 μmol/L whereas a large UK based study from 1996 found the 95th percentile in adult males to be 0.49 μmol/L. The decline in average blood lead concentration can be directly related to reduced levels of lead in the environment. Lead pollution in the UK has decreased dramatically over the last 30 years largely due to the phasing out and ultimately the removal of lead from petrol between 1986 and 1999. Crucially, as lead exposure from the environment has been reduced, we have begun to realise that lead has toxic effects that are apparent at concentrations previously thought to be “normal”. As knowledge of the effects of lower blood lead concentrations has increased, the action limits for intervention in childhood lead poisoning have been reduced. In 1982, the UK action limit for initiating the investigation of elevated blood lead in a child was reduced from 1.7-1.2 μmol/L and is currently set at 0.48 μmol/L. In 2012, the Centres for Disease Control in the USA issued a statement to recommend that the maximum acceptable blood lead concentration in a child should be reduced to 0.24 μmol/L. Our aims here are to summarise the correlation between environmental lead pollution and blood lead concentration and to review the effects that this man-made problem has had on the population.

  • P061 Male testosterone and laboratory practice: an urgent need for a standardised, consistent approach

    Paul Downie (University Hospitals of Bristol NHS Foundation Trust, Bristol, United Kingdom), Paul Thomas (Severn Pathology, Bristol, United Kingdom)

    To gain a better understanding of potential variations in laboratory practice with respect to testosterone measurement and reporting in men, a pilot questionnaire was developed and sent to selected United Kingdom laboratories. Questions reflected aspects of requesting, general laboratory practice, reference ranges and reporting. Three hypothetical clinical scenarios were provided with laboratory data and respondents asked what comments they would add.

    Results were received from 11 laboratories. Only 4 laboratories provided specific advice to the user recommending measurement of early morning (09:00) testosterone. There was considerable variation in the reference ranges used by laboratories with only 4 laboratories provided some form of adult age related reference range. Only 2 laboratories had developed an ‘in-house’ reference range. Only 3 laboratories had a cut-off value indicating a result definitely consistent with testosterone deficiency or a range which could indicate borderline or possible hypogonadism. Free testosterone was calculated by 8 laboratories with variation in the equation used to do this. Responses to the clinical scenarios showed laboratories consistently recommending repeating a 09:00 sample following a low result taken in the afternoon as well as reflexing appropriate tests on a result clearly consistent with hypogonadism. However, 9 laboratories given a borderline result would not add any comment at all.

    The absence in several instances of guidance to a range of testosterone concentrations whereby results could still be consistent with testosterone deficiency is a concern. This could result in men with genuine testosterone deficiency not being identified and not undergoing necessary further biochemical and clinical assessment, especially where the laboratory do not specifically comment on such results.

    There is a need for better standardisation of testosterone assays and a more consistent approach between laboratories to lessen differences in patient care. A national audit would help confirm these findings.

  • P062 An unusual case of thyrotoxicosis: when numbers don’t add up

    Helen Leveret (St George's Hospital NHS Foundation Trust, London, United Kingdom), Christina Wei (St George's Hospital NHS Foundation Trust, London, United Kingdom), Lisa Garrison (South West London Pathology, London, United Kingdom)

    A premature male infant, with multiple comorbidities, began having seizures on day 13 of life. He received numerous medications to achieve seizure control and to treat any underlying undiagnosed metabolic disorder including phenobarbitone, phenytoin, midazolam, levetiracetam, pyridoxine, L-carnitine and biotin. Thyroid function tests (TFTs) measured on day 34 were consistent with thyrotoxicosis: free-T4 (fT4) >100 pmol/L, thyroid stimulating hormone (TSH) <0.02 µ/L. Clinically the patient was euthyroid. TFTs were repeated and confirmed original results so a detailed family history was taken. The family history revealed a first cousin once-removed and her daughter had the autosomal dominant inherited condition dysalbuminemic thyrotoxicosis, a condition that biochemically mimics thyrotoxicosis but presents in clinically euthyroid patients. Since the male infant’s mother’s TFTs were normal, and there was no personal history of thyroid problems, this was considered not to be the diagnosis. The endocrinologist advised starting carbimazole and monitoring the TFTs over the coming days. On day 38, the lab advised that the TFTs had returned to normal. Carbimazole was stopped. What had caused this sudden change in TFTs? On review of the baby’s history it was noted that the biotin therapy had been stopped on day 35. At this point, a diagnosis of pseudo-thyrotoxicosis secondary to biotin interference was made. Most immunoassays rely on the interaction between biotin and streptavidin in order to work. If a patient is taking biotin, depending on the assay mechanism, it can lead to either a spuriously high or low result. Biotin is a water soluble molecule with a half-life of 2 hours. When the TFTs were assayed on day 38, biotin had been completely excreted. The manufacturers state in their TFT kit inserts that biotin may cause analytical interference.

  • P063 Maternal circulating leptin and adiponectin in postpartum depression: correlation with foetal physiological well-being

    Pooja Dhiman (JIPMER, Puducherry, India), Raji R (JIPMER, Puducherry, India), Anand Wilson (JIPMER, Puducherry, India), Leena Sharon (JIPMER, Puducherry, India), Nancy Premkumar (JIPMER, Puducherry, India), Shivanand Kattimani (JIPMER, Puducherry, India), Soundravally Rajendiran (JIPMER, Puducherry, India)

    Background: Various biological factors implicated in the pathogenesis of postpartum depression (PPD) have been described, although exact mechanism is still unclear. Adipokines leptin and adiponectin, have been independently associated with depression in general population. We aim to assess the concentration of these adipokines in women with and without postpartum depression, and observe their association with parameters of foetal physiological well-being (APGAR score and birth weight).

    Material & Method: Women were recruited at 6 weeks post-partum, where socio-demographic, obstetrical, and dietary history, along with APGAR score and birth weight of the newborn was noted. These women were screened for the presence of depression using Edinburgh postnatal depression scale (EPDS). Serum was analysed for lipid profile by spectrophotometric method, leptin and adiponectin by commercially available ELISA kits.

    Results: Ninety women with PPD and 90 age-matched controls were included in the final analysis. Leptin and adiponectin concentration did not differ statistically between two groups, although a higher level of leptin was observed in depressed women. A positive correlation of maternal leptin levels with infant birth weight in all subjects (r=0.178, p=0.017), and in control group (r=0.225, p=0.033) was observed. This was not seen in women with PPD (r=0.134, p=0.207), suggesting an altered leptin dynamics with an influence on foetal development, although no significant differences in the birth weight was observed between two groups. APGAR score, which is an expression of infant’s physiological status, was found be positively correlating with maternal adiponectin levels at one minute (r=0.207, p=0.050), in non-depressed women. Also, leptin:adiponectin ratio negatively correlated with APGAR score at five minutes (r=-0.250, p=0.018) in non-depressed women.

    Conclusion: Leptin levels were positively correlated with infant birth weight, whereas APGAR score at one minute was positively correlated with adiponectin levels, and negatively correlated with leptin/adiponectin ratio at five minutes in women without depression. This relationship was lost in women with postpartum depression.

  • P064 Screening for primary aldosteronism: clinical audit of patient preparation prior to measurement of the aldosterone:renin ratio

    Elodie Hanon (The Newcastle upon Tyne Hospitals NHS Foundation Trust, Newcastle upon Tyne, United Kingdom), Christopher Boot (The Newcastle upon Tyne Hospitals NHS Foundation Trust, Newcastle upon Tyne, United Kingdom)

    Primary hyperaldosteronism (PA) is characterised by an inappropriately increased aldosterone concentration compared to the plasma renin concentration. Previously thought to be always associated with hypokalaemia and an uncommon cause of hypertension, PA is present in ~5% of the hypertensive population and frequently occurs with normokalaemia. Screening for PA is performed by calculation of the aldosterone renin ratio (ARR).

    The ARR is affected by a range of pre-analytical factors that may compromise its interpretation. We sought to investigate how well these factors are accounted for in patients undergoing ARR measurement in our Trust. The approach to patient preparation for measurement of aldosterone and renin suggested by the Endocrine Society in 2016 was used as audit criteria; this includes correction of hypokalaemia and withdrawal of anti-hypertensive agents markedly affecting the ARR. Laboratory and clinical data including ARR, treatment and follow-up were gathered over a 3 month period.

    With exclusion of saline suppression tests, 96 cases from the Trust were identified. At the time of analysis 11 patients had a serum potassium <4 mmol/L, and 58 patients were taking up to 5 agents with known effects on ARR. Most importantly, 15 patients were on agents that should have been withdrawn according to Endocrine Society guidance due to known marked effects on the ARR (aldosterone antagonists and potassium-wasting diuretics). ARR was interpreted as “PA unlikely” in 12 of these patients, amongst which 8 were not followed up. Results obtained from patients on such agents are of limited clinical use and constitute a waste of laboratory resources. More seriously, a normal ARR result while on these medications may give false reassurance to the requestor that the patient does not have PA.

    The audit findings will be discussed with the Endocrine and Hypertension clinical teams in order to improve Trust protocols for investigation of primary aldosteronism.

  • P065 Urine catecholamine analysis cannot differentiate neuroblastoma from L-Dopa therapy in children: a case series

    Alison U Kelly (NHS GG&C, Glasgow, United Kingdom), Ellie Dow (NHS Tayside, Dundee, United Kingdom), Allan Dunlop (NHS GG&C, Glasgow, United Kingdom), Rajeev Srivastava (NHS GG&C, Glasgow, United Kingdom)

    Background: Neuroblastoma is the most common solid extracranial malignancy diagnosed in childhood. Clinical presentation is variable, and metastatic disease is common at diagnosis. Analyses of urinary catecholamines and their metabolites are commonly requested as a first-line investigation when clinical suspicion exists. Levodopa (L-Dopa) therapy is utilised as a treatment for a number of disorders in childhood, including Dopa-responsive dystonia. Neuroblastoma may mimic some of the clinical features of this disorder. L-Dopa can interfere with analysis of urinary catecholamines and their metabolites, and complicate interpretation of results.

    Methods: Results of urine catecholamine and metanephrine profiles from 3 children without neuroblastoma but receiving L-Dopa therapy (measured by high performance liquid chromatography with electrochemical detection) were compared with results from 10 patients with true neuroblastoma.

    Results: Urine catecholamine analysis cannot reliably distinguish patients with neuroblastoma from those receiving L-Dopa therapy.

    Conclusions: We recommend that patients should be off L-Dopa therapy if possible when urinary catecholamine analyses are performed. These 3 cases illustrate the importance of providing clinical details and drug history to the laboratory in order to avoid diagnostic confusion and unnecessary further investigations, often causing anxiety to patients.

  • P066 The importance of genetics in routine biochemistry

    Sally Thirkettle (Prince Charles Hospital, Merthyr Tydfil, United Kingdom), Gaynor Harrison (Prince Charles Hospital, Merthyr Tydfil, United Kingdom), Gautam Das (Prince Charles Hospital, Merthyr Tydfil, United Kingdom), John Geen (Prince Charles Hospital, Merthyr Tydfil, United Kingdom)

    Maturity-Onset Diabetes of the Young (MODY) is a clinically heterogeneous group of non-ketotic diabetes mellitus. It is estimated to be the underlying cause of diabetes in 1-2% of the diabetic population, although it is often misdiagnosed as either type 1 or type 2 diabetes. Eighty seven percent of MODY diagnoses are caused by mutations in one of 6 genes: HNF1ɑ, GCK, HNF4ɑ, IPF1, HNF1β and NeuroDI, with the most common mutations being in the transcription factor HNF1ɑ (accounting for 61% of cases), the gene encoding glucokinase (22%) and HNF4ɑ (4%).

    Here we present the case of a family identified with a mutation in the gene encoding the transcription factor HNF1α. The proband presented with young onset diabetes which was misdiagnosed and treated as T2DM for 18 years. Full family history showed the proband had three generations of family members with a T2DM diagnosis, suggestive of a genetic cause. The correct diagnosis for this patient resulted in a change in treatment, greatly improving glycaemic control. Genetic testing has now been carried out in close relatives. This case highlights the importance of being cognisant of potential inherited forms of aberrant glycaemic control and considering genetic testing to obtain a correct diagnosis. A correct diagnosis allows the clinician to tailor the treatment to the individual to improve their quality of life and also for risk factor management. Additionally this case emphasises the necessity to monitor and test family members when appropriate to get an early diagnosis which will reduce the long-term damage caused by a potential misdiagnosis.

  • P067 Have you got a PPI claim?

    Rebecca Powney (Luton and Dunstable University Hospital, Luton, United Kingdom), Danielle Freedman (Luton and Dunstable University Hospital, Luton, United Kingdom), David Housley (Luton and Dunstable University Hospital, Luton, United Kingdom)

    Background: Literature suggests that the incidence of PPI induced severe hypomagnesaemia is low. Anecdotally duty biochemists have noted that a high proportion of patients with severe hypomagnesemia are on a proton pump inhibitor (PPI). Severe hypomagnesaemia can cause significant clinical findings including tetany, convulsions, ventricular arrhythmias, hypocalcaemia and hypokalaemia.

    Aim: To determine the aetiology of severe hypomagnesaemia (serum magnesium <0.4 mmol/L) at the Luton and Dunstable hospital.

    Method: All magnesium results <0.4 mmol/L from 01/03/2013-28/2/2014 were extracted from the hospital information system, Sunquest ICE. The 40 lowest results from adult patients were included in the study. Information from Sunquest ICE including requests, discharge letters and test results were reviewed.

    Results: The location of the request for the 40 patients was: 24 inpatients; 8 Emergency Department (ED); 3 acute assessment units; 3 outpatients and 2 primary care. The aetiology of severe hypomagnesaemia was recorded in the discharge letter for 10/36 patients who were admitted to hospital and the aetiology for 6 of these 10 patients was diarrhoea.

    The possible aetiology was determined for the remaining 30 patients (23 with no aetiology given, 3 with no discharge letter and 4 that weren’t admitted): 16 PPI; 2 severe diarrhoea; 2 diarrhoea ± PPI ± inadequate intake; 1 diarrhoea ± PPI; 2 alcoholic liver disease ±PPI ± inadequate intake; 1 acute pancreatitis; 1 small bowel resection and ileostomy; 2 unknown; 3 had no letter.

    Conclusions: This study shows and confirms severe hypomagnesaemia is not well documented in patients’ discharge letters and PPIs are not a well recognised cause or contributory cause. The duty biochemist has the opportunity to educate and support the investigation of severe hypomagnesaemia, to improve the recognition of PPI-induced severe hypomagnesaemia and contribute to improving patient outcomes.

  • P068 Using a prolactin cut-off of >1000 mU/L to reflex test for macroprolactin misses a high proportion of positive macroproactin cases

    Helen Beeston (Manchester Royal Infirmary, Manchester, United Kingdom), Edward Hinchliffe (Manchester Royal Infirmary, Manchester, United Kingdom), Katharine Hayden (Manchester Royal Infirmary, Manchester, United Kingdom)

    To determine the validity of using a prolactin threshold of >1000 mU/L to reflex for macroprolactin analysis. Following integration of two separate biochemistry laboratories, an improvement audit was undertaken to harmonise the macroprolactin assay and reporting procedures. Macroprolactin analysis using PEG 6000 precipitation in combination with the Roche prolactin generation II assay was observed at both sites. A LIMS gather was performed covering a 12-month period for both laboratories. The data was investigated to determine whether all raised prolactin results should be reflexed for macroprolactin analysis. Reporting methods of the results were also compared. Twenty-one percent (360) of all female prolactin results were found to be elevated. Seventy percent (249) had values of between the upper limit of normal (ULN) and 1000 mU/L. Of these samples, 27% (68) were positive for macroprolactin. In samples with a macroprolactin result >1000 mU/L, 3% (3) were positive for macroprolactin. Identical real data sets were subjected to the two different reporting algorithms used by the two labs. One of these methods which report monomeric prolactin recovery in stratified ranges of >60%, 40-60% and <40% recovery, reported that 11% (28) of female prolactin results between ULN and 1000 mU/L as positive for macroprolactin. The other method, which reports a free prolactin value (mU/L) along with a locally derived reference range, reported 27% of these same samples positive for macroprolactin. Using a prolactin cut-off of >1000 mU/L to reflex test for macroprolactin misses a high proportion of positive macroproactin cases between the ULN and 1000 mU/L. This may lead to unnecessary investigations. Reporting of macroprolactin results as free prolactin (mU/L) with an appropriate reference range is superior; it detects a greater proportion of samples as being positive for macroprolactin and allows the result to be interpreted in a standalone fashion.

  • P069 Failure to thrive with severe hyponatraemia and hyperkalaemia: the role of clinical validation in reaching a diagnosis

    Simon J Whitehead (Royal Wolverhampton NHS Trust, Wolverhampton, United Kingdom), Bhavna Singham (Royal Wolverhampton NHS Trust, Wolverhampton, United Kingdom), Rajeshkumar Jayaraman (Royal Wolverhampton NHS Trust, Wolverhampton, United Kingdom), Clare Ford (New Cross Hospital, Wolverhampton, United Kingdom), Rousseau Gama (New Cross Hospital, Wolverhampton, United Kingdom)

    We report a four-week old boy referred by the midwives for failure to thrive and weight loss (11% in preceding 4 weeks). On presentation he was asymptomatic with normal genitalia but routine tests revealed severe hyponatraemia (Na+ = 119 mmol/L) and hyperkalaemia (K+ = 7.8 mmol/L). Renal failure (normal creatinine), congenital adrenal hyperplasia (normal 17-hydroxyprogesterone) and adrenal failure (normal synacthen test) were excluded. On reviewing results in clinical validation it was recommended that renin and aldosterone be measured. Elevated plasma renin (>2562 mU/L) and serum aldosterone (>8500 pmol/L) levels made the diagnosis of pseudohypoaldosteronism type 1 (PHA1). A renal tract ultrasound scan excluded structural abnormalities, a feature of transient PHA1. He was treated with sodium supplements and a potassium binder. His electrolytes normalised and he gained weight. Following discharge, his electrolytes have remained normal on treatment. We are awaiting the results of genetic analysis to differentiate between the autosomal dominant and autosomal recessive forms of PHA-I because this will have an impact on genetic counselling and long-term management.

  • P070 An unusual case of steroid 17alpha-hydroxylase deficiency

    Francis Lam (Imperial College Healthcare NHS Trust, London, United Kingdom), Gill Rumsby (University College London Hospitals, London, United Kingdom), Florian Wernig (Imperial College Healthcare NHS Trust, London, United Kingdom), Emma Williams (Imperial College Healthcare NHS Trust, London, United Kingdom)

    A 26 year-old Middle Eastern lady of consanguineous parents presented with primary infertility. She was normotensive, had regular menses and was otherwise in good health. At age 14 y she had been investigated for primary amenorrhoea but subsequently underwent spontaneous menarche at 18 y. 17alpha-hydroxylase deficiency (17OHD) had initially been suspected but this was never confirmed, and she was lost to follow up.

    Initial biochemical investigations showed unremarkable U&Es, gonadotrophin and oestradiol concentrations. DHEAS, testosterone and aldosterone were all undetectable with a concomitant plasma renin activity within the normal reference interval. A short synacthen stimulation test showed a peak serum cortisol of 139 nmol/L with no increase in 17-hydroxyprogesterone ruling out 21-hydroxylase deficiency.

    A subsequent urine steroid profile revealed minimal cortisol but raised progesterone and corticosterone metabolite excretion, a pattern consistent with 17OHD. Oddly there were also metabolites of 17-hydroxyprogesterone but not of aldosterone, the latter raising the added possibility of aldosterone synthase deficiency. Genetic testing confirmed homozygosity for the c.160_162del (p.Phe54del) mutation which is known to have residual 17alpha-hydroxylation (37% of normal) but almost complete loss of 17,20-lyase (8% of normal) activities. No mutations were detected for aldosterone synthase.

    Patients with 17OHD typically present with hypertension, hypokalaemia and an absence of secondary sexual development. The patient described in this case did not exhibit these hallmarks. Although biochemical and genetic investigations diagnosed 17OHD there were a number of unexplained observations including normal oestradiol and undetectable aldosterone concentrations, making this an unusual case of 17OHD.

  • P071 Post-saline aldosterone suppression after adrenalectomy for primary aldosteronism

    Ruvini N Ranasinghe (King's College Hospital NHS Foundation Trust, London, United Kingdom), Royce P Vincent (King's College Hospital NHS Foundation Trust, London, United Kingdom), David R Taylor (King's College Hospital NHS Foundation Trust, London, United Kingdom), Benjamin Whitelaw (King's College Hospital NHS Foundation Trust, London, United Kingdom), Simon Aylwin (King's College Hospital NHS Foundation Trust, London, United Kingdom)

    Introduction: Primary aldosteronism (PA) is the most common endocrine cause of hypertension affecting up to 10% of hypertensives. Saline suppression, a confirmatory test for PA, helps avoid patients undergoing invasive adrenal venous sampling due to a false positive aldosterone-to-renin ratio (ARR). The proposed cut-off to exclude PA is post-saline aldosterone suppression to <140 pmol/L. We reviewed our saline suppression data pre- and post-adrenalectomy.

    Method: This retrospective audit reviewed adult patients who underwent saline suppression test between January 2014 and December 2016. Pathology and hospital IT systems were used to obtain information on investigations, multi-disciplinary meeting (MDM) outcomes, clinical management and histology. Data is reported as median (IQR).

    Results: In total there were 77 patients (45M) aged 52 (43-62) years. MDM diagnosed PA in 52 (post-saline aldosterone >140 pmol/L) out of which 25 were managed surgically. Twenty-two, histology confirmed PA had pre- and post-surgery (1-8 weeks after surgery) saline suppression. After surgery, baseline and post-saline aldosterone decreased from 814 (555-1192) to 232 (141-291) pmol/L and 648 (389-951) to 122 (92-159) pmol/L respectively (both, p<0.0001). Post-surgery, only 13 suppressed to <140 pmol/L and 9 suppressed between 140 and 325 pmol/L. The mean aldosterone suppression after saline infusion was 79% (range 62-96%) lower than the pre-surgery supressed values.

    Conclusion: In our cohort the proposed post-saline aldosterone cut-off <140 pmol/L correctly identified all patients with confirmed PA but, only 59% had post-saline aldosterone <140 pmol/L after surgery. However, all had >60% suppression in aldosterone post-surgery. Long-term follow-up studies are needed to establish biochemical confirmation of successful surgical resection in primary aldosteronism.

  • P072 T3 toxicosis occurs in seventeen percent of new cases of hyperthyroidism

    Stephen M GIbbons (Leeds Teaching Hospitals Trust, Leeds, United Kingdom), Mary Moulding (Leeds Teaching Hospitals Trust, Leeds, United Kingdom), Julian Barth (Leeds Teaching Hospitals Trust, Leeds, United Kingdom)

    Background: T3 toxicosis is defined as a subset of hyperthyroid patients who present with suppressed TSH, elevated triiodothyronine (TT3) and normal free thyroxine (fT4). Despite the data available, there is little information on the incidence of this condition in comparison to hyperthyroidism. Indeed, many articles fail to address the true incidence and simply qualitatively describe T3 toxicosis as ‘rare’. Hence, the aim of this retrospective study was to determine the incidence. The practice of thyroid testing in LTHT is based on a frontline TSH and fT4 with added total T3 and TPO abs as considered appropriate by the duty biochemist. All new cases of hyperthyroidism in primary care will then be identifiable with a suppressed TSH and raised TT3.

    Methods: All TFTs from primary care were identified from our LIMS over a 12 month period. Hyperthyroid patients were identified as having suppressed TSH and raised TT3. Only one result per patient was collected. Patients with previously diagnosed hyperthyroidism were excluded.

    Results: Two hundred and ninety-six patients were identified. Of these, 244 patients were hyperthyroid (raised fT4 and TT3) and 51 had T3 toxicosis (raised TT3 and normal fT4) giving an incidence of 17% in the Leeds population. Partitioned data for sex revealed the incidence of T3 toxicosis to be 20% in females and 8% in men.

    Discussion: These data suggest the incidence of T3 toxicosis may be less ‘rare’ than previously described, potentially accounting for one fifth of all female hyperthyroid patients. Moreover, we have provided a quantitative incidence for T3 toxicosis. We propose that T3 toxicosis is not a disease entity but merely an early phase of hyperthyroidism. Through monitoring the 51 patients identified with T3 toxicosis over a 6 month period, rate of progression to hyperthyroidism will also be determined.

  • P073 Positive pregnancy test in a male

    Thomas G Morris (St George's Hospital, London, United Kingdom), Lisa Garrison (South West London Pathology, London, United Kingdom), Gul Bano (St George's Hospital, London, United Kingdom)

    A 50 year old physical trainer of Afro-Caribbean descent presented to the endocrine clinic with a 3 month history of painful gynaecomastia. He was otherwise asymptomatic. He had no history of nipple discharge and was not on any medication. The patient denied use of exogenous steroids. On examination he had a bilateral palpable gynaecomastia. He had normal testes with no palpable mass. His blood tests revealed an elevated testosterone (26.2 nmol/L, (6.68-25.7 nmol/L)), estradiol (298 pmol/L (99-192 pmol/L)) and HCG (250 IU/L (0-2 IU/L)) with suppressed gonadotrophins and a normal SHBG. His CT chest abdomen and pelvis was reported to be normal. In a repeat measurement his HCG had increased to 443 IU/L. Assay interference was considered, and the sample was sent to another laboratory, where again it came back at 461 IU/L. A urine dipstick test was positive for HCG, excluding heterophilic antibody interference. Over the next few months the HCG continued to increase, to >3000 IU/L. No obvious cause of the elevated HCG was found so a diagnosis of paraneoplastic syndrome was made. Various endocrine tests to determine the cause were performed but were all normal. A PET scan demonstrated a hot spot within the anterior mediastinum, and a biopsy confirmed this to be a primary choriocarcinoma. Following this diagnosis the patient is undergoing treatment at Charing Cross. HCG is very similar to LH and high levels can stimulate testosterone production. Testosterone is aromatised to estradiol, high levels of which cause gynaecomastia. The elevated HCG and testosterone also suppress the production of the gonadotrophins via the normal negative feedback loop. All the results in this patient suggested that the hypothalamic-pituitary axis was functioning normally and the mediastinal choriocarcinoma was the source of HCG in this male patient.

  • P074 Graves’ disease in a patient previously diagnosed and treated for Hashimoto’s thyroiditis

    Helen Holt (Luton and Dunstable University Hospital NHS FT, Luton, United Kingdom), Danielle Freedman (Luton and Dunstable University Hospital NHS FT, Luton, United Kingdom)

    Context: Hashimoto’s thyroiditis and Graves’ disease are autoimmune thyroid disorders considered to be opposite ends of the spectrum. Hashimoto’s is caused by thyroid receptor blocking antibodies, whereas Graves’ disease is caused by thyroid receptor stimulatory antibodies. In rare cases, transition from Graves’ to Hashimoto’s has been shown to occur, apparently driven by a change in the ratio of blocking to stimulatory thyroid antibodies.

    Case: We present a 66 year-old lady who had been diagnosed with Hashimoto’s thyroiditis in 2005. Annual monitoring shows the patient was biochemically euthyroid on 50 mcg thyroxine until 2009, however, in November 2010 thyroid function tests revealed a TSH of <0.06 mU/L. Thyroxine was reduced to 25 mcg daily and subsequently stopped in 2014 yet the TSH remained suppressed. Thyroid function tests following withdrawal of thyroxine were fT4 21.3 pmol/L (RR 7.8-17), TSH <0.03 mU/L and anti-TPO antibodies >1000 mU/L. The lady complained of palpitations, feeling excessively hot and increased bowel frequency. On examination she was clinically toxic with a pulse rate of 92, a slight tremor and a diffuse smooth goitre. Ultrasound revealed increased vascularisation of the thyroid in keeping with Graves’ disease. The patient was started on carbimazole 10 mg 3 times daily. At follow up the lady appeared clinically euthyroid but was biochemically hypothyroid (fT4 <3 pmol/L, TSH 47.7mU/L). Carbimazole was therefore withdrawn, however, 6 weeks later, results showed overt thyrotoxicosis (fT4 29.4 pmol/L and TSH < 0.03mU/L). Multiple dose adjustments were required over the following months, however, the patient currently remains stable on 5 mg carbimazole once daily.

    Conclusion: Clinicians should be aware of the potential for switching of thyroid disease states in the long term. It is likely that fluctuating thyroid function due changes in thyroid receptor antibody status is more common than currently accepted but miss-diagnosed as over-treatment.

  • P075 A review of water deprivation tests: an audit of 39 tests over a 5 year period

    Mary Moulding (Leeds Teaching Hospitals, Leeds, United Kingdom), Julie Andrew (Leeds Teaching Hospitals, Leeds, United Kingdom), Heather Cooke (Leeds Teaching Hospitals, Leeds, United Kingdom), Steve Orme (Leeds Teaching Hospitals, Leeds, United Kingdom), Emma Ward (Leeds Teaching Hospitals, Leeds, United Kingdom), Julian Barth (Leeds Teaching Hospitals, Leeds, United Kingdom)

    Water deprivation tests (WDT) are used in the investigation of suspected diabetes insipidus or primary polydipsia. We reviewed the results of WDTs performed in the Day Unit of the Regional Endocrine Unit based at Leeds Teaching Hospitals between 2011 and March 2016 as well as reviewing urine and serum osmolalities performed prior to the WDTs.

    Thirty-nine WDTs were performed in the time period studied. Thirty-seven of these were included in the review; 2 were excluded due to difficulties gaining further information on their subsequent diagnosis.

    WDTs excluded diabetes insipidus in 25 of the 37 patients. Ten of 37 patients had results consistent with cranial diabetes insipidus; 2/37 patients had results suggestive of nephrogenic diabetes insipidus. Nineteen of 37 of the patients had paired urine and serum osmolalities prior to their WDT. Four of 37 patients had only a urine osmolality prior to the test and the remaining 14/37 patients did not have a urine or serum osmolality measured prior to the test.

    Twenty-five patients had WDT results that excluded diabetes insipidus, 17 had a baseline urine osmolality of >500 mOsm/kg; 3 had a baseline urine osmolality of 400-499 mOsm/kg; and 5 had a baseline urine osmolality of <400 mOsm/kg. Twelve patients had WDT results consistent with diabetes insipidus, none of them had a baseline urine osmolality >400 mOsm/kg.

    We therefore suggest that further studies with a larger sample size be conducted to investigate whether a lower cut-off for urine osmolality could be used, rather than the generally accepted diagnostic value of >750 mOsm/kg, to exclude diabetes insipidus. This would consequently result in fewer patients having to proceed to WDT. We also conclude that all patients should have paired urine and serum osmolalities prior to WDT.

  • P076 Determination of renin activity in human plasma: a UPLC-MS/MS method utilising semi-automated solid phase extraction following generation of angiotensin 1

    Sophie C Barnes (Imperial College Healthcare NHS Trust, London, United Kingdom), Amal M Bashir (Imperial College Healthcare NHS Trust, London, United Kingdom), Francis Lam (Imperial College Healthcare NHS Trust, London, United Kingdom), Emma Williams (Imperial College Healthcare NHS Trust, London, United Kingdom)

    Accurate and sensitive measurement of renin is critical to demonstrate autonomous secretion in primary hyperaldosteronism. It is also essential for monitoring mineralocorticoid replacement and investigating other aldosterone production disorders. Traditional radioimmunoassay methods had long incubation times to maximise analytical sensitivity. We report a semi-automated, offline solid phase extraction method with ultra-performance liquid chromatography tandem mass spectrometry (UPLC-MS/MS) detection, offering enhanced sensitivity and increased analytical range for quantifying plasma renin activity. AT1 was generated by renin from endogenous angiotensinogen under controlled conditions. AT1 was extracted from 280 µL of generate using solid phase extraction. Chromatographic separation was performed using a CORTECS UPLC C18 column (2.1x100 mm, 1.6 µm) on a Waters Acquity UPLC system. Chromatographic separation was achieved by reversed-phase gradient elution at 0.5 mL/min, with a total run time of 5.5 minutes. AT1 and its 13C15N internal standard (IS) were detected using a Waters Xevo TQ-S, operating in ESI positive mode. We detected MRM transitions of m/z 649.1>784.2 and 655.6>797.6 for AT1 and IS respectively, co-eluting at 3.4 min. The assay is linear up to 190 nmol/L (r2>0.995, n=10). A lower limit of quantitation was determined at 0.1 nmol/L/h. Good analytical imprecision is obtained for quality control samples at 0.4, 1.3, 3.3, 11.6 nmol/L/h, with CVs of 7.1-9.2% in routine use. No system carryover was observed from a high concentration sample pool (192 nmol/L/h) into subsequent injections. Mean (range) matrix effects were 1.04 (0.99-1.07) for AT1 compared to IS. Comparison of the UPLC-MS/MS assay (y) with a radioimmunoassay (x) method in routine clinical use produced the regression equation y=1.13x-0.04 with excellent agreement below the reference range. Our semi-automated extraction procedure is non-isotopic and offers a significantly increased analytical range allowing 99% of our SAS workload to be reported without dilution compared to 60% by RIA.

  • P077 A child with seizures and hyponatraemia: a case of idiopathic SIADH

    Sakunthala Jayasinghe (Oxford University Hospitals NHS Trust, Oxford, United Kingdom), Rahul Ravindran (Oxford University Hospitals NHS Trust, Oxford, United Kingdom), Gayani Weerasinghe (John Radcliffe Hospital, Oxford, United Kingdom), Zoe Rooney (Oxford University Hospitals NHS Trust, Oxford, United Kingdom), Brian Shine (Oxford University Hospitals NHS Trust, Oxford, United Kingdom), Nishan Guha (Oxford University Hospitals NHS Trust, Oxford, United Kingdom), Tony McShane (Oxford University Hospitals NHS Trust, Oxford, United Kingdom), Taffy Makaya (Oxford University Hospitals NHS Trust, Oxford, United Kingdom)

    A 4½ year-old girl presented with three afebrile seizures occurring within 3 months. She had a history of previous febrile seizures. At the time of first afebrile seizure, she was found to have low plasma sodium and repeat levels ranged from 127 to 134 mmol/L. She had no other significant medical history, and was not on any medications. The plasma osmolality ranged between 270 and 277 mOsmol/kg, urine sodium 63 mmol/L and urine osmolality 452 mOsmol/kg. Other renal, liver and thyroid profile tests and plasma glucose were normal. The response to low dose synacthen test was appropriate and baseline ACTH level was normal. Since the investigations excluded other possible causes of hyponatraemia, SIADH was diagnosed. She had all essential criteria for the diagnosis of SIADH and one supplementary criteria of >0.5% FENa (patient result 0.57%). Blood was sent for co-peptin analysis. Second line investigations were carried out to find out the cause of SIADH. MRI brain revealed a 1.5 cm partly cystic/solid pineal cyst with normal brain parenchyma. However, in the opinion of the neurologist and the oncologist this was unlikely to be the cause for seizures or SIADH. Chest X-ray did not show a space occupying lesion or evidence of infection. Serum tumour markers, AFP and hCG were undetectable and adrenal cortical antibody was negative. Negative CSF cultures excluded cerebral infection. Results for CSF tumour markers are pending. Sodium valproate was started but replaced after a few days with levetiracetam. Since hyponatraemia was present at the time of the first seizure these drugs were not considered to be the cause for SIADH. The most probable diagnosis is idiopathic SIADH.

    This case emphasises the importance of excluding all the possible causes of hyponatraemia and causes of SIADH before concluding it is idiopathic.

  • P078 A case of classical salt-wasting congenital adrenal hyperplasia: lessons to be learned?

    Sasala Wickramasinghe (Barking, Havering and Redbridge University NHS Trust, Romford, United Kingdom), Kausik Banerjee (Barking, Havering and Redbridge University NHS Trust, Romford, United Kingdom), Norman Taylor (Viapath, London, United Kingdom)

    A term baby born of non-consanguineous parents, following an uncomplicated pregnancy, was noted at birth to have clitoromegaly. She was otherwise well. Initial biochemical investigations were unremarkable with normal electrolytes, liver and bone profile. Samples were sent for 17-hydroxyprogesterone (17-OHP; 876 nmol/L), testosterone (>52 nmol/L), cortisol (367 nmol/L) and progesterone (131 nmol/L). Karyotyping (46XX) confirmed female sex. Based on these results, a working diagnosis of congenital adrenal hyperplasia (CAH) secondary to 21-hydroxylase deficiency was made. Urine steroid profiling (USP) confirmed the diagnosis with a characteristic pattern of excess metabolites of 17-OHP, which are virtually absent in the normal neonate, high levels of 3β-hydroxy-5-ene steroids, with relative increase of 16α-hydroxypregnenolone. Cortisol metabolites were not distinguishable, indicating a severe defect. Genetic screening for CYP21 mutations is awaiting confirmation. Over the first 8 days, serum sodium decreased from 140-133 mmol/L, and potassium increased from 4.5-6.2 mmol/L. Hydrocortisone, fludrocortisone and salt replacement was then commenced. Once her electrolytes had stabilised, she was discharged home. Her family was educated on how to administer medication (standard and ‘sick day’ doses) and manage her life-long condition. Deficiency of 21-hydroxylase comprises 90% of all cases of CAH. Of these, ~75% will be salt-wasters. There is generally a ‘lag phase’, approximately 1-week post-partum, before onset of salt-wasting. Females are usually identified early due to virilisation, while affected males are only diagnosed after the clinical presentation of salt-wasting. Therefore, it is important to closely monitor patients for early signs of adrenal crisis. There are also important learning points from a laboratory perspective: USP securely identifies 21-hydroxylase deficiency but cannot differentiate between salt-wasting and non-salt-wasting forms; a ‘normal’ immunoassay cortisol value is likely due to cross reaction, which risks providing false reassurance of a less severe defect; and requesting USP confirmation before treatment is vital for accurate characterisation of the defect.

  • P079 A semi-automated solid phase extraction LC-MS/MS method for combined androstenedione, 17-hydroxyprogesterone and testosterone measurement

    Amy Dunne (Imperial College Healthcare NHS Trust, London, United Kingdom), Francis Lam (Imperial College Healthcare NHS Trust, London, United Kingdom), Emma L Williams (Imperial College Healthcare NHS Trust, London, United Kingdom)

    Measurement of androstenedione, 17-hydroxyprogesterone (17-OHP) and testosterone is important for the investigation and diagnosis of polycystic ovarian syndrome (PCOS) and congenital adrenal hyperplasia (CAH). We present a combined LC-MS/MS method for the simultaneous measurement of androstenedione, 17-OHP and testosterone utilising semi-automated off-line solid phase extraction (SPE).

    SPE was performed using a Tecan Freedom Evo Liquid Handler with Waters Oasis PRiME HLB μ-elution plates. LC-MS/MS used a Waters Acquity-TQD, with separation by high strength silica T3 reverse phase chromatography. Analytes and their co-eluting isotopic internal standards were detected by Multiple Reaction Monitoring (MRM) in electrospray ionisation positive mode. Quantifier transitions (m/z) monitored were 287>97 for androstenedione, 289>97 for testosterone and 331>97 for 17-OHP.

    The method demonstrated good linearity over the analytical range, with all r2 values ≥0.99. Overall process efficiency was 100-108.3%, demonstrating excellent recovery and minimal ion suppression/enhancement. Intra-assay precision was 2.6-8.1% for all analytes across a range of concentrations, and inter-assay precision varied from 4.9-10.8%. Analysis of NEQAS samples revealed a small negative bias and the high specificity of the assay was confirmed by spiking and interference studies. The newly developed assay compared favourably with the stand alone LC-MS/MS methods in use previously, with no requirement to re-derive reference intervals.

    In conclusion, a combined LC-MS/MS method for the simultaneous measurement of androstenedione, 17-OHP and testosterone, using semi-automated off-line SPE, has been validated. This supra-regional assay service (SAS) accredited assay has been in routine use in the LC-MS/MS laboratory at Imperial College Healthcare NHS Trust since April 2016, streamlining our analytical service.

  • P080 Female testosterone results measured by the Abbott Architect immunoassay compared to LC-MS/MS analysis: evaluation of a reflex testing strategy

    Cassandra Porchetta (Imperial College Healthcare NHS Trust, London, United Kingdom), Alexander Ross (Imperial College Healthcare NHS Trust, London, United Kingdom), Emma L Williams (Imperial College Healthcare NHS Trust, London, United Kingdom)

    The second generation Abbott Architect testosterone immunoassay was launched in 2015, with improved performance for the measurement of low testosterone levels found in female patients. This assay has been in routine use at Imperial College Healthcare NHS Trust for the first-line measurement of female samples since January 2015. All samples yielding results above the upper limit of normal (ULN = 2.0 nmol/L) are automatically reflexed for confirmation by LC-MS/MS analysis.

    We present an evaluation of our reflex testing strategy over an 18-month period from 1st April 2015 – 30th September 2016. During this time, 8117 samples were received in the laboratory for testosterone measurement, of which 5019 were from female patients. Of these samples 17 (0.3%) were unsuitable for analysis, 854 (17%) were >2.0 nmol/L and 4148 (83%) yielded results within the reference interval. All samples with elevated testosterone results were auto-reflexed for confirmation by LC-MS/MS analysis.

    Of the 854 samples reflexed, 12 (1.4%) were insufficient for analysis, 198 (23%) were within the reference interval and 644 (75%) remained elevated. For the 198 samples with normal LC-MS/MS results, the original elevated immunoassay results were available for 125 samples. The mean (range) immunoassay result was 2.5 (2.1-11.0) nmol/L and the mean (range) LC-MS/MS result was 1.7 (0.5-2.0) nmol/L. Seven samples showed a significant positive bias on immunoassay with results >4.0 nmol/L and for these samples all LC-MS/MS results were below 1.6 nmol/L. The cause of this positive interference in the immunoassay remains unknown.

    In conclusion, elevated female testosterone results on immunoassay should always be reflexed for confirmation by LC-MS/MS analysis, since a significant proportion of these (23%) yield LC-MS/MS results within the reference interval.

  • P081 Comparison of Roche cortisol generation I and II assays

    Mfon Ewang-Emukowhate (Royal Free Hospital, London, United Kingdom), Nandini Rao (Royal Free Hospital, London, United Kingdom), Efthimia Karra (Royal Free Hospital, London, United Kingdom), Devaki Nair (Royal Free Hospital, London, United Kingdom)

    Background: Cortisol is the major glucocorticoid produced by the adrenal cortex. Its measurement is important in the diagnosis of disease states associated with its deficiency or excess. The Roche cortisol assays (generation I and II) are electrochemiluminescence immunoassays that employ a competitive principle. The second generation assay was developed to improve the specificity of the cortisol assay. It uses a more specific monoclonal antibody directed against cortisol. Previous unpublished work has shown minimal difference in both assays at lower cortisol concentrations. This has implications in determining the cortisol cut-off for the dexamethasone suppression test.

    Aim: To compare the performance of the generation 1 and 2 Roche assays at cortisol concentrations below 50 nmol/L.

    Method: Cortisol measurement was performed simultaneously using both assays on 100 anonymised patient serum samples. These patients were either having a 9am cortisol checked or investigation with dynamic function test. Comparison was made between the results of both assays. Six patients were excluded from statistical analysis; 2 had cortisol levels below the detection limit of 1.5 nmol/L, 2 were on prednisolone, 1 was on fluticasone and dexamethasone, 1 patient’s sample was processed beyond the limit of sample stability.

    Results: There was a correlation between both assays, with a regression equation of; generation II assay = 0.77 (generation I assay) + 3.05 nmol/L, (r2=0.76). Both assays showed good agreement with an average bias of 3.61 nmol/L.

    Conclusion: At cortisol concentrations below 50 nmol/L, this study shows a good agreement between the two Roche assays. We recommend that the cut-off for the dexamethasone suppression test which is <50 nmol/L with the generation I assay should also apply to the generation II assay.

  • P082 Is the LH to FSH ratio abnormal in polycystic ovary syndrome and what’s it got to do with anti-Müllerian hormone?

    Ian Laing (Royal Preston Hospital, Preston, United Kingdom)

    Elevated LH and LH to FSH ratio are well described in polycystic ovary syndrome (PCOS) particularly in earlier studies using competitive LH immunoassays. Circulating LH is heterogeneous due to varied glycosylation and different glycoforms predominate in PCOS. More specific two-site labelled antibody assays exhibit different, poorly defined reactivity with different glycoforms, limiting their usefulness and giving different results in patients with PCOS. Anti-Müllerian hormone (AMH) receptor 2 is expressed in GnRH neurones in the hypothalamus and AMH modulates both neuronal excitability and GnRH secretion in vitro. Intra-cerebroventricular AMH increases LH pulsatility and secretion. This suggests that AMH may be involved in gonadotrophin regulation. This was investigated further by post hoc data analysis of LH and FSH (Roche II) and AMH (DSL Kit) during the early follicular phase in a total of 235 IVF candidates. LH was significantly positively correlated with AMH (rs = 0.24; p=0.001) and negatively with FSH (rs = -0.49; p<0.001). The ratio of LH to FSH was positively correlated to AMH (rs = 0.51, p<0.001). In a subset of patients with PCOS (n=51, Rotterdam criteria) median AMH was higher (6.1 vs 2.0 ng/mL; p<0.001) LH was higher (8.0 vs 6.1 IU/L; p<0.001) FSH was lower (6.0 vs 7.7 IU/L; p<0.001) and the LH to FSH ratio higher (1.30 vs 0.79; p<0.001) These results are consistent with a role for circulating AMH in modulating pituitary LH and FSH secretion and contributing to the increased LH and LH/FSH ratio in PCOS.

  • P083 Curious clinical and biochemical conundrum in a young girl with primary amenorrhoea

    Angela D Burns (Leicester Royal Infirmary, Leicester, United Kingdom), Elaine Maddocks (Leicester Royal Infirmary, Leicester, United Kingdom), Timothy G Davies (Leicester Royal Infirmary, Leicester, United Kingdom), Richard Cole (Leicester Royal Infirmary, Leicester, United Kingdom), David R Taylor (King’s College Hospital, London, United Kingdom), Norman Taylor (King’s College Hospital, London, United Kingdom), Ragani C Bhake (Leicester Royal Infirmary, Leicester, United Kingdom), Faizanur Rahman (Leicester Royal Infirmary, Leicester, United Kingdom)

    A 16 year-old female presented to her GP with primary amenorrhea. Initial findings: LH 28 IU/L, FSH 16 IU/L, 17-beta oestradiol 97 pmol/L. Further analyses recommended by Clinical Biochemistry staff: progesterone 54.8 nmol/L, testosterone <0.3 nmol/L, 9 am cortisol 125 nmol/L. The elevated gonadotrophins suggested that oestradiol was inadequate, while elevated progesterone with a low oestradiol level was unexpected. These results suggested a diagnosis of 17α-hydroxylase deficiency. In this disorder, cortisol is expected to be low/absent, but may be explained by the high levels of precursor steroids cross-reacting with our cortisol immunoassay. On referral to Endocrinology, the patient was tall in comparison to her non-consanguineous parents; she was overweight by WHO criteria for obesity, normotensive and in a pre-pubertal female state. There was no family history of endocrinopathy. Investigations showed an elevated ACTH and inadequate cortisol response in a short synacthen test. The karyotype was 46XX. Samples for a urine steroid profile and serum 17-hydroxyprogesterone (17-OHP) were collected. The urine steroid profile indicated a virtually complete 17α-hydroxylase deficiency. However, the serum 17-OHP (4.1 nmol/L) and 11-deoxycortisol (S) (61 nmol/L (RR 7-13)) by liquid chromatography tandem mass spectrometry (LC-MS/MS) at another site were not consistent with this diagnosis. This led us to consider isobaric interference from corticosterone. In 17α-hydroxylase deficient patients corticosterone is raised which might explain the elevated S result. Using an alternative steroid LC-MS/MS assay, designed to resolve the major isobars, the values were (nmol/L): S, <0.36, 17-OHP 3.4, corticosterone 819. This was consistent with the diagnosis and indicated that S had been read as corticosterone in the previous assay. This case is of interest as a rare cause of primary amenorrhea and highlights the importance of understanding biochemical methods and their limitations when interpreting results and the laboratory’s role in instigating further investigations. Additionally it reaffirms the importance of collaboration between laboratories and endocrinology which was key in solving this case.

  • P084 A case of insulin autoimmune syndrome presenting with hypoglycaemia associated with a raised insulin to C-peptide ratio

    Catriona A Clarke (Western General Hospital, Edinburgh, United Kingdom), Anna Dover (Edinburgh Centre for Endocrinology and Diabetes, Edinburgh, United Kingdom), David Church (The University of Cambridge Metabolic Research Laboratories, Cambridge, United Kingdom), Robert Semple (The University of Cambridge Metabolic Research Laboratories, Cambridge, United Kingdom), Rebecca Pattenden (Western General Hospital, Edinburgh, United Kingdom)

    Background: Insulin autoimmune syndrome is a rare condition characterised by recurrent hypoglcyaemic episodes often following periods of hyperglycaemia, associated with inappropriately high insulin and insulin autoantibodies. It is the third leading cause of spontaneous hypoglycaemia in Japan but is increasingly being recognised worldwide in non-Asian populations. Failure to consider this rare diagnosis could result in inappropriate investigation and treatment. Insulin and C-peptide are released at equimolar concentrations but due to the longer half-life of C-peptide, insulin is present at lower concentrations. When insulin is found at higher concentrations than C-peptide this classically indicates the presence of exogenous insulin but may also indicate the presence of insulin autoantibodies.

    Case description: A 37 year-old non-diabetic woman presented with symptoms of recurrent hypoglycaemia. Using the Abbott Architect immunoassays, a high insulin level with relatively low C-peptide was found (insulin 39181 pmol/L; C-peptide 1046 pmol/L; glucose 2.8 mmol/L). Measurement with the Mercodia immunoassays at a referral laboratory indicated a more typical ratio of results (insulin 307 pmol/L; C-peptide 515 pmol/L). Autoimmune hypoglycaemia syndrome was considered in the differential diagnosis. The presence of insulin antibodies was confirmed and following treatment, antibody titre and immunoreactive insulin levels reduced and symptoms improved.

    Conclusion: This case highlights that in rare circumstances high levels of insulin relative to C-peptide could be due to the presence of autoantibodies bound to insulin. Further investigation is required to confirm the presence of autoantibodies and establish if they are pathological. Insulin autoimmune syndrome as a rare cause of hypoglycaemia should be considered in the differential diagnosis of spontaneous hypoglycaemia in non-diabetic patients.

  • P085 Radioimmunoassay and immunometric assay discordance as an indicator of thyroglobulin antibody interference

    Lorna E Rashid (Western General Hospital, Edinburgh, United Kingdom), Catherine Shearing (Western General Hospital, Edinburgh, United Kingdom), Karen Smith (University Hospitals Birmingham NHS Foundation Trust, Birmingham, United Kingdom), Rebecca Pattenden (Western General Hospital, Edinburgh, United Kingdom)

    Introduction: Endogenous thyrogloulin antibodies are reported in up to 30% of differentiated thyroid cancer (DTC) patients and can lead to falsely positive thyroglobulin results by radioimmunoassay (RIA) and falsely negative results in immunometric assays (IMA). Assay discordance between RIA and IMA methods has been suggested as a method to assess thyroglobulin antibody interference, although there is no definition as to what level of discordance is indicative of interference.

    Methods: The results of 3019 patient samples that had been analysed for thyroglobulin by both IMA (Beckman Access; analysed at Western General Hospital, Edinburgh) and RIA (in-house method analysed at Queen Elizabeth Hospital, Birmingham) methods were identified from the laboratory computer system between 1st January 2012 and 31st May 2016.

    Results: There were 230 samples with thyroglobulin (Tg) results within the reporting limits of both RIA and IMA assays, with good agreement between the two methods (y=1.065x–0.080; r2=0.931). The ratio of IMA:RIA Tg levels was calculated for these samples, mean = 0.98, 2.5th and 97.5th percentiles were 0.75 and 1.99 respectively. One hundred and forty samples were identified with an undetectable IMA Tg levels (<0.5 ug/L) paired with a reportable RIA Tg level (>5 µg/L), the mean RIA Tg level = 18 µg/L, 2.5th and 97.5th percentiles were 7 and 66 µg/L respectively.

    Conclusions: There is limited data in the literature on the degree of disconcordance between RIA and IMA methods which is indicative of assay antibody interference. We have reviewed a large number of paired patient results to define the level of agreement between the IMA and RIA methods and determined the ratio in thyroglobulin levels between the two assays which may aid identification of antibody interference.

  • P086 How should prednisolone be administered in glucocorticoid replacement therapy?

    Sirazum M Choudhury (Imperial College Healthcare NHS Trust, London, United Kingdom), Emma L Williams (Imperial College Healthcare NHS Trust, London, United Kingdom), Tricia Tan (Imperial College NHS Trust, London, United Kingdom), Karim Meeran (Imperial College Healthcare NHS Trust, London, United Kingdom)

    Aim: Patients with adrenal insufficiency are unable to produce cortisol and require oral glucocorticoids to mimic the physiological diurnal cortisol rhythm. Cortisol typically peaks within half an hour of waking. Medication inserts advise that prednisolone is taken with food. We investigated the effect of concurrent food ingestion on the pharmacokinetic profile of prednisolone, with a view to achieving peak concentrations as early as possible.

    Methods: Three participants (A, B and C) took prednisolone in controlled conditions: 1) baseline fasting state, 2) within 30 minutes before completing breakfast, and 3) within 10 minutes after completing breakfast. Prednisolone day profiles were performed using a UPLC-MS/MS method. Samples were taken at defined time points and plotted against time.

    Results: Taking prednisolone in association with breakfast caused small changes in its metabolism. The time to peak concentration (TMax) was prolonged in participants B and C, when prednisolone was taken before breakfast, and shortened in participant A. When prednisolone was taken after breakfast, participants A and B had a quicker TMax. It was however delayed in participant C. This suggests that the effect of concurrent food intake on prednisolone metabolism cannot be reliably predicted. The largest observed reduction in TMax was 24 minutes and the largest increase was 102 minutes when compared to the individual participant’s baseline fasting profile.

    Conclusion: In order to mimic the early morning physiological cortisol peak as closely as possible, patients should be advised to take their prednisolone tablets immediately, as soon as they wake. Taking prednisolone with breakfast will only add an unnecessary delay to achieving TMax and is unlikely to confer any benefit to absorption. This is contrary to the instructions contained in the medication information leaflets, which may be more appropriate for anti-inflammatory doses.

  • P087 Audit of female testosterone results measured by the Abbott Architect 2GENTESTO immunoassay method: is confirmation by tandem mass spectrometry necessary?

    Katie A Hadfield (Barnet Hospital, Royal Free London Hospital NHS Foundation Trust, London, United Kingdom), Penny Skinner (Barnet Hospital, Royal Free London Hospital NHS Foundation Trust, London, United Kingdom), Anne Trewick (Barnet Hospital, Royal Free London Hospital NHS Foundation Trust, London, United Kingdom)

    Traditionally immunoassays for testosterone measurement have been subject to interference from other endogenous androgens in female patients. As a result these methods are often considered unsuitable for the accurate measurement of female testosterone results with many laboratories confirming raised immunoassay testosterone concentrations in female patients by tandem mass spectrometry (LC-MS/MS) methods. Since their introduction immunoassay testosterone methods have undergone much development. The Abbott Diagnostics 2GENTESTO method was launched in December 2015 to improve the accuracy of measurement at low testosterone concentrations.

    Procedure in Clinical Biochemistry at Barnet Hospital, Royal Free London Hospital NHS Foundation Trust, is to confirm female testosterone results >2.0 nmol/L by LC-MS/MS. An audit was undertaken comparing Abbott Architect (AA) and LC-MS/MS testosterone results for samples referred 01/09/16 – 31/10/16 and 9 samples with AA testosterone results within the reference range. Clinical details were reviewed for all patients with discordant results by the two methods.

    LC-MS/MS and immunoassay results were compared for 52 samples. Results were concordant ie. either consistently above or within the reference range by both methods for 41/52 (79%) samples. Eight of 52 samples had results above the immunoassay reference range, but within the LC-MS/MS reference range. Clinical details were available for 7/8 discordant patients. All 7 had clinical features of hyperandrogenism, oligo-/amenorrhoea or pelvic ultrasound scan features consistent with polycystic ovaries (PCO). This was in contrast to the 6 patients with results within both the immunoassay and LC-MS/MS reference ranges all of whom were being investigated for hair loss with no other hyperandrogenic or PCO features.

    In conclusion the Abbott Diagnostics 2GENTESTO assay performs comparably to LC-MS/MS for the majority of female samples. AA testosterone concentrations were consistent with clinical details, where available, for patients with discordant results. Routine confirmation of female testosterone concentrations by LC-MS/MS is not required for results generated by the 2GENTESTO assay.

  • P088 The first in vitro clinical thyroid receptor antibody assay: historical contributions to the clinical application of autoimmunity in Grave’s disease

    Dilini Peiris (Basingstoke and North Hampshire Hospital, Basingstoke, United Kingdom)

    The development of the first thyroid receptor antibody (TRAb) assay is a fine example of how the application of biochemical principles can have real impact with regards to innovations in healthcare science and clinical application. TRAb assays are immunoassays that are used to aid clinicians in the diagnosis of Grave’s disease (GD), owing to the involvement of autoantibodies in its pathogenesis. The autoantibodies, thyroid stimulating immunoglobulins (TSI), act to stimulate thyroid cells to produce thyroid hormones uncontrollably, by binding to their TSH receptors. This causes the hyperthyroidism seen in GD.

    This concept was first described by Adams and Purves in 1956. They described a ‘thyroid stimulating factor’ which was later defined as an autoantibody by Kriss et al. in 1964. However, it was Rees Smith and Hall who collaborated on the first in vitro clinical TRAb assay in 1974, by utilising the autoimmunity of GD analytically. It involved a radioreceptor assay, based on the ability of autoantibodies to inhibit 125I-labelled TSH binding to thyroidal TSH receptors.

    The development of the first TRAb assay has had several implications. It enabled further elucidation of the pathogenesis of GD and has become a clinical tool to inform the management and treatment of GD. From an analytical perspective, it also laid the foundations for the progression of TRAb assays, with regards to increased sensitivity and specificity.

    The TRAb assay has evolved from the radioreceptor assay, to liquid and solid phase assays, through to more specific assays based on human thyroid stimulating monoclonal antibodies. This innovation in healthcare would not have been realised without the important contributions of Rees Smith and Hall. By reviewing their early work, it is hoped that the role of biochemistry is highlighted and valued in terms of its integral part in assay development and clinical application, particularly with regards to GD.

  • P089 Historical perspectives of testosterone: from antiquity to the modern day

    Paul Downie (University Hospitals of Bristol NHS Foundation Trust, Bristol, United Kingdom), Frances Palmer (University Hospitals of Bristol NHS Foundation Trust, Bristol, United Kingdom), Corey Pritchard (University Hospitals of Bristol NHS Foundation Trust, Bristol, United Kingdom)

    Testosterone plays a critical role in various physiological functions in men and women, its name derived from a combination of the words testicle, sterol and the suffix ketone. We review the history of this fascinating steroid and chart its progress from ancient times through to its place in modern medicine. John Hunter (1728-1793) the founder of experimental pathology reported that castrated roosters recovered some of their ‘maleness’ after testicular implantation. Arnold Berthold furthered this work and in 1849 he discovered that testicular action in male chickens was correlated to circulating fractions (androgens). In 1889 the very famous clinician Charles Brown Séquard extracted a “rejuvenating elixir” from dog and guinea pig testes. Following self-injection with his elixir he subsequently reported improved vigour and well-being. His work attracted considerable notoriety so much so that livestock dealers in San Francisco could not keep guinea pigs in stock! It is probably accepted now that the reported response was a placebo effect. In 1927 Fred Koch obtained a large number of bull testicles and extracted isolates of pure testosterone. When injected into de-masculinized roosters fascinatingly they became re-masculinized! Butenandt and Hanisch created the first synthetic version of testosterone from cholesterol in August 1935 and at a similar time a group in The Netherlands isolated a few pure milligrams of testosterone after determining the hormone’s molecular structure. In the 1950s testosterone was developed for use as a medical treatment however this could only be administered by injection or a subcutaneous pellet and proved ineffective if taken orally. It was not until the year 2000 that transdermal gels were developed which became a standard and popular delivery method. Despite it being a widely researched and understood hormone, the investigation and treatment of male testosterone deficiency continues to be misunderstood by many within the medical community.

  • P090 Reference range of serum anti-Müllerian hormone in Omani healthy women in their reproductive age

    Elham Alrisi (Oman Medical Speciality Board, Muscat, Oman), Waad A. S. Mula-Abed (Oman Medical Specialty Board, Muscat, Oman)

    Introduction: Anti-Müllerian hormone (AMH) is a glycoprotein belonging to transforming growth factor beta family that are involved in tissue growth and differentiation. It has a lot of clinical application in the field of women’s health, it assess the ovarian reserve, prediction of response to controlled ovarian stimulation, monitoring granules cell tumours, surrogate biomarker for antral follicle count in diagnosis of polycystic ovarian syndrome (PCOS), and it helps in prediction of time of menopause. The objective of this study is to establish the reference range of AMH in healthy Omani women during their reproductive life as an overall reference range and as age-dependent reference range in different age groups (20-45 years).

    Methods: This is a pilot study conducted in healthy fertile Omani women aged between 20-45 years. The subjects divided into five age groups, total sample size will be 350-400 and each interval had a number that is calculated by using statistical analysis. Subjects were healthy volunteers. The inclusion criteria included females with a regular period, no history of any chronic illness, not using any type of hormonal contraceptive, no history of PCOS and any gynaecological operation, body mass index (BMI) <30. Blood collection will be done on day 21 of the period. Progesterone, LH, and FSH hormones along with AMH measured to insure the fertility status of the women.

    Results: This is the primary data of 165 women, for age-group (20-25 years) AMH median was 22.86 and the range was (2.38-53.72), age group (26-30) median was 18.91 (2.31-61.95), age group (31-35) median was 13.48 (1.37-55.83), age group (36-40) median was 11.39 (1.56-41.54), age group (41-45) median was 4.68 (1.15-47.77). Distribution of AMH is not the same across the age group (p value was 0.0000).

    Conclusion: Age-dependent reference range of AMH was determined. The level of AMH is decrease with age and this support the physiology of this hormone and previous studies done in this field.

  • P185 Evaluation of a novel third generation parathyroid hormone assay

    Rachel Cullen (Mater Misericordiae University Hospital, Dublin, Ireland), Meg M O'Riordan (Mater Misericordiae University Hospital, Dublin, Ireland), David A Foley (Mater Misericordiae University Hospital, Dublin, Ireland), Maria C Fitzgibbon (Mater Misericordiae University Hospital, Dublin, Ireland)

    Parathyroid hormone (PTH) levels are used to guide treatment decisions in patients with chronic kidney disease (CKD). Second generation intact PTH (iPTH) immunoassays measure the biologically active PTH (1-84). However, they have been shown to cross react with PTH fragments, including PTH (7-84) which is increased in CKD. Our aim was to evaluate the performance of a novel third generation “bio-intact” PTH assay which claims to exclusively detect PTH (1-84).

    The analytical performance of the PTH (1-84) assay by Roche Diagnostics was assessed on a Cobas e411 platform using the ACB assay verification protocol. Specificity was evaluated with a UKNEQAS sample spiked with PTH (7-84). Correlation of both PTH assays with creatinine, eGFR, corrected calcium and phosphate was assessed in a cohort of CKD patients (n=77). The % of PTH (1-84)/iPTH was determined across a range of (eGFR).

    The PTH (1-84) assay correlated with the Roche Diagnostics iPTH assay but was on average 28% lower (r=0.997, Passing-Bablok: PTH (1-84) = (0.72) iPTH-0.4 n= 428). No cross reactivity with the PTH (7-84) fragment was detected. Similar correlations for the two PTH assays (Intact PTH and PTH (1-84) respectively) were found when compared to creatinine (r=0.7710, r=0.7496), eGFR (r=-0.4374, r=-0.4403), corrected calcium (r=-0.1440, r=-0.1206) and phosphate levels (r=0.4813, r=0.4843). The % of PTH (1-84)/iPTH decreased with decreasing eGFR (p=0.0037).

    We have analytically verified the performance of this novel assay which offers improved specificity for PTH (1-84). This assay may provide superior diagnostic interpretation of PTH in the management of CKD patients.

  • P091 Roche Generation 3 turbidimetric immunoassay for HbA1c: an evaluation with a focus on haemoglobin variants

    Kathryn McDade (NHS Lanarkshire, Wishaw, United Kingdom), Ravinder Sodi (University Hospitals of Morecambe Bay NHS Foundation Trust, Lancaster, United Kingdom), Linda Jones (Glasgow Caledonian University, Glasgow, United Kingdom)

    Introduction: Glycated haemoglobin (HbA1c) is used in monitoring and diagnosis of diabetes mellitus. The commonly used cation-exchange chromatographic methods (e.g Tosoh G8) are usually affected by haemoglobin (Hb) variants and are expensive. The Roche generation 3 turbidimetric (Roche A1C-3) immunoassay is an alternative that offers fast sample throughput, is not affected by Hb variants and is automatable. The main aim of this study was to evaluate the Roche A1C-3 method.

    Methods: Samples (60 males; 60 females; age range 5-98 years) were compared by the Roche A1C-3 and Tosoh G8 methods. Accuracy was determined using 27 (range: 33-98 mmol/mol Hb) external quality assurance (EQA) samples. Precision on the Roche A1C-3 method was assessed by determining the within- and between-day variation. Interference in the Roche A1C-3 method by Hb variants, lipaemia (triglycerides) and icterus (bilirubin) was evaluated. The sample types with EDTA and fluoride-oxalate additives were compared.

    Results: The Roche A1C-3 and Tosoh G8 methods were highly correlated (r=0.986). There was a statistically significant difference between the Roche A1C-3 and Tosoh G8 methods (mean [SD] 49.0 [17.7] vs. 51.5 [19.2] mmol/mol Hb; p=<0.0001); however, the difference of 2.5 mmol/mol Hb was not deemed clinically significant. In general all EQA samples were comparable to the specified target. The overall within- and between-run precision for the Roche A1C-3 method was 2.92% and 2.98%, respectively. There was no interference by any variants studied or triglyceride and bilirubin concentrations up to 11.8 mmol/L and 420 µmol/L, respectively. There was no significant bias between fluoride-oxalate and EDTA samples.

    Conclusions: This study has shown that the Roche A1C-3 method is comparable to the Tosoh G8 method, has good precision and accuracy and is not affected by Hb variants, lipaemia or icterus. Fluoride-oxalate samples used for glucose measurement can also be used to analyse HbA1c using the Roche A1C-3 method. This study shows that this immunoassay for HbA1c is ready for prime time.

  • P092 Post-partum screening for dysglycaemia in women with gestational diabetes mellitus is poor, despite the high risk of subsequent diabetes

    Rebecca J Ward (Department of Clinical Biochemistry, Stoke-on-Trent, United Kingdom), Fahmy Hanna (Department of Diabetes and Endocrinology, Stoke-on-Trent, United Kingdom), Ann Shelley-Hitchen (Department of Diabetes and Endocrinology, Stoke-on-Trent, United Kingdom), Ellen Hodgson (Department of Diabetes and Endocrinology, Stoke-on-Trent, United Kingdom), Adrian Heald (Department of Medicine, Crewe, United Kingdom), Anthony A Fryer (Department of Clinical Biochemistry, Stoke-on-Trent, United Kingdom), Christopher J Duff (Department of Clinical Biochemistry, Stoke-on-Trent, United Kingdom)

    Background: Women diagnosed with gestational diabetes (GDM) have elevated risk of developing type 2 diabetes (T2DM). NICE recommend women who develop GDM are screened 6 weeks post-partum and annually thereafter.

    Aims: To (i) evaluate conformity to guidance on screening in women with GDM by 6-week post-partum fasting plasma glucose (FPG) and annual FPG and (ii) determine time between delivery and development of T2DM.

    Methods: Records at a teaching hospital were used to identify women (n=252) diagnosed with GDM by antenatal oral glucose tolerance test between July 1999 and January 2007. Data from laboratory records were used to collect further assessments of glycaemic status during follow-up (median: 12.4 years [range 9.5-17.1]).

    Results: Of the 252 women with GDM, 150 (59.5%) had the 6-week post-partum FPG (±2 weeks) while 1 (0.4%) had a FPG before 4 weeks post-partum and 101 (40.1%) did not have a FPG prior to 8 weeks. Twenty-nine of the 252 women (11.5%) had a follow-up FPG at 52±8 weeks. Of the 150 tested at 6 weeks post-partum, 17 (11.3%) had no further tests during follow-up, 43 (28.7%) were tested too soon (<44 weeks), 73 (48.7%) were tested too late (>60 weeks) and 17 (11.3%) were tested on time. Of the 150 women, 7 had post-partum results suggesting pre-existing T2DM and 45 (31.5%) developed T2DM during follow-up. Of the 101 women who did not have the 6 week post-partum test, 5 (5.0%) had no further tests during follow-up, 35 (34.7%) were tested too soon (<44 weeks), 49 (48.5%) were tested too late (>60 weeks) and 12 (11.9%) were tested on time. Conclusion: Our data suggest an alternative approach is needed for monitoring women with a history of GDM. This will need to be appropriate for a generally healthy patient group in which traditional screening mechanisms may not be adequate.

  • P093 Utility of measuring both GAD and ICA in diagnosis of latent autoimmune diabetes in adults

    Lauren E Hughes (New Cross Hospital, Wolverhampton, United Kingdom), Clare Ford (New Cross Hospital, Wolverhampton, United Kingdom), Anjali Ekbote (New Cross Hospital, Wolverhampton, United Kingdom), Rousseau Gama (New Cross Hospital, Wolverhampton, United Kingdom)

    Introduction: Latent Autoimmune Diabetes in Adults (LADA) is characterised by onset of diabetes in adulthood, with the presence of islet specific autoantibodies at diagnosis of diabetes. Patients demonstrate slow progression to beta-cell failure, thus can present with clinical characteristics of both Type 1 and Type 2 diabetes mellitus. This makes it difficult to diagnose and treat appropriately. Estimates for prevalence of LADA range from 10% to 60% of patients initially diagnosed as Type 2. Several auto-antibodies have been identified that contribute to autoimmune diabetes. Royal Wolverhampton NHS Trust (RWT) currently offers islet cell antibodies (ICA) and glutamic acid decarboxylase (GAD) autoantibodies to aid with diagnosis of LADA. ICA analysis by immunofluorescence is time consuming. A service evaluation was performed to determine the utility of measuring both ICA and GAD in diagnosis of LADA.

    Methods: All GAD and ICA requests for patients ≥18 years from Jan 2015 to May 2016 were identified using the Laboratory Information System. Concordance between GAD and ICA results was assessed.

    Results: One hundred and seventy-eight requests for GAD, ICA, or both, were identified; 109 requested both GAD and ICA. Of these, 85 were double negative, 8 were double positive, 25 were GAD +ve, ICA –ve, and 1 was GAD –ve, ICA +ve. Discussion with a diabetiologist confirmed the GAD –ve, ICA +ve patient was an 18 year old Type 1 diabetic for whom islet autoantibody testing was not clinically indicated.

    Conclusion: The service evaluation suggests that ICA in combination with GAD does not have a significant impact of improved diagnosis of LADA over measurement of GAD alone and therefore ICA is no longer offered for the diagnosis of LADA at RWT. Investigation of the cost effectiveness of offering new islet specific autoantibodies such as zinc transporter 8 and islet antigen 2 in place of ICA is currently ongoing.

  • P094 A case of autoimmune haemolytic anaemia leading the diagnosis of ‘no diabetes’

    Leanne Evans (Liverpool Clinical Laboratories, Liverpool, United Kingdom), Andrew Davison (Liverpool Clinical Laboratories, Liverpool, United Kingdom)

    Background: Measurement of glycated haemoglobin (HbA1c) is widely acknowledged as a useful clinical tool for the diagnosis and monitoring of diabetes mellitus. Despite this there are a number of limitations to its use, including situations where the life of the red cell is altered. This can lead to falsely high or low HbA1c results. For example from blood loss, haemoglobinopathies, and anaemia (including haemolytic and iron deficiency).

    Case Description: A 79 year old male was diagnosed with impaired fasting glycaemia in April 2014, glucose 6.8 mmol/L and HbA1c 45 mmol/mol (Menarini 8160). At this time he was commenced on metformin to aid with glycaemic control. In February 2016 repeat HbA1c was 43 mmol/mol (Tosoh G8). By August 2016 it had dropped to <20 mmol/moL. Metformin was stopped and the patient reported by clinician as having “no diabetes”. In November 2016 following clinical investigation for tiredness and shortness of breath, the patient was diagnosed with autoimmune haemolytic anaemia (Hb = 91 g/L, RBC = 2.18x1012/L, LDH = 683 U/L, bilirubin = 54 µmol/L, haptoglobin = <0.1 g/L). Fasting glucose at this time was 6.3 mmol/L and HbA1c <20 mmol/mol (Tosoh G8). When the team realised the HbA1c result was unreliable due to increased red cell turn over the patient was re-assessed by diabetes specialist team and re-prescribed metformin.

    Conclusion: This case demonstrates the importance of questioning results and interpreting them in the context of disease. Here HbA1c was found to be falsely low due to increased red cell turnover in the context of haemolytic anaemia and was therefore an unreliable index of glycaemic control. Clinicians should consider alternative approaches to monitoring glycaemic control including more frequent measurement of plasma glucose and glycated albumin in such patients.

  • P095 A short verification study of gestational glucose tolerance test containers and collection procedure

    Elinor V Hanna (Northern Health and Social Care Trust, Antrim, United Kingdom), David A Wright (Northern Health and Social Care Trust, Antrim, United Kingdom), Ryan Da Prato (Northern Health and Social Care Trust, Antrim, United Kingdom), Sumana Gidwani (Northern Health and Care Trust, Antrim, United Kingdom), Adele Kennedy (Northern Health and Social Care Trust, Antrim, United Kingdom)

    The aim of this study was to compare the glucose results between standard NaF/EDTA tube both at room temperature and on ice to the Greiner citrated FC Mix tube and assess impact on diagnosis of gestational diabetes.

    Twenty antenatal patients with risk factors, booked to have an OGTT performed, consented to have additional blood samples taken at fasting (t=0), 1 h and 2 h post OGTT. Comparison between the diagnosis based on IADPSG and NICE criteria was made.

    The glucose measurement on the citrate tube correlated well with NaF/EDTA on ice, mean bias -0.03 mmol/L (y=0.99x+0.01; r2=0.990). However, the mean bias was significantly lower at -0.47 mmol/L (y=0.99x-0.41; r2=0.988) when citrate was compared to the NaF/EDTA tube at room temperature.

    Ten patients met the IADPSG criteria on the basis of glucose result using either a citrated tube or NaF/EDTA on ice while only 3 of these would be diagnosed using NaF/EDTA at room temperature alone. Three patients met the NICE criteria for gestational diabetes using the citrated tubes and 4 using NaF/EDTA tube on ice but only 1 was diagnosed using the NaF/EDTA tube at room temperature.

    The three-fold increase in patients diagnosed with gestational diabetes using the IADPSG criteria when a citrated tube was used needs verification with a larger study. The increase in cases of gestational diabetes will lead to significant resource implications.

  • P096 Socioeconomic deprivation as measured by the index of multiple deprivation and its association with low sex hormone binding globulin in women

    Adrian H Heald (Salford Royal Hospital, Salford, United Kingdom), Ian Laing (Royal Preston Hospital, Preston, United Kingdom), Andrew Hartland (Walsall Manor Hospital, Walsall, United Kingdom), Mark Livingston (Walsall Manor Hospital, Walsall, United Kingdom)

    Objective: Sex hormone binding globulin (SHBG) is a marker of insulin resistance. Given established links between BMI and socioeconomic disadvantage, we investigated how SHBG varies by index of multiple deprivation (IMD).

    Research Design and Methods: Using laboratory data from a Midlands UK population of mixed ethnicity, we examined the relation between blood concentrations of SHBG and IMD in 1160 women aged between 17 and 71 years. Women with a serum SHBG >250 nmol/L were excluded.

    Results: Mean age was 28.7 (95% CI 28.2-29.1) years; 48.2% of women were of Caucasian origin, 15.5% of Southern Asian ethnicity and 2.6% were of African or other origin (33.7% were of unknown origin). SHBG increased with age (Spearman’s; p=0.195; p<0.001). A higher proportion of women of South Asian origin vs other ethnic groups had an SHBG <30 nmol/L (OR 1.93 (95%CI 1.3-2.71)). SHBG level was lower in individuals with greater socioeconomic disadvantage as measured by IMD (Spearmans p=-0-09; p=0.004 for SHBG versus IMD). In multivariate logistic regression, IMD women in the quartiles 2-5 (higher socioeconomic disadvantage) were more likely to have an SHBG of less than 30 nmol/L (compatible with significant insulin resistance) versus quartile 1 (odds ratio (OR) 1.71 (95% confidence interval (CI) 1.17-2.53), adjusted for age (OR=0.97 (95% CI 0.95-0.98)) and ethnicity (for South Asian ethnicity OR=2.00 (95% CI 1.42-2.81) vs the rest).

    Conclusion: We have shown that lower SHBG levels in women are associated with a higher level of socioeconomic disadvantage. Given the known association between lower SHBG and higher plasma glucose, our findings suggest a link between socio-economic disadvantage and future risk of type 2 diabetes. Separately, South Asian origin is an independent determinant of lower SHBG levels.

  • P097 Demand optimised introduction of HbA1c as a test for diagnosis in primary care

    Katie Booth (NHS Grampian, Aberdeen, United Kingdom), Mark Lum (Aberdeen Royal Infirmary, Aberdeen, United Kingdom), Emma L Dewar (Aberdeen Royal Infirmary, Aberdeen, United Kingdom), Fiona M Brandie (Aberdeen Royal Infirmary, Aberdeen, United Kingdom), Ian Rothnie (Aberdeen Royal Infirmary, Aberdeen, United Kingdom), Bernard Croal (Aberdeen Royal Infirmary, Aberdeen, United Kingdom)

    Background: In 2011 the World Health Organisation (WHO) stated that HbA1c could be used for diagnosing Type 2 diabetes mellitus (DM). This prompted funding concerns across UK Pathology services given the likely significant impact that this could have on laboratory silo budgets. NHS Grampian took the decision to introduce this change in parallel with a Primary Care engagement plan and a demand optimisation intervention aimed at appropriate testing.

    Methods: As part of a Demand Optimisation Programme, Primary Care was targeted with enhanced educational feedback on the use of HbA1c for diagnosis of Type 2 DM and monitoring of both Type 1 and Type 2 DM. The introduction of this programme was aligned with other measures such as promotion through the primary care laboratory newsletter and direct engagement with the GP community. Order Communications entry also recorded whether any HbA1c request was for diagnosis or monitoring. The overall impact of this collection of interventions was assessed by comparing HbA1c requesting data for equivalent 5 month periods before and after their introduction.

    Results: The total number of HbA1c requests received in the 5 months (Aug-Dec 2016) following introduction of the new service for diagnostic purposes was 24,662, compared to 25,430 requests for the equivalent 5 month period in the previous year (p=0.247), and 25,771 requests for the preceding 5 months (p=0.374). Requests for diagnosis versus monitoring were approximately 1:3.

    Conclusions: This observational study demonstrated that despite concern, the introduction of the availability of HbA1c for diagnosis purposes did not lead to an increase in request numbers when a parallel demand optimisation approach was simultaneously introduced. The importance of good planning, developing agreed clinical protocols, and a demand optimisation strategy cannot be over emphasised and has led to a better diagnostic service without increased laboratory costs.

  • P098 Misleading HbA1c results in a case of chronic liver disease

    Jennifer D Spencer (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Julian Barth (Leeds Teaching Hospitals Trust, Leeds, United Kingdom)

    The use of HbA1c in the diagnosis and management of diabetes is well established; however, its limitations are often underestimated. HbA1c is a poor indicator of glycaemic control in the presence of conditions affecting red cell lifespan. An increase occurs in iron deficiency and alcoholism, while a decrease is seen where there is haemolysis or administration of erythropoietin. Chronic liver disease, even without cirrhosis, is also associated with lower than expected HbA1c values however the mechanism involved remains unclear. We present a case of chronic liver disease demonstrating a gradual fall in HbA1c with disease progression, despite continued poor glycaemic control. This 61-year-old female presented to hepatology with a history of alcoholic liver disease and cirrhosis. This was on a background of diabetes, iron deficiency anaemia and continued alcohol consumption. Routine investigations identified two liver lesions consistent with hepatocellular carcinomas. On examination, she was jaundiced with spider naevi but was generally well and underwent radiofrequency-tumour-ablation. However, within weeks of treatment she decompensated and over the following 10 months was admitted 7 times with recurrent episodes of hepatic encephalopathy, confusion, hepatic flap, portal hypertension and ascites. Biochemical investigations demonstrated persistently high glucose concentrations (13.9 to 31 mmol/L) despite treatment. Conversely, HbA1c concentrations over the same time frame decreased from 98 to 31 mmol/mol (<58 mmol/mol). Liver function tests were abnormal but remained relatively stable: bilirubin 30-62 µmol/L (<21), ALT 21-90 U/L (<40), ALP 254-542 U/L (70-300), and albumin 27-34 g/L (35-50). Following multiple hospital admissions, improved glycaemic control and abstinence from alcohol for over 1 year the patient’s condition improved. Glucose concentrations were maintained at 4.5 to 11.3 mmol/L and her HbA1c values remained low at 24-31 mmol/mol. There were no further episodes of hepatic encephalopathy. The conflicting glucose and HbA1c results in this case emphasised how misleading the use of HbA1c analysis can be without a clear understanding of its pitfalls in relation to certain medical conditions.

  • P099 Over 100 A&E visits, 40 raised troponins and a heterophilic antibody: a standard pathway goes wrong

    Ceri Parfitt (University Hospital of the North Midlands, Stoke-on-Trent, United Kingdom), Christopher J Duff (Department of Clinical Biochemisty, Stoke-on-Trent, United Kingdom)

    Emergency care departments commonly rely on troponin measurements in order to aid diagnosis of myocardial infarction in patients presenting with chest pain. Patients with elevated troponin measurements are quickly referred for further investigation and management. A 34 year-old male patient presented to the Emergency Department more than 100 times over a period of three years complaining of chest/back pain. He was found to have consistently elevated troponin I levels of up to 4030 ng/L. The patient underwent a number of invasive and costly investigations, including several CT-guided pulmonary angiograms, which showed no significant abnormality. Due to the discrepancy between clinical findings and laboratory results, the possibility of assay interference was investigated. A sample from the patient was analysed for troponin T at another laboratory, which showed discrepant results (troponin I: 398 ng/L (reference range <40 ng/L); troponin T: <6 ng/L (reference range <14 ng/mL). Dilution experiments with patient serum showed a non-linear decrease in troponin I concentration. A serum sample was pre-treated with heterophilic antibody blocking tubes before analysis, which resulted in a dramatic decrease in troponin I (initial concentration: 4228 ng/L; post-treatment concentration: 158 ng/L). Together, these findings suggest that the abnormal troponin I results in this patient were due to presence of heterophilic antibodies. This kind of interference is a relatively rare phenomenon, but should always be considered in individual patients where results do not fit with clinical presentation. In particular, clinicians should communicate with the relevant laboratory to discuss results where these appear unusual.

  • P100 Development of a liquid chromatography tandem mass spectrometry method for the detection of antihypertensive drugs in human urine

    Roxanne Farnon (University of Manchester, Manchester, United Kingdom), Adam MacDiarmaid-Gordon (Brighton and Sussex University Hospitals Trust, Brighton, United Kingdom), Michael Okorie (Brighton and Sussex University Hospitals Trust, Brighton, United Kingdom), Gary Weaving (Brighton and Sussex University Hospitals Trust, Brighton, United Kingdom)

    Hypertension is a significant global health burden and a leading cause of death worldwide. Antihypertensive drugs have a prominent role in the treatment of hypertension, which is directly linked to reduction of cardiovascular risk. However, the full benefit of drug treatment is often hindered by non-adherence. An objective method of monitoring adherence might increase the utility of drugs in the treatment process and lead to improved patient outcomes.

    A method was developed using an AB Sciex 4000 QTRAP ESI tandem mass spectrometer to measure 16 antihypertensive drugs. Chromatographic conditions were optimised using a Supelco Discovery HSF5 3.3 cm 3 μm column. Spiked urine was used to determine cut-offs, based on lower limits of detection and quantitation. Precision, recovery and linearity were also assessed. Thirty anonymised random patient urine samples (surplus from albumin-creatinine ratio tests) were then analysed following liquid-liquid extraction.

    Optimal peak separation was achieved using a flow gradient of 100 μL/min increasing to 600 μL/min in the first 3 minutes, then remaining at 600 μL/min for a further 4 minutes, with 50% acetonitrile. Lower limits of detection varied between 0.2 and 250 μg/L. Limits of quantitation varied between 0.7 and 775 μg/L. From this data cut-offs were determined, which ranged between 1 and 500 μg/L. Precision, recovery and linearity were deemed acceptable for a routine screening method. Antihypertensive drugs were found in 70% (n=30) of patients tested. The most commonly detected drugs were amlodipine, bisoprolol and ramipril.

    A qualitative screen has been developed for detection of antihypertensive drugs in urine. This method will undergo further refinement to validate cut-offs with analysis of more patients on known drugs. The introduction of this method into clinical practice might help optimise the drug treatment of hypertension and reduce prescribing costs.

  • P101 Clinical utility of troponin measurement in primary care samples with elevated creatine kinase

    Ushani Gayathri Wijethunga Jayawardane (King's College University Hospital NHS Foundation Trust, London, United Kingdom), David R Taylor (King's College Hospital NHS Foundation Trust, London, United Kingdom), Royce P Vincent (King's College Hospital NHS Foundation Trust, London, United Kingdom)

    Background: Cardiac enzymes creatine kinase (CK), aspartate aminotransferase (AST) and lactate dehydrogenase (LDH) are non-specific; hence the universal biochemical definition of myocardial infarction relies on the more specific cardiac troponin assays. Our laboratory currently provides AST as part of the standard liver profile. For Primary Care samples, we reflex CK testing when AST is elevated and add-on troponin if CK >1000 IU/L. In this audit we retrospectively assessed the clinical utility of this approach.

    Methods: All Primary Care troponin I results (and accompanying CK and AST data) from November 2015 to November 2016 were extracted using pathology IT systems. Clinical data was obtained from request forms and hospital electronic patient records. High-sensitive troponin I was measured using Centaur (Siemens, UK; LoD 6 ng/L) and CK and AST by ADVIA 2400 (Siemens, UK). Data is reported as median (IQR).

    Results: In total there were 294 patients out of which reflex troponin were performed in 160 (116 males) aged 35 (27-47) years. Sixty-seven patients (58%) had a measurable troponin [12 (8-21) ng/L] including seven with troponin >40 ng/L. All elevated (2 analysed out-of-hours) results were phoned to Primary Care. The CK (reference range <150 IU/L) and AST (3-35 IU/L) concentrations were 2562 (1497-5104) IU/L and 78 (63-112) IU/L respectively. Follow-up data; only one patient presented to ED the following day, one had a repeat via Primary Care and one was known to have left ventricular hypertrophy, none presented with acute coronary syndrome (ACS).

    Conclusions: Within the audit period we identified 4% of patients with unsuspected elevated (>40 ng/L) troponin. However, given the poor clinical utility the routine measurement of troponin in Primary Care samples with elevated CK is not indicated. Thus, troponin measurement should be limited to patients with clinical symptoms of ACS.

  • P102 The association of methylenetetrahydrofolate reductase gene polymorphism and hyperhomocysteinaemia with coronary artery disease in Sudanese patients with type 2 diabetes

    Nisreen O Mohammed (Ahfad University for Women, Omdurman, Sudan), Ibtisam A Ali (International University of Africa, Khartoum, Sudan), Bahaeldin K Elamin (University of Khartoum, Khartoum, Sudan), Bakri O Saeed (Sudan International University, Khartoum, Sudan)

    Objective: Individuals with Type 2 diabetes mellitus (T2DM) are at increased risk of coronary artery disease (CAD). The mutation (C677T) of methylenetetrahydrofolate reductase (MTHFR) gene is associated with elevated plasma levels of homocysteine. The association of the MTHFR gene and the level of homocysteine with development of CAD has been studied in various population groups but the results have been variable. In this study, we determined the distribution of the MTHFR genotypes and plasma homocysteine levels, and their association with CAD in patients with T2DM. Methods: We screened 226 patients with T2DM, <60 years of age. Of these, 113 had CAD confirmed by angiography (cases) and 113 had no evidence of CAD (controls). Polymerase chain reaction (PCR) and restriction fragment length polymorphism (RFLP) using Hinf1 restriction enzyme were used to determine MTHFR genotypes. Plasma homocysteine level was determined by enzymatic assay on the Hitachi Cobas Integra® 400 plus.

    Results: The frequencies of TT, CT, and CC genotypes among patients with T2DM and CAD were 16, 40, and 44%, and were 0.0, 19, and 83% in patients with T2DM without CAD. The T allele showed strong association with CAD among T2DM patients (odds ratio 6.2, p<0.0001). Patients with CAD showed higher plasma homocysteine concentrations than patients without CAD (15.4±5.8 µ/L versus 11±5.8 µ/L, p<0.001). The T allele had significant effect on homocysteine level (p<0.05). Plasma homocysteine levels were higher in individuals with TT genotype than those with CT or CC genotypes in patients with CAD (16±5, 14±5 and 12±5 µ/L (p<0.000). Homocysteine levels showed significant association with CAD (odds ratio 3.2, p<0.00).

    Conclusions: The C677T polymorphism of the MTHFR gene is associated with hyperhomocysteinaemia, and the two are independently associated with the presence of CAD in patients with T2DM.

  • P103 A case of macroCK in a patient with hypercholesterolaemia

    Julian Waldron (Royal Stoke University Hospital, Stoke on Trent, United Kingdom), Anthony Fryer (Royal Stoke University Hospital, Stoke on Trent, United Kingdom)

    A 52 year-old gentleman was referred to a lipid clinic for investigation of a raised cholesterol. He had been commenced on atorvastatin but cholesterol level remained elevated (7 mmol/L). Patient was not diabetic, had normal blood pressure and there was no family history of cardiovascular disease. Blood test results revealed normal renal, liver and thyroid function but an elevated creatine kinase (CK) of 2500 U/L (reference range 15-185). A repeat test after 2 weeks off statin therapy was 2200 U/L. The patient was a builder but denied any muscle aches or pains.

    Further screening for causes of a raised CK were arranged including urine organic acids and acylcarnitines but results were unremarkable. In light of the lack of symptoms a screening test for macroCK was arranged. This is a benign cause for an elevated CK. The test was carried out by polyethylene glycol (PEG) precipitation and revealed a percentage recovery of 3% (reference range >50%). The results were consistent with macroCK. The patient has now been recommenced on statin therapy.

    This case highlights that baseline measurement of CK should be undertaken prior to commencing statin therapy and the importance of screening for macroCK in patients with an unexplained raised CK.

  • P104 Mutation detection rate of Randox Familial Hypercholesterolaemia Array in a multi-ethnic patient population with clinically diagnosed definite familial hypercholesterolaemia

    Sukhbir Kaur (Heart of England NHS Foundation Trust, Birmingham, United Kingdom), Christina Jewkes (Heart of England NHS Foundation Trust, Birmingham, United Kingdom), Craig Webster (Heart of England NHS Foundation Trust, Birmingham, United Kingdom), Alan F Jones (Heart of England NHS Foundation Trust, Birmingham, United Kingdom)

    Familial hypercholesterolaemia (FH) is a genetic disorder that is characterised by high levels of circulating low-density lipoprotein cholesterol (LDL-C) due to defective clearance. The Dutch Lipid Clinic Network (DLCN) criteria can be used to clinically determine whether a diagnosis of FH is unlikely, possible, probable or definite based on family history, clinical history, physical signs and serum LDL-C concentration. However, genetic testing is helpful for patients with low Dutch scores with poor family history and in cascade testing relatives. The Randox FH Array screens for 40 mutations in 3 genes (LDL-receptor [LDLR], apolipoprotein B [APOB] and proprotein convertase subtilisin/kexin type 9 [PCSK9]) said to be responsible for 70-80% of FH cases in UK and Ireland. Since the Heart of England NHS Foundation Trust treats a multi-ethnic patient population, the effectiveness of the FH Array to detect mutations in definite FH patients of different ethnic backgrounds was evaluated.

    Whole blood was collected with informed consent from 32 lipid clinic patients with definite FH on DLCN criteria (10 Caucasian, 11 South Asian and 11 Afro-Caribbean). DNA was extracted and analysed by the FH Array for 38 LDLR, 1 APOB and 1 PCSK9 mutations. The array identified mutations in 5 (16%) of the 32 patients: 4 were Caucasian and 1 was Afro-Caribbean. The detection rates within the Caucasian, Afro-Caribbean and South Asian groups were 40%, 9% and 0%, respectively.

    These results suggest that different mutations may exist in different ethnicities and the FH Array may be useful in screening a predominantly Caucasian population. However, due to the presence of other mutations and the large number of such mutations, targeted mutation screening may not be the best genetic approach for identifying mutations in ethnically diverse communities. Although costly, the most appropriate method for detecting mutations in such populations may be whole gene sequencing.

  • P105 Analytical variation at extreme high density lipoprotein cholesterol concentration

    Louise J Ward (Viapath, Bedford Hospital NHS Trust, Bedford, United Kingdom), Jane Williams (Viapath, Bedford Hospital NHS Trust, Bedford, United Kingdom; Derby Teaching Hospitals NHS Trust, Derby), WS Wassif (Viapath, Bedford Hospital NHS Trust, Bedford, United Kingdom)

    Grossly elevated serum HDL-C results are a relatively recent phenomenon and its significance uncertain. Possible causes include dysfunctional HDL-C or analytical interference. Additional analysis of apolipoprotein A1, apolipoprotein B and direct LDL-C has been used to investigate some patients with high HDL-C but these tests did not produce any additional useful information. We analysed lipid profiles on 54 subjects; 36 with HDL-C ≥3.5 mmol/L and 18 with HDL-C ≤0.5 mmol/L using two different platforms (Roche COBAS 6000 and Siemens Advia 2400). Significant difference in HDL-C at high concentrations was noted between the two methods with Siemens Advia reporting lower results on all samples (mean mmol/L ± SD, 3.3±0.4) compared to Roche Cobas (4.0±0.5, r=0.88, p=<0.001). The methods correlated well for total cholesterol and triglycerides (r=0.988 and r=0.997; respectively) with no significant difference noted. Interestingly, the calculated parameters reported by each method, LDL-C and non HDL-C, were also significantly different, although to a lesser extent (p=0.04 and p=0.03; respectively). Although poor correlation was observed at low HDL-C concentrations, no significant difference was noted between the two methods; Siemens Advia (0.6±0.3), Roche Cobas (0.4±0.2, r=0.38, p=0.06). No other significant difference was noted in the remaining lipid profile. In contrast to analytical performance at low HDL-C concentrations, there is uncertainty of analytical performance at extreme high HDL-C concentration. Such difference at high HDL cannot be explained by uncertainty of measurement since both Siemens and Roche have similar RCV values (20.8% & 21.2%; respectively). In summary, grossly elevated HDL-C concentrations should be interpreted with caution in the light of clinical scenario and individual patient. In those patients with occlusive vascular disease, such high HDL-C may not be cardioprotective and may reflect assay interference or dysfunctional HDL-C.

  • P106 Effect of PCSK9 inhibition with alirocumab on lipoprotein (a) in patients with high risk of cardiovascular disease

    Samreen Safdar (Queen Elizabeth Hospital, Birmingham, United Kingdom), Rob Cramb (Queen Elizabeth Hospital, Birmingham, United Kingdom), Charlotte Dawson (Queen Elizabeth Hospital, Birmingham, United Kingdom)

    Background: Association of atherosclerotic cardiovascular disease and abnormal lipoprotein metabolism is well established and is associated with high morbidity and mortality. More targeted therapies, beyond statins, are emerging to effectively reduce highly atherogenic lipoproteins particularly in patients with familial hypercholesterolemia and failure/intolerance to statins. Studies have shown that, alirocumab, a monoclonal antibody that inhibits the proprotein convertase subtilisin/kexin type 9 (PCSK9), significantly lowers the low density lipoprotein cholesterol (LDL) and apolipoprotein B100 (apoB100), but the effect on lipoprotein (a) regulation and clearance is shown in limited studies only. Moreover, role of LDL receptors in clearance of lipoprotein (a) is poorly defined. Therefore, we measured the effect of alirocumab on lipoprotein (a) in patients with high risk of cardiovascular disease.

    Methods: Thirteen patients, with either heterozygous familial hyperlipidaemia (FH), primary non-FH or mixed hyperlipidaemia and LDL-cholesterol level of >3.5 mmol/L, despite maximum tolerated dose of other lipid lowering agents, received 6 doses of alirocumab, 150 mg two weeks apart via subcutaneous injections. Total cholesterol, LDL cholesterol, non-HDL cholesterol and Lipoprotein (a) levels were measured before and after treatment.

    Results: Seven patients had FH and six had either primary non-FH or mixed hyperlipidaemia. In the 7 patients with FH, alirocumab reduced LDL-C by 60%, non HDL-C by 72% and lipoprotein (a) by 35%. Whereas, in other high-risk patients, LDL-C is reduced by 58%, non-HDL-C by 60% and lipoprotein (a) lowered by 31%.

    Conclusion: Alirocumab, significantly reduced LDL-C, non LDL-C and lipoprotein (a) by increasing availability of LDL receptors to clear these highly atherogenic particles. Whilst patients with FH are relatively resistant to the lipid lowering effect of statins, their response to alirocumab is similar to that of the patients with non-FH dyslipidaemia.

  • P107 Analytical sensitivity of current troponin assays

    Gareth J Davies (WEQAS, Cardiff, United Kingdom), Gwyn Pritchard (WEQAS, Cardiff, United Kingdom), Mary A Thomas (WEQAS, Cardiff, United Kingdom)

    Introduction: The aim of the study was to determine the intra-laboratory variation at the upper limit of normal (99th percentile) and near the limit of detection of two hsTroponin assays in Wales.

    Method: A base pool of human serum from a healthy donor was spiked with ternary Troponin complex (ITC) to an approximate concentration of 35 ng/L. Doubling dilutions using the base serum were prepared to give final concentrations ranging from 2.3 to 29.3 ng/L for the Abbott TnI and 3.8 to 38.2 ng/L for the Roche hs-cTnT. Ten sets of 4 pools were dispatched to all laboratories and stored at -20°C. The assay protocol consisted of two replicates per pool per run, and two runs per day for 5 days (10 runs). Each laboratory carried out 80 assays.

    Results: For the Abbott TnI method (n=7), the coefficient of variation (CV) at a concentration of 15.7 ng/L varied between 3.5% and 6.6%. At a concentration of 8.8 ng/L all laboratories had CV <8%. All laboratories had CV <10% at 5 ng/L with CVs ranging from 10.8 to 15.2% at 2.3 ng/L. For the Roche hs-cTnT (n=10), the CV at a concentration of 13.9 ng/L varied between 1.5% and 4.8%. All laboratories achieved a CV of <5% at 12 ng/L and with the exception of 1 laboratory, all achieved a CV <10% at 3.8 ng/L.

    Conclusion: The study has established the performance at or near the limit of detection of two hsTroponin assays. The data suggests that the performance for these 2 assays are acceptable for use as part of the 0 and 1 hr rule in/rule out algorithm in the ECS guidelines for the management of acute coronary syndrome in patients presenting without persistent ST-segment elevation.

  • P108 Low HDL cholesterol and Apo A1 on fenofibrate therapy

    Thomas G Morris (St George's Hospital, London, United Kingdom)

    A 60 year-old male presented with DM, hypertriglyceridemia and a BMI of 34.9 (target <25.0). He was a heavy drinker and ex-smoker (30 cigarettes a day, 1961-2008). He has no family history of premature vascular disease but previously had high total cholesterol, which peaked in 2006 at 7.8 mmol/L with a total cholesterol/HDL cholesterol ratio of 9.8 (ref. range 0.0-5.0). Since then, he has been on lipid lowering medication: atorvastatin (80 mg a day), ezetimibe (10 mg a day) and fenofibrate (267 mg a day), and his level of total cholesterol returned to within normal reference ranges. It was noted by his GP in late 2015 that his levels of HDL cholesterol had decreased to 0.10 mmol/L (ref. range 0.9-1.9) and further investigation found that his Apo A1 concentration was also reduced at 0.21 g/L (ref. range 1.04-2.02). Following these findings the patient was referred to the lipid clinic at St George’s Hospital and it was recommended that he should discontinue his use of fenofibrate and be reviewed in 2 months’ time. At review, it was found that the patient’s HDL cholesterol and Apo A1 levels had both increased to within the normal reference ranges (0.99 mmol/L and 1.43 g/L, respectively) and his total cholesterol and total cholesterol/HDL cholesterol remained normal. This case highlights that although fenofibrate is a well-recognised medication for treating lipid abnormalities, specifically hypertriglyceridemia, it may also have some unwanted side-effects on HDL cholesterol in a small number of patients. It has been reported in a handful of publications that fenofibrate can reduce HDL cholesterol, although the mechanism of action is unknown. Even though this is a relatively uncommon occurrence, clinicians and scientists should be aware of this phenomenon.

  • P109 Seventy years of the Journal of Clinical Pathology: contributions to clinical biochemistry

    Tahir S Pillay (University of Pretoria, Pretoria, South Africa)

    In 2017, the Journal of Clinical Pathology (JCP) will celebrate 70 years. Although a multidisciplinary journal, many of the publications have been in clinical biochemistry. The aim was to examine the most highly cited papers in clinical biochemistry from a historical perspective. The Web of Science was used for this purpose.

    Of the top 5 most highly cited papers, 3 are in clinical biochemistry. The most highly cited paper in the JCP is a publication on a method for the analysis of urea by Fawcett and Scott, published in 1960. This is followed by the description of Mattingly in 1962 on a simple fluorometric method for the estimation of free 11-hydroxycorticoids and a paper by Trinder on the determination of blood glucose using an oxidase-peroxidase system. The measurement of urea was based on the hydrolysis of urea with urease and the generation of ammonia which was used detected using a colour reaction. The 1960 publication improved the sensitivity allowing the method to be used on 10 µL of plasma, a considerable advance at that time. The first method for estimated cortisol and corticosterone using fluorescence was developed in 1954. The method of Mattingly made it possible for a single person to complete 6 plasma estimations in 90 minutes and this could be used to confirm adrenal insufficiency within one hour of having taken blood; 18-24 estimations could be performed in a working day. The method of Trinder involved the development of a manual and automated method for measuring glucose via oxidase/peroxidase and then with adrenaline as an oxygen acceptor. The automated method was able to perform 40 glucose estimations/hour which was also a major advance in 1969.

    This analysis does provide an historical overview of some of the fundamental methods developed in the 1960s and the contribution of a multidisciplinary journal such as the JCP to clinical biochemistry.

  • P110 Does linking a chemical pathology learning point to a relevant fictional character enhance knowledge acquisition?

    Michelle Muscat (Nottingham Trent University, Nottingham, United Kingdom), Gren Ireson (Nottingham Trent University, Nottingham, United Kingdom), Ruth Richards (Nottingham Trent University, Nottingham, United Kingdom), Mark Turner (Nottingham Trent University, Nottingham, United Kingdom)

    Objectives: The objective of this study was the elucidation of the hypothesis as to if linking a chemical pathology learning point to a relevant fictional character enhances knowledge acquisition.

    Methods: Two validated questionnaires were distributed, one to a medical student cohort (n=88), and the other to a school age student cohort (n=678) Both these questionnaires included the question as to if participants would enjoy relevant movie clip incorporation within a teaching session, as well as another question as to their personal preference with respect to inclusion of stories. A separate questionnaire was distributed at a Comic Convention to a sizeable number of participants (n=542). This convention questionnaire specifically elicited, not just general preferences to story inclusion, as did the previous questionnaires, but also specific questions as to linking learning points to a fictional character.

    Results: In both the medical cohort as well as the school-aged children cohort, the positive response towards reported clarity of use of film clips and inclusion of stories in a chemical pathology related session was statistically significant. However, in the medical student group, movie enjoyment was not a predictive variable in the proportional odds model for interest in chemical pathology, whereas story inclusion was. The comic convention questionnaire responses with respect to motivation and helpfulness of linking a learning point to an animated character in remembering a topic were significant.

    Conclusions: Linkage of learning points to fictional characters may have a definite niche in the education of chemical pathology, albeit may be favoured to varied extents by different individuals.

  • P111 A somewhat ‘cameo appearance’ of gamma-glutamyltransferase: a magical garment girl using alcohol to break a spell

    Michelle Muscat (Nottingham Trent University, Nottingham, United Kingdom)

    In a conjectural universe where a boy is resurrected as a zombie by a necromancer, and a young girl is cursed to assume the semblance of a middle aged man, being able to revert to her original self when drunk, it is interesting, somewhat unexpected and refreshing to hear the said character specifically voice concern that her gamma-glutamyltransferase (GGT) levels were abnormally high. In such a show where comedy reigns supreme, a young girl, who temporarily takes the semblance of an older man, complains specifically that in that form her gamma-glutamyltransferase (GGT) and cholesterol levels in the blood were elevated.

    In “Kore wa Zombie Desu ka?” (written by Shinichi Kimura and adapted to anime by Studio Deen), the girl in question was constantly seen to be drinking alcoholic beverages, which were somehow related to breaking the spell that made her assume the body of a middle-aged man. Gamma-glutamyltransferase (GGT) is mentioned briefly, in a somewhat ‘cameo appearance’ style in episode 7 of the second series, “Kore wa Zombie Desu ka? Of the Dead”. It is worthy of note that this line was modified in the English adaptation to reflect terminology more well known to the general public, namely reference to enlargement of the prostate with advancing age. Gamma-GT, which made somewhat of a cameo appearance in this show, is indeed one of the traditional and well known biochemical indicators of alcohol abuse, however lacks high specificity for this purpose.

    Some other biochemical markers of alcohol consumption include use of carbohydrate-deficient transferrin (CDT), phosphatidylethanol (PEth) and urine ethyl glucuronide (EtG) and ethysulfate (EtS). The ratio of serum aspartate aminotransferase to alanine aminotransferase (AST/ALT ratio) can be used as a potential indicator of alcoholic liver disease.

  • P112 New facets of chemical pathology education: an unconventional weekend teaching break

    Michelle Muscat (Nottingham Trent University, Nottingham, United Kingdom), Gren Ireson (Nottingham Trent University, Nottingham, United Kingdom), Ruth Richards (Nottingham Trent University, Nottingham, United Kingdom), Mark Turner (Nottingham Trent University, Nottingham, United Kingdom)

    Objectives: There is a growing impetus towards implementation of newer multimodal teaching modalities. This study assesses the impact of a hotel weekend teaching break, with incorporation of relevant video clips and specifically drawn up poetry, on the learning experience for medical students.

    Methods: A preliminary study was carried out first with regards to potential interest in such a teaching weekend. The subsequent study proper involved feedback obtained through designated questionnaires distributed during a two-day teaching weekend. The interested 36 attendees were identified via distribution of event flyers amongst class representatives, and the event was also included in a local medical magazine. Data analysis involved quantification and graphical output of the various Likert-like scale responses.

    Results: In the preliminary study, a total of 102 responded ‘Yes’ whereas 54 answered ‘No’ (n=156). The test of sample proportions confirmed a significant difference between the two groups. Out of the 36 attendees of the weekend seminar, 31 completed the questionnaire. A trend was first assessed for personal enjoyment of movies and poetry. A total of 61.3% of participants selected the option that engaging in fun activities related to the chemical pathology topic they are studying is ‘very helpful’ for them recollect it better and 22.6% selecting ‘moderately helpful’. Also, 64.5% selected that incorporation of a relevant story within the chemical pathology teaching weekend met their needs ‘very well’. The majority of participants favoured movie clips to a greater extent than poetry.

    Conclusions: Adoption of non-traditional formats outside the traditional monolithic lecture and tutorial theatres aids the learning experience.

  • P113 Investigation of raised platelet and white cell counts as a cause of pseudohyperkalaemia

    Rosie Forster (University Hospitals Birmingham NHS Foundation Trust, Birmingham, United Kingdom), Karen Smith (University Hospitals Birmingham NHS Foundation Trust, Birmingham, United Kingdom)

    Introduction: Pseudohyperkalaemia is a relatively common laboratory finding, with a major cause being a delay in separation in serum samples. There is also an association between high platelet and white cell (WBC) counts and falsely elevated potassium in serum samples due to the release of potassium during the clotting process. This can be circumvented by the use of lithium heparin as an anti-coagulant. During the validation process, a raised serum potassium in the presence of a raised platelet or WBC count may be considered to be artefactual. The aim of this project was to investigate the correlation of raised platelet/white cell counts with either pseudohyperkalaemia or true hyperkalaemia. This data will help to assess if further care is needed to identify true hyperkalaemic patients, which is important, as hyperkalaemia is a medical emergency.

    Methods: WBC and platelet count data was gathered from the laboratory IT system for patients with paired serum and plasma potassium measurements between May 2015 and March 2016, giving a dataset of 337 patients from the primary care population, and 285 patients from the hospital population (both in- and outpatients). The difference between the serum and plasma potassium results was calculated. Linear regression analysis was performed separately for the platelet count and WBC count against this difference in each population.

    Results: The correlation between platelet count or WBC count and the difference between the serum and plasma potassium result was weak in all of the groups studied (Rs in range 0.08 to 0.28), apart from a moderate correlation for platelet count in the hospital population (Rs=0.58).

    Discussion: When technically validating results, laboratory staff should not dismiss elevated serum potassium in patients with elevated platelets or white blood cells as pseudohyperkalaemia without further investigation.

  • P114 The value of vetting sendaway requests by the duty biochemist

    Eleanor McLaughlan (York Teaching Hospital NHS Foundation Trust, York, United Kingdom), Claire Chapman (York Teaching Hospital NHS Foundation Trust, York, United Kingdom), Deepak Chandrajay (York Teaching Hospital NHS Foundation Trust, York, United Kingdom), Daniel Turnock (York Teaching Hospitals NHS Foundation Trust, York, United Kingdom)

    Introduction: Checking the validity and clinical appropriateness of tests sent to external laboratories is commonly undertaken by many clinical biochemistry departments. Historically in our trust, requests from one laboratory site (York) were vetted in this way but requests from the other site (Scarborough) were not checked by the duty biochemist prior to dispatch. Tests requested by specialists were generally not vetted prior to referral from either laboratory site. The aim of this work was to look at the impact of introducing vetting procedures for sendaway requests in Scarborough and to assess potential impact from introducing minimum re-testing intervals for certain sendaway tests requested by specialists.

    Methods: A total of 8 weeks’ worth of sendaway tests from Scarborough were reviewed from historical records to identify the number and type of tests being rejected by the Duty Biochemist prior to dispatch. Retrospective data extraction from LIMS (Telepath) was used to identify requests for free light chains, thiopurine metabolites and P3NP from specialists that had been referred externally.

    Results: A total of 77/463 requests (17%) of sendaway requests from Scarborough were rejected by the Duty Biochemist prior to dispatch. Of the rejected tests, only 3 were subsequently repeated. The five most frequently rejected requests comprised 60% of the total rejected requests. The estimated minimum cost saving from sendaways vetting in Scarborough was £9126 annually, compared to an estimated average cost of £2369 annually in duty biochemist time to perform vetting. Use of minimum re-testing intervals for free light chains, thiopurine metabolites and P3NP would identify a total of 166 inappropriate requests across both laboratory sites in 1 year.

    Conclusion: There is significant recurrent saving to the laboratory from vetting sendaway tests. The application of agreed minimum re-testing intervals to requests from specialists will provide further cost savings in some instances.

  • P115 A re-audit of routine sample flow through the Blood Sciences Department

    Joseph M Taylor (Royal Liverpool University Hospital, Liverpool, United Kingdom), Charlotte E Harborow (Royal Liverpool University Hospital, Liverpool, United Kingdom), Lisa Bailey (Royal Liverpool University Hospital, Liverpool, United Kingdom)

    Background: Annually, an audit is performed to examine sample flow and evaluate specimen reception processes to identify potential efficiency improvements. Previous audits highlighted delays in analysis and turn-around times with corrective action being of limited success. The 2016 re-audit was of particular interest due to the implementation of new staff rotas giving the opportunity to audit pre and post implementation.

    Methods: Routine requests were chosen at random from specimen reception during normal working hours. Patient ID, requesting location, time of sample collection and time of arrival were recorded. Time booked in, analysed and authorised were collected retrospectively from the LIMS and turnaround times calculated. Audit standards were: 1) 95% of samples booked into the LIMS <2 hours of arrival, 2) 95% of samples analysed <4 hours of arrival and 3) 90% of samples analysed <6 hours of collection.

    Results: Pre-rota conformation of audited requests to standards 1, 2 and 3 for biochemistry was 84%, 94% and 94% respectively. Haematology conformation was 97%, 100% and 79%. Post-shifts, biochemistry conformation was 86%, 93% and 89% and haematology 86%, 81% and 52%. Segregating data by request location revealed inpatient work was consistently meeting audit standards.

    Conclusion: Both departments showed improved performance compared to previous audits. Reasons for this include increased staff recruitment, presentation of audit data in more appropriate formats and greater involvement of specimen reception staff in audit processes, communication of findings, consultation on perceived problems and generation of corrective action. Post-rota data highlighted areas for further improvement and rota adjustments were made to ensure staff availability at times of high workload. Segregation of data into inpatient and outpatient requests can give further useful information. In our audit this showed steps taken to prioritise workflows and ensure internal hospital requests are dealt with quickly were successful.

  • P116 The creation of a partial retrospective clinical validation review system

    Jonathan Vernazza (Gateshead Health NHS Foundation Trust, Gateshead, United Kingdom), Ian Dunn (Gateshead Health NHS Foundation Trust, Gateshead, United Kingdom)

    The aim of the project is to introduce a new ‘partial retrospective clinical validation review’ system. Real time review allows results to be trapped on the basis of validation limits for review prior to release. Retrospective review allows results to be issued directly, prior to being pulled back for retrospective clinical review. Partial retrospective review allows the laboratory to dictate different sets of results which can be released by real time review and by retrospective review.

    A series of rules were set up in the middleware system, which add a validation test to any sample with an analyte which is outside the validation limits. The validation test is held in the middleware system until all tests within that sample are complete (this avoids the potential for subsequent results to become available and miss review). A series of rules were set up to allow a subsequent validation test to be added to any samples with additional requests which require review, after the initial batch of tests have been reviewed. The validation tests are sent to the laboratory information management system, then a list of outstanding validation tests can be pulled for retrospective clinical review. Multiple validation tests were set up to allow different tests to be grouped within a sample, for example routine chemistry tests and routine endocrine tests.

    Test scripts with tester patients have been used to confirm the system is working as expected and test scripts with live patients have been used to validate the system.

    A new ‘partial retrospective clinical validation review’ system has been created, which allows users to define which results require different types of clinical review. This has the benefit of reducing turn- around times for targeted locations, and allowing laboratory staff to prioritise workloads and target results which require detailed clinical review.

  • P117 Duty biochemists beware! Think? pseudohypokalaemia

    Amy Kiley (Kettering General Hospital, Kettering, United Kingdom), Gwyn McCreanor (Kettering General Hospital, Kettering, United Kingdom)

    Case 1: A consultant haematologist enquired with the duty biochemist with regards to a patient under their care with a discrepancy in the potassium results provided by the laboratory and a blood gas analyser. Potassium on a venous blood sample collected in the emergency department was 1.9 mmol/L, however blood gas analysis performed at the same time gave a potassium on 3.6 mmol/L. The patient had been referred to the haematologist for a new diagnosis of acute myeloid leukaemia, with a white cell count of 495 x 109/L. The biochemist advised of the need to exclude pseudohypokalaemia in the laboratory due to routine use of serum samples compared to the lithium heparin blood gas syringe. Paired serum and lithium heparin venous blood samples analysed in the laboratory revealed a serum potassium of 2.3 mmol/L and plasma potassium 5.5 mmol/L.

    Case 2: An out-of-hours BMS obtained a potassium of 1.2 mmol/L on a sample sent by the emergency department. Identifying the result as invalid, a repeat sample was requested which gave a result of 1.8 mmol/L. The BMS contacted the on-call biochemist for assistance, and was advised to perform a blood gas analysis. A fresh sample in a lithium heparin blood gas syringe sample gave a potassium result of 3.5 mmol/L. Further investigation revealed the patient had received a new diagnosis of acute leukaemia, with a white cell count of 312 x 109/L.

    Pseudohypokalaemia is a rare phenomenon that may occur in hyperleukocytosis due to cellular metabolism in vitro. It may be identified by observation of discrepancies between lithium heparin and serum samples, or spuriously low serum potassium. These cases reflect the need to identify the phenomenon in order to facilitate appropriate patient management.

  • P118 Raising the profile of laboratory medicine research

    Owen J Driskell (NIHR CRN West Midlands, Stafford, United Kingdom)

    Research is a core activity in the NHS. We need the evidence from research to deliver better care. The Accelerated Access Review and the National Institute for Health Research (NIHR) Push the Pace initiative recognise the need to speed up patients’ access to innovative healthcare and how research findings should influence patient care more quickly. Pathology’s contribution can often be an afterthought.

    NHS research infrastructure has changed significantly in the last ten years at a time when pressures reduced laboratory consultant research time. The Association for Clinical Biochemistry and The Royal College of Pathologists (Consultants in Clinical Biochemistry: The future, May 2009) recommended professional bodies and other stakeholders in clinical biochemistry should promote the opportunities for innovation, research and development and encourage consultants to increase their activity in this area.

    NHS laboratory scientists lead research and a spectrum of improvement activates that are incorporated into practice. However, obtaining funding, navigating the ethics and research governance procedures, and demonstrating the patient benefit, can sometimes be a challenge.

    NHS laboratories also play a vital role in delivering NHS research. Over a third of NIHR Clinical Research Network (CRN) portfolio studies have pathology as an element of either inclusion or exclusion criteria and a fifth of studies that state research outcomes on the portfolio include pathology elements. Despite this, in a recent survey, a third of laboratories report no research lead, a third reported no research coordinator role and only 14% of laboratories actively manage the set up and delivery of studies to achieve metrics supporting the timeliness of NHS research. This indicates room for improvement.

    Engaging with NHS research infrastructure can raise the profile of pathology leading to development opportunities. This poster will describe the work of the Lead for Laboratory Medicine to promote pathology departments and laboratory scientists in research.

  • P119 Visualising requesting patterns in primary care

    Philippa J Croxford (Heart of England Foundation Trust, Birmingham, United Kingdom)

    Introduction: On a national scale it has been shown that there is significant variation in clinician’s pathology requesting patterns. Some of this is due to patient population variances, but a significant proportion is due to differences in clinical practice and interpretation of guidelines (Smellie 2002). We propose that, by creating a data visualisation tool to display these differences within our local GP population, we can provide clinicians with a measure of their practice against other local GPs. It is anticipated that the information will generate dialogue amongst GPs into the most appropriate use of the tests we offer. The desired outcome for this project would be to decrease variability and reduce inappropriate test requesting.

    Methods: We have created a website using open source software (Python, Django and D3.js) to create interactive data visualisations.

    We presented this data at local GP forums and collected feedback responses. GPs will eventually have direct access to the website for GPs so that we can assess the long-term impact of this service.

    Results: Feedback from local GP forum indicates support for this project. When asked if this was a service they would like to use the median result was Agree (4 on the Likert scale, n=15). We have found a strong correlation between those who understand the data and those who would like to see this development implemented (p=0.0118). We could conclude that a better explanation of the data is necessary for those GPs not supporting this project.

    This is a new concept for local GPs, since only 11% of those questioned (n=17) reported that they analysed their pathology requesting levels and none of them compared their requesting levels to other practices.

  • P120 Evaluation of 7 point-of-care human chorionic gonadotropin devices for susceptibility to high dose hook in the first trimester of pregnancy

    Sarah J Glover (Harrogate and District NHS Foundation Trust, Harrogate, United Kingdom), Stephen R Goodall (Bradford Teaching Hospital NHS Foundation Trust, Bradford, United Kingdom)

    Background and Aim: Previous reports have demonstrated urine point of care testing (POCT) human chorionic gonadotrophin (hCG) pregnancy tests can exhibit the high dose hook effect. Following several reports of false negative POCT pregnancy tests at Bradford Royal Infirmary (BRI), we aimed to evaluate 7 POCT pregnancy hCG devices commonly available and used in UK hospitals.

    Methods: Spare urine samples routinely collected from 5 pregnant women (6-11 weeks gestation) attending the Early Pregnancy Assessment Unit at BRI were sent to the Blood Sciences laboratory. Neat samples were diluted 1:10 and 1:100 using urine collected from a male control. Neat and diluted samples were analysed using 6 manual and 1 automated POCT pregnancy test device according to the package inserts. Results of the manual kits were interpreted by 10 untrained personnel.

    Results: Visible differences in POCT device performance were evident. Four devices performed well with 100% of readers interpreting all results correctly as positive. Three devices failed to generate a result considered positive by all readers and 2 of these devices had poor performance, with less than 50% of readers interpreting a result as positive in some cases.

    Conclusion: This study demonstrates several commonly used POCT urine pregnancy testing kits are subject to false negative results due to the high dose hook effect, when analysing samples from women in the second half of their first trimester of pregnancy. Our findings should be a major concern for healthcare professionals using POCT pregnancy tests and all such kits should be fully evaluated and deemed fit for purpose before use.

  • P121 An evaluation of the performance of the Point of Care Test i-CHROMA™ ferritin method in the Randox International Quality Assessment Service

    Olu Coker (JB Consulting MDP Limited, Upper Heyford, United Kingdom), Suman Bains (JB Consulting MDP Limited, Upper Heyford, United Kingdom), John Bolodeoku (JB Consulting MDP Limited, Upper Heyford, United Kingdom)

    The estimation of ferritin levels plays an important role in the diagnosis, monitoring and management of haemochromatosis and iron deficiency disorders in men, women and children. Ferritin can be measured using immunoassays e.g. ELISA, immunochemiluminescence and immunoturbidimetric assays. As a result of the variable structure of ferritin, different antibodies to ferritin may not react equally with all forms, therefore sometimes making the method not comparable. The i-CHROMA™ ferritin is a POC lateral flow chromatography, fluorescence immunoassay (FIA) for the quantitative determination of ferritin in serum or plasma. In this study, the performance of the i-CHROMA™ ferritin method was evaluated with other laboratory ferritin methods using correlation analysis and Bland-Altman difference plots using Cycle 41 from the Randox International Quality Assessment Service (RIQAS). The results showed that the estimations of the POCT i-CHROMA™ ferritin correlated well with the all method means: Abbott {Architect Chemiluminescence (r²=0.91), Beckman {AU400/600/640/2700/5400 (r²=0.00055), DxI 600/800 (r²=0.92) Access/LXi725 (r²=0.92), Biomerieux {VIDAS (r²=0.90)}, Roche {COBAS® C501/502/701 (r²=0.92), Elecsys/COBAS® e411 (r²=0.92), COBAS® 4000/c311 (r²=0.93) Modular E170/COBAS® e601/e602 (r²=0.92), Siemens (Centaur CP (r²=0.91), Centaur XP/XPT/Classic (r²=0.92), Dimension (r²=0.92), DPC Immunilite 2000/2500, (r²=0.91) DPC Immunilite 1000 (r²=0.91), Diasorin Liaision (r²=0.93) Monobind Inc ELISA (r²=0.96), and Ortho Vitros 3600/5600/ECi (r²=0.93). However, the method biases ranged from -95.37 to -20.2 suggesting that there is an under-estimation with the i-CHROMA™ ferritin method compared to the others. In summary, the POC i-CHROMA™ ferritin method correlated very well with the traditional laboratory methods in the RIQAS, although it was under-estimating compared to the other methods.

  • P122 The performance of a quantitative thyroid stimulating hormone point of care test method using the i-CHROMA™ TSH method in the Randox International Quality Assessment Service

    Olu Coker (JB Consulting MDP Limited, Upper Heyford, United Kingdom), Suman Bains (JB Consulting MDP Limited, Upper Heyford, United Kingdom), John Bolodeoku (JB Consulting MDP Limited, Upper Heyford, United Kingdom)

    The estimation of thyroid stimulating hormone (TSH) levels in the blood plays an important role in the diagnosis of hypothyroidism. There are several qualitative point of care (POC) TSH methods available such as Thyrotest Rapid TSH, ThyroChek, HomeTest Thyroid, Rapid Response, ThyroScreen and Vivacta Thyroid test designed to identify TSH at concentration >5 µIU/mL. The i-CHROMA™ TSH is a lateral flow chromatography, fluorescence immunoassay (FIA) for the quantitative determination of TSH in serum or plasma with a sensitivity of 0.1 µIU/mL. In this study, the performance of the i-CHROMA™ POC TSH method was evaluated alongside other laboratory TSH method means using correlation analysis and Bland-Altman difference plots using Cycle 38 samples from the Randox Internal Quality Assessment Service (RIQAS). The results showed that the estimations of the POCT i-CHROMA™ TSH correlated well with the all method means: Abbott {AxSyn (r2=0.96) Ultrasensitive (r2=0.96), Architect (r2=0.97), Axsym 3rd generation (r2=0.96); Biomerieux {VIDAS TSH (r2=0.98) VIDAS TSH3 Ultrasensitive (r2=0.96); Roche COBAS® 6000/8000 (r²=0.97) COBAS® 4000/e411 (r²=0.97) Modular 170 (r²=0.97) Elecsys (r²=0.97) Siemens Centaur XP/XPT/Classic (r2=0.97) Siemens Centaur XP/Classic 3rd generation (r²=0.96) Centaur XP/Classic TSH3-Ultra (r²=0.97) Siemens DPC Immunilite 2000/2500 (r²=0.97), DPC Immunilite 1000 (r²=0.97) Beckman DXI 600/800 Fast TSH (r²=0.97) Hyper TSH (r²=0.97), DXI 600/800 1st generation (r²=0.98) Diasorin Liaision (r²=0.96291), Monobind (r²=0.96) Access/LXi725 hyper TSH 3rd generation (r²=0.97) and Ortho Vitros 3600/5600/ECi (r²=0.97). The method biases ranged from -1.15 to 0.64 therefore indicating that there was no major bias with the i-CHROMA™ TSH method. In summary, the POC i-CHROMA™ TSH method correlated very well with the other traditional laboratory methods in the RIQAS.

  • P123 The performance of the point of care test i-CHROMA™ prostate specific antigen and alpha-fetoprotein methods using distributions from the Randox International Quality Assessment Service

    Suman Bains (JB Consulting MDP Limited, Upper Heyford, United Kingdom), Olu Coker (JB Consulting MDP Limited, Upper Heyford, United Kingdom), John Bolodeoku (JB Consulting MDP Limited, Upper Heyford, United Kingdom)

    The estimation of faecal occult blood (FOB), prostate specific antigen (PSA), and α-fetoprotein (AFP) are a few of the tumour markers that are undergoing evaluation for screening than are estimated on the i-CHROMA™. In this study, the performance of the PSA and AFP methods were evaluated with other laboratory methods using Cycle 38 for PSA and Cycle 40 for AFP from the RIQAS. The POCT i-CHROMA™ PSA correlated very well with all methods: the best: Abbott {Architect (r²=0.99) AxSym monoclonal (r²=0.99) Elecsys (r²=0.99) COBAS® 4000/4e11 (r²= 0.99) Modular E170 (r²=0.99)} Siemens {Centaur XP/XPT/Classic (r²=0.99) Dade Dimension (r²=0.99) Immulite 1000 3rd gen (r²=0.99) Immulite 2000/2500, 3rd gen (r²=0.99)} Beckman {DXI standardised to Hybritech (r²=0.99) DXI standard to WHO IRP96/670 (r²=0.99). Access standard to WHO IRP96/670 (r²=0.99)} DiaSorin, Liaison (r²=0.99), there was a positive bias with some assays. The POCT i-CHROMA™ AFP also correlated well with: Abbot {Architect (r²=0.97)} Biomerieux {VIDAS (r²=0.97)} Roche {COBAS® 6000/8000 (r²=0.97), COBAS® 4000/e411 (r²=0.97), Modular E170 (r²=0.97) Elecsys (r²=0.97)} Siemens {Centaur XP/XPT Classic (r²=0.96) DPC Immunilite 2000/2500 (r²=0.97) DPC Immunilite 1000 (r²=0.97)} Beckman {DXI 600/800 (r²=0.97)} SNIBE Maglumi Analysers (r²=0.95) and Ortho Vitros 3600/5600/ECi (r²=0.97). There was bias of 43.87 to 57.99, indicating an overestimation with the i-CHROMA™ AFP method. In summary, both the POC i-CHROMA™ PSA and AFP methods correlated very well with the other traditional laboratory methods enrolled in the RIQAS, although there was a slight overestimation with both the POC i-CHROMA™ PSA and AFP methods compared to the others.

  • P124 An evaluation of the point of care test i-CHROMA™ prostate specific antigen method for screening in the community

    Suman Bains (JB Consulting MDP Limited, Upper Heyford, United Kingdom), Vivek Chand (Vekindo Limited, Greenford, United Kingdom), Roger Bacon (Prostate Cancer Support Organisation, Emsworth, United Kingdom), Peter Weir (Prostate Cancer Support Organisation, Emsworth, United Kingdom), Vivian Miles (Prostate Cancer Support Organisation, Emsworth, United Kingdom), John Bolodeoku (JB Consulting MDP Limited, Upper Heyford, United Kingdom), Frank Chinegwundoh (School of Health Sciences, City, University of London, London)

    We have previously, demonstrated that i-CHROMA™ POCT prostate specific antigen (PSA) method correlated well with a number of traditional laboratory methods. The aim of this study was to evaluate the performance of the i-CHROMA™ POCT method for PSA screening in the community. Blood samples [venous whole blood (143) and serum (143)] and capillary (55), were collected from 143 volunteers who attended a PSA screening campaign. Venous whole blood and capillary samples were analysed using the i-CHROMA™ PSA method, serum samples were analysed using the Abbott Architect PSA method. The results were compared using linear regression and a RAG (Red Amber Green) analysis where red indicated a raised PSA; amber, a slightly raised PSA and green, a normal PSA, this scoring system was based on the volunteer’s age and PSA level. The data showed that, the PSA results obtained using the i-CHROMA™ method on the venous whole blood samples and the capillary samples showed a good correlation (r2=0.9841 and r2=0.90845, respectively) with the serum samples tested with the laboratory method (Abbot Architect method). At lower PSA levels, (below 10.0 ng/mL), the i-CHROMA™ results were accurate. At higher levels (above 10.0 ng/mL) there was a positive bias particularly with the venous results. Using the RAG analysis: the i-CHROMA™ PSA method identified from the venous whole blood samples, 15 reds, 13 ambers 115 greens compared to 9 reds, 3 ambers and 126 greens identified from the serum samples by Abbot Architect Method. In summary, the i-CHROMA™ POCT PSA method showed good correlation with the Abbott Architect PSA, although there were a higher number of raised and abnormal PSA identified by the i-CHROMA™ POCT PSA method due to the positive bias observed with the method especially above 10 ng/mL. The i-CHROMA™ POCT PSA method is a reliable method for screening within its limitations.

  • P125 Impact of point of care management of critical care outreach patients using ePOC biochemistry and blood gas analysis

    Natalie Hunt (Lancashire Teaching Hospitals NHS Foundation Trust, Preston, United Kingdom), Robert Bolton (Lancashire Teaching Hospitals NHS Foundation Trust, Preston, United Kingdom), Denise Brooks (Lancashire Teaching Hospitals NHS Foundation Trust, Preston, United Kingdom), John Walmsley (Lancashire Teaching Hospitals NHS Foundation Trust, Preston, United Kingdom), Martin A Myers (Lancashire Teaching Hospitals NHS Foundation Trust, Preston, United Kingdom)

    Our Critical Care Outreach Team clinically assesses ward patients with new onset AKI or sepsis. If decision diagnostics are required blood samples are sent to the laboratory for analysis and the outreach team revisit the patient once results are available. To improve this process we explored the use of extended point of care testing at initial assessment.

    Method: Thirty patients attended by the critical care outreach team in September-October 2016 were included in this study. Venous samples were collected and analysed using ePOC BGEM test cards (Alere) and sent to the laboratory (Cobas, Roche) for parallel analysis. In addition, if patients required arterial blood gas analysis the sample was simultaneously analysed using ePOC and ABL90 (Radiometer) blood gas analyser. BGEM test cards include 11 analytes (pH, pO2, pCO2, pSO2, Hct, Na+, K+, Cl, Ca++ Urea, Creatinine, glucose, lactate) with results available in 30 seconds. Statistical analysis was carried out using Analyse It.

    Results: Good correlation was seen between ePOC and laboratory analysis for Na+, K+, Cl, Urea, Creatinine, glucose and lactate. Excellent correlation was seen between ePOC and ABL90 pH and pCO2. Good correlation between chloride ePOC and ABL90 was seen; ionised calcium and pO2 compared less well due to volatile nature of these analytes. Acceptable precision was obtained using control material across the analyte panel. Creatinine has a positive bias using ePOC compared with Cobas laboratory method (12 µmol/L across the measuring range).

    Conclusion: We found the ePOC to be analytically suitable for the Critical Care Outreach team and they thought that the ePOC was easy to use. Having the results at initial visit improved decision making with less patient contact time. A business case is being prepared to introduce ePOC routinely for these patient groups and to further assess patient outcomes.

  • P126 Verification of two CRP point-of-care devices for use in primary care

    Chris Maynard (Lancashire Teaching Hospitals NHS Foundation Trust, Preston, United Kingdom), Natalie Hunt (Lancashire Teaching Hospitals NHS Foundation Trust, Preston, United Kingdom), Robert Bolton (Lancashire Teaching Hospitals NHS Foundation Trust, Preston, United Kingdom), Martin A Myers (Lancashire Teaching Hospitals NHS Foundation Trust, Preston, United Kingdom)

    Introduction: Over half of antibiotics prescribed in primary care are for respiratory tract infections (RTIs) yet the majority of these RTIs are viral or self-limiting bacterial infections that do not require antibiotic treatment. Reducing inappropriate antibiotic use is necessary to limit the spread of anti-microbial resistance as well as to minimise unnecessary expenditure. Accordingly, NICE CG191 recommends point-of-care (POC) C-reactive protein (CRP) testing in patients in primary care presenting with symptoms consistent with lower RTIs, if, after clinical assessment, a diagnosis of pneumonia has not been made, and it is not clear whether or not antibiotics should be prescribed. The guidance recommends withholding antibiotics if CRP concentration is less than 20 mg/L; prescribing antibiotics if greater than 100 mg/L, and considering a delayed prescription if 20-100 mg/L. We have verified two CRP POCT devices against our laboratory method prior to their trial in primary care.

    Methods: The Alere Afinion and QuikRead go POCT devices were verified for CRP analysis and compared against the Roche Cobas 8000 701. Pooled serum samples from patients were prepared to cover the range of expected CRP values. Imprecision was assessed using manufacturer-supplied quality control materials.

    Results: Both devices were easy to use and performed well, with good agreement with the laboratory method at both decision cut-offs. The Afinion results were: Afinion = 1.09 x Cobas -2.38 (Passing & Bablok), r2=0.99 (Pearson correlation); imprecision: 2.56% (CV) at 19.5 mg/L, 6.48% (CV) at 56.6 mg/L. The QuikRead results were: QuikRead = 1.04 x Cobas -1.77 (Passing & Bablok), r2=0.99 (Pearson correlation); imprecision: 3.23% (CV) at 23.2 mg/L, 3.83% (CV) at 70.8 mg/L.

    Conclusion: Our verification has shown that both devices are suitably accurate for measurement of CRP at the NICE recommended decision cut-offs. Both devices are now being trialled in primary care.

  • P127 The integration of the rapid plasma separation device into the point of care i-CHROMA™ immunoassay system

    Stuart Pinkney (JB Consulting MDP Limited, Upper Heyford, United Kingdom), Suman Bains (JB Consulting MDP Limited, Upper Heyford, United Kingdom), Abhinav Goyal (Advanced Microdevices Pvt. Ltd, Ambala Cantt – 133006, India), Nalini Gupta (Advanced Microdevices Pvt. Ltd, Ambala Cantt – 133006, India), John Bolodeoku (JB Consulting MDP Limited, Upper Heyford, United Kingdom)

    It has been shown that plasma derived from Rapid Plasma Separation Device (RPSD) is comparable to plasma obtained by centrifugation for immunological (HBsAg, HCV) and biochemistry (uric acid, urea, glucose) applications, with minimal white cell and platelet contamination but no published data for hormones. The aim of this study was to evaluate the integration of the RPSD into a Point of Care i-CHROMA™ cortisol system comparing plasma separated using the RPSDs and plasma separated by centrifugation. Ten healthy volunteers had 10 mL of whole blood collected into lithium heparin tubes from each of them, this was separated into two sets of tubes (A and B). Tubes A were centrifuged and plasma collected (centrifuged plasma), tubes B, had its hematocrit estimated and plasma was extracted using the RPSDs using about 460 µL per device (RSPD plasma), and also for the whole blood cortisol analysis (50 µL). Thirty µL of plasma were used from the plasma samples (centrifuged plasma and RPSD plasma) and 50 µL for the whole blood samples, these were pipetted into the detection buffer and applied onto the cortisol cartridge and incubated for 10 minutes, then read on the i-CHROMA™ reader. The cortisol estimations (whole blood, centrifuged plasma and RPSD plasma) were analysed by correlation and Bland Altman difference plots. All cortisol estimations correlated very well: RPSD plasma cortisol vs centrifuged plasma cortisol (r2= 0.9497), RPSD plasma cortisol vs whole blood cortisol (r2=0.9146), centrifuged plasma cortisol vs whole blood cortisol (r2=0.9738). The Bland-Altman difference plots also highlighted that centrifuged plasma cortisol vs whole blood cortisol was the best of the comparisons and there was positive bias in estimates from the RPSD plasma cortisol. This evaluation demonstrated that plasma filtered using the RPSD is comparable to plasma obtained by centrifugation for immunoassays such as cortisol on the POC i-CHROMA™ system.

  • P128 Evaluation of two point of care tests for measuring CRP in whole blood

    Mfon Ewang-Emukowhate (Royal Free Hospital, London, United Kingdom), Felicity Blake (Royal Free Hospital, London, United Kingdom), Anjly Jain (Royal Free Hospital, London, United Kingdom), Devaki Nair (Royal Free Hospital, London, United Kingdom)

    Background: CRP is a biochemical marker used in the diagnosis of infection and in monitoring its treatment. Inappropriate antibiotic prescription contributes to the burden of antibiotic resistance. NICE has offered guidance on antibiotic prescription based on CRP concentration. A CRP point of care test (POCT) will be more convenient for patients and will ensure results are available quickly to help general practitioners (GP) make appropriate treatment decisions.

    Aim: To evaluate whole blood CRP measurement in Alere Afinion™ and Finecare™ POCT and compare it with the Roche Cobas® CRP method in serum.

    Method: A random search was performed on the Pathology IT system to identify patients who were having CRP measured in serum on the routine laboratory method the Roche Cobas®. Their corresponding EDTA sample was used to perform CRP POCT in whole blood. A total of 50 anonymised patient samples was used for this study. Precision studies were also carried out for each CRP POCT using quality control (QC) materials at two CRP levels provided by the manufacturers.

    Results: Very good correlation was seen between the Alere Afinion™ POCT and the Roche Cobas® method (r2=0.98). Both methods showed good agreement with a mean negative bias of 0.31 mg/L. A good correlation (r2=0.86), and agreement with a mean negative bias of 4.08 mg/L was observed between the Finecare™ POCT and the Roche Cobas® method.

    Conclusion: Both POCT showed better agreement with the laboratory method at CRP levels <50 mg/L. Different challenges were identified with the use of these CRP POCT. They include; multiple and manual preparation steps that can lead to errors, interference, issues with the quality control material, storage and customer care.

    Based on the NICE guidance and our findings we conclude that the two CRP POCT will be useful in determining which patients will not require antibiotics.

  • P129 Plasma conversion factors may not be appropriate when measuring point of care glucose in neonatal samples

    Ruth O'Kelly (Coombe Women and Infants University Hospital, Dublin, Ireland), Ann O'Donnell Pentony (Coombe Women and Infants University Hospital, Dublin, Ireland), Anne Killalea (Coombe Women and Infants University Hospital, Dublin, Ireland), James Kelly (Coombe Women and Infants University Hospital, Dublin, Ireland), Mary Stapleton (North Devon District Hospital, Devon, United Kingdom)

    To avoid unnecessary blood loss in neonates, point of care (POC) devices (such as the HemoCue 201+) are frequently used. However, discrepancies between ward POC and laboratory glucose results have been noted in the literature and may be a cause of lack of confidence in bedside technology. The IFCC has recommended the reporting of the concentration of glucose in plasma irrespective of sample type or measurement technique using a conversion factor. However this may not be appropriate in the neonatal population. We compared glucose measurements in neonatal and adult samples by a point of care device (without a plasma correction factor) and a laboratory method. Samples from 25 neonates (aged 3 days or less) received into the laboratory for routine glucose estimation were included. Similarly 34 samples from pregnant women were also analysed for whole blood and plasma glucose. (Inclusion criteria: adequate blood volume to perform both tests, glucose concentration ≤5 mmol/L). The mean whole blood glucose result for neonatal samples analysed on receipt using the POC device was 2.8 mmol/L (range 0.1-4.3). The mean plasma glucose result from the same samples immediately assayed on the laboratory analyser was 2.7 mmol/L (range 0.1-4.8). Bland Altman analysis showed these values were not significantly different (p=0.3446). The mean whole blood glucose in adult samples was 3.8 mmol/L (range 3.2-4.9) while the mean plasma glucose result was 4.2 mmol/L (range 3.8-5.0) [Significant at p=0.0001]. The difference between plasma and whole blood glucose was similar to the recommended IFCC “plasma equivalent factor” of 1.11. The lack of difference between plasma and whole blood glucose in neonatal samples may be explained by the increased mean corpuscular volume or the presence of resistant red cells that do not undergo full lysis in the POC device reaction.

  • P130 Do air-bubbles or transport method affect blood gas samples analysis?

    Mariana Ionescu (Cwm Taf University Health Board, Merthyr Tydfil, United Kingdom; Cardiff Metropolitan University, Cardiff, United Kingdom), Brian P Tennant (Cwm Taf University Health Board, Merthyr Tydfil, United Kingdom), Alan Dodd (Cwm Taf University Health Board, Merthyr Tydfil, United Kingdom), Jason Weaver (Cwm Taf University Health Board, Merthyr Tydfil, United Kingdom), Sally Thirkettle (Cwm Taf University Health Board, Merthyr Tydfil, United Kingdom), Mark Henry (Cwm Taf University Health Board, Merthyr Tydfil, United Kingdom), David Cassidy (Cwm Taf University Health Board, Merthyr Tydfil, United Kingdom), John Geen (Prince Charles Hospital, Merthyr Tydfil, United Kingdom; University of South Wales, Pontypridd, United Kingdom)

     

    We wished to establish if our local upgraded PTS would be suitable for BG samples without adversely affecting results.

    Aim: Establish if PTS transportation and/or air bubbles affect blood gas results.

    Methodology: Emergency Department (ED) BG samples (n=38) were transported to Clinical Biochemistry for repeat analysis, or vice versa, following transportation by PTS or walking, randomised by alternate allocation. BG results, time of analysis and number/size of any air bubbles present in ED/Lab; sample type; and transportation method, were recorded. Time and measured pO2 change (%), between repeat measures, and surface area/volume of air-bubbles were calculated. Gaussian and non-Gaussian data groups were compared by unpaired T test and Mann-Whitney U test, respectively. Correlations were made between pre-transport air-bubble surface area/volume and change in pO2 (%). A p-value <0.05 was deemed statistically significant.

    Results: Samples transported between ED and laboratory by PTS or walking, showed no significant difference in pO2 change (median 7.9 and 7.8%, respectively), or in time to repeat measurement (mean 11.5 and 14 minutes, respectively). Both air-bubble surface area and volume correlated significantly with change in pO2 (Spearman r=0.42 and 0.41, respectively). When broken down by transport method, only air-bubble surface area and change in pO2, for samples walked to the laboratory significantly correlated (Spearman r=0.50).

    Conclusion: These data suggest BG samples air-bubbles may increase pO2, in samples transported on foot. Previous studies have not measured or considered air-bubble surface area. We plan further sample analysis, to ensure appropriate sample number and examine other BG parameters.

  • P131 Every virtual ward needs a virtual laboratory

    Chris Cockcroft (Bradford Teaching Hospitals NHS Foundation Trust, Bradford, United Kingdom), Rafaq Azad (Bradford Teaching Hospitals NHS Foundation Trust, Bradford, United Kingdom)

    The Virtual Ward team at Bradford Teaching Hospitals provides community based care for elderly patients who have recently left hospital and require regular monitoring. The team currently use two i-STAT (Abbott) POCT devices for carrying out blood tests in the patient’s own home. We carried out laboratory-based validation of the i-STAT CHEM8+ cartridge system to verify analyte agreement with the existing laboratory methods for U&Es, chloride, bicarbonate, glucose and haematocrit. In consideration of clinical risk management, we set clinically agreed action limits in order to avoid missing abnormal results that would typically be telephoned by the laboratory. The timely confirmation of abnormal results outside the action limits by the laboratory was advised unless Medical Staff were already aware of similar recent laboratory results (e.g. elevated creatinine in CKD patients). Sample collection for confirmation testing took place either in the community or following urgent re-hospitalisation depending on the clinical history and presentation of the patient. Evaluation of i-STAT performance in the community setting was undertaken by comparing i-STAT U&E results with the laboratory results when re-tested within 6 hrs. Approximately 500 i-STAT tests were carried out on patients in the first 13 months of operation. The proportion of U&E results that triggered the action limits was 17%. The majority of the patients in the action limit group received a confirmatory laboratory test (77%) and were re-tested within 6 hrs (63%). Comparison of the initial U&E validation studies and community based patient studies indicated no significant change in accuracy. In summary, these findings support the validity of i-STAT results in the community setting.

  • P132 Use of A3 methodology to put the ‘monitor’ back into PKU monitoring

    Erin C Mozley (Viapath, London, United Kingdom), Sasala Wickramasinghe (Viapath, London, United Kingdom), Farzana Ghoni (Viapath, London, United Kingdom), Nicole Delange (Viapath, London, United Kingdom), Samantha Hayes (Viapath, London, United Kingdom), Rachel Carling (Viapath, London, United Kingdom)

    Patients with the inherited metabolic disease phenylketonuria (PKU) require routine monitoring of their blood phenylalanine concentration to optimise their ongoing treatment. Phenylalanine is measured in bloodspots by tandem mass spectrometry and the results are communicated to the Adult and Paediatric metabolic dietetic teams at the Evelina London as soon as they are available. The Dieticians require timely reporting of results to enable them to contact the patients with details of any required changes in treatment.

    A Continuous Quality Improvement project using A3 methodology was performed in Biochemical Sciences in collaboration with the metabolic clinical and dietetic teams. The problem statement was “The PKU monitoring turnaround time does not currently meet the needs of the Metabolic Dieticians who require the results to be available by 4 pm daily”.

    Multiple waste steps were identified in sample postage and receipt, processes within the laboratory, and result reporting. A number of changes were introduced to enable a faster and more efficient pathway from sample collection to reporting of results. A Key Performance Indicator (KPI) was subsequently introduced that states 90% of batches should be reported by 4 pm. This KPI has been consistently achieved over the last year.

    Ongoing work includes further raising awareness of why these results are so crucial to the management of patients with PKU. A session with the dieticians and a PKU patient representative is planned for next year, and included in our user engagement surveys this year is a survey for all monitoring patients, including an invite for them and their families to visit the laboratory.

  • P133 Faecal and urine reducing substances: a report of current practice

    Julie Davies (Blood Sciences, Wye Valley NHS Trust, Hereford, United Kingdom)

    Aim: To review of current practice for faecal and urine reducing substances.

    Methods: A pilot sample exchange scheme was begun in March 2016 with 10 participants recruited via the ACB mail base. A questionnaire was sent to all participants concerning current requesting and reporting practices for these two tests.

    Results: Urine reducing substances were reported in 100% of laboratories; faecal reducing substances in 60%. All laboratories only accepted samples for analysis if received on the same day of collection. Ninety percent of laboratories use Benedict’s reagent as their screening method, the other used HPLC. Of the 80% of laboratories with QC in place, the split between commercial and in house QC was 50:50. Reporting methods varied greatly between laboratories with 50% producing a semi-quantitative value and 30% a positive or negative result but only 4 laboratories had verified these cut-off values. Where this study had been done, both units and cut-off levels varied significantly between sites.

    Conclusion: Reporting practices varied greatly between laboratories and should be standardised. All should ensure users are aware of how to interpret results in light of the clinical context.

  • P134 Assessing the impact of electronic add-on requesting in North East Yorkshire

    Maria G Flenley (York Teaching Hospitals NHS Foundation Trust, York, United Kingdom), Emma Lovie (York Teaching Hospitals NHS Foundation Trust, York, United Kingdom), Rick Gillyon (York Teaching Hospitals NHS Foundation Trust, York, United Kingdom), Daniel Turnock (York Teaching Hospitals NHS Foundation Trust, York, United Kingdom)

    In our Trust, clinicians have historically requested additional tests on samples already delivered to the laboratory (add-ons) using either telephone communication or handwritten forms. This places a significant burden on laboratory staff, prolongs test turnaround times, and is subject to complications (e.g. illegible handwriting, missing forms). We set out to gather more information on the nature of add-on requesting and identify common requesting patterns (which could be targeted for quality improvement) by performing an audit at our two main laboratory sites in York and Scarborough. A dummy ‘ADD’ test was designed to allow add-on details (e.g. requestor name, role, location and test requirements) to be recorded in our laboratory information management system (LIMS) for each request received. A search of the LIMS identified requests received during 1 week in July 2016 for our audit. This was repeated for a re-audit in December 2016. The most striking trends in the July results were a large proportion of CRP requests originating from Emergency Departments (ED) and 40-60% of haematinics requests originating from GP surgeries. Prior to re-audit, we liaised with acute physicians to increase the inclusion of CRP in ED test profiles. We also introduced an electronic add-on request system to try and improve the efficiency and documentation of add-on requesting. Whilst this showed little impact on the absolute number of CRP add-ons received, the proportion of requests featuring CRP decreased, due to an 11% increase in the total number of add-ons received. The increased add-on volume may reflect either an increased ease of adding tests, or improved documentation of requests using the new electronic system.

    We conclude that introduction of the electronic add-on request system provides a useful tool for monitoring the burden of add-on requesting and assessing the impact of changes made by the laboratory, but may increase add-on requesting in the short-term.

  • P135 Audit of an internal quality assurance scheme for biomedical scientist staff reporting xanthochromia scans

    Emma J Lewis (Countess of Chester NHS Trust, Chester, United Kingdom)

    Scanning CSF for the presence of xanthochromia is used as part of the diagnostic workup for possible subarachnoid haemorrhage (SAH). The results rely on accurate analysis and interpretation of the resultant spectrophotometric scan. In order to ensure that biomedical scientist (BMS) staff are correctly reporting the scans we have introduced an internal quality assurance scheme for all staff who work within the biochemistry department, utilising the IMMQAS CSF xanthochromia external quality assurance (EQA) samples. This audit retrospectively looked at 2 years’ worth of data, recording the number of results returned, results obtained and how these related to the final EQA results.

    Eighty-five percent of staff returned results for all scans sent out with only 1 member of staff missing more than 1 return. Overall 94% of staff obtained correct results as compared to the consensus results from IMMQAS for all the scans sent out. There were 9/144 incorrect results, 6 of these were during the first year as people were getting used to the format of the scheme. During the second year of the trial there were only 3 minor errors.

    During the period of the audit it was not compulsory for staff to take part in the scheme, however the majority of staff were willing to take part and found that they were more confident about reporting xanthochromia scans as they had seen a range of scenarios that they may not see during routine scan reporting.

    We are looking to continue the scheme and use the results as part of on-going competency records and for training new members of staff.

  • P136 Clinical audit of performance of a new tandem mass spectrometry method for urine catecholamines

    Helen L Wiggins (Heart of England NHS Foundation Trust, Birmingham, United Kingdom), Briony Johnson (Heart of England NHS Foundation Trust, Birmingham, United Kingdom), Craig Webster (Heart of England NHS Foundation Trust, Birmingham, United Kingdom)

    Urinary catecholamines are a critical screening tool for phaeochromocytoma and paraganglioma, and are commonly used in combination with urine metanephrines to aid clinical decisions. Analysis of catecholamines at Heart of England NHS Foundation Trust changed from HPLC-ECD to tandem mass spectrometry in September 2015, largely necessitated by aging, unsupported equipment and the limitations of non-mass spectrometry based technologies. This audit was conducted to compare laboratory performance and clinical utility of the two methods pre and post method change.

    The new method was associated with a 37% decrease in rejected catecholamine samples, with the proportion rejected due to difficulty identifying peaks falling by 98%. This was the major cause of rejected samples in 2015 and the fall reflects increased analytical performance of the new assay. In comparison the majority of rejected samples in 2016 were caused by an unknown period of urine collection; however, this only represented 5% of requests.

    Further analysis established the difference in positive results produced by the two methods. The new method increased positive results by 35%, however the proportion of positive noradrenaline, adrenaline and dopamine results were unchanged. Clinical information was available for 15 patients in 2015 and 30 patients in 2016, although due to the rarity of the disease, there was only one example of a true positive (metastatic paraganglioma). Supplementary investigations for positive results included imaging, repeat catecholamine analysis and plasma metanephrines.

    The utility of catecholamine measurement in phaeochromocytoma was established in five histologically confirmed cases spanning two years. Catecholamines were raised in the most recent sample in all cases before surgery, and correlated with abnormal urine and plasma metanephrine results.

    In summary, the new catecholamine method improved the number of reportable and positive results, and was an important tool in the diagnosis of phaeochromocytoma.

  • P137 Identification through audit of a panacea to the age old problem of Emergency Department 1 hour turnaround time expectations

    Paul D O'Neill (Southern Health and Social Care Trust, Portadown, United Kingdom), Derek J McKillop (Southern Health and Social Care Trust, Portadown, United Kingdom), Sarah Duffy (Southern Health and Social Care Trust, Portadown, United Kingdom), Mark Feenan (Southern Health and Social Care Trust, Portadown, United Kingdom)

    Introduction: Emergency Departments (ED) frequently cite laboratory turnaround times (TAT) as a contributing factor in breaches of the ED 4 hour target. Improved TAT’s are perceived to reduce ED stay, improve ED and laboratory efficiency as well as enhancing patient safety and satisfaction.

    Aim: The aim of this project is to apply audit tools to the achievement of the Royal College of Pathologists’ (RCPath) key performance indicator (KPI) standard of reporting 90% of core blood tests within 1 hour (sample receipt to vet).

    Method: Baseline data from April 2016 and post intervention were extracted from Business Objects XI (BOXI). Four Plan Do Study Act (PDSA) cycles were developed to evaluate capacity and demand. Questionnaires were used to gauge service user and laboratory staff opinion on issues that it was felt impacted on TAT of blood samples. The following corrective actions were implemented: A red form was developed to aid staff in identification of ED samples; disease specific profiles were agreed with stakeholders with the objective of minimising ED stay; a staff member was assigned the role of runner between specimen reception and the laboratory; and use of a centrifuge was trialled in specimen reception.

    Results: At baseline, 41% of core ED blood tests were available within 60 minutes. The 90th percentile was 77 minutes. This improved to 91% and 36 minutes for the 90th percentile after implementation of the corrective actions. Laboratory testing was shown to represent 44% of the total patient journey in ED. The average transport time for ED samples to the laboratory using the pneumatic tube system was 26.9 minutes. Use of a dedicated ED runner evidenced a major improvement in turnaround times (18-24 minutes). Automatic release of results could save a further 11.4 minutes.

    Conclusion: The RCPath KPI standard of reporting 90% of core blood tests within one hour is achievable but will require investment in staffing and IT infrastructure.

  • P138 Application of a framework for the quality assessment of measurement procedures to the utility of neutrophil gelatinase-associated lipocalin in acute kidney injury

    Gemma L Jones (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Rebecca Kift (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), David Cairns (University of Leeds, Leeds, United Kingdom), Alison Smith (University of Leeds, Leeds, United Kingdom), Michelle Hutchinson (University of Leeds, Leeds, United Kingdom), Judy Wright (University of Leeds, Leeds, United Kingdom), Nicola Calder (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Catharine Sturgeon (Royal Infirmary of Edinburgh, Edinburgh, United Kingdom), Ashley Garner (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Elizabeth Mitchell (University of Leeds, Leeds, United Kingdom), Peter Hall (University of Edinburgh, Edinburgh, United Kingdom), Michael Messenger (University of Leeds, Leeds, United Kingdom)

    In vitro diagnostic (IVD) medical devices form the basis of ~70% of clinical decision making in the NHS. The accuracy and associated uncertainty surrounding diagnostic testing consequently has a major impact on the overall quality of clinical decisions and subsequent clinical and cost effectiveness. Numerous pre-analytical, analytical and biological factors can contribute to the measurement uncertainty in diagnostic testing procedures. These uncertainties accumulate through the measurement system and may introduce bias into clinical trials, contribute towards heterogeneity between biomarker research studies, and limit the applicability of research findings to clinical practice. Whilst prior reports have highlighted the scale and impact of these issues and reporting guidelines have been published (e.g. BRISQ, PROBE-ME), we are not aware of any methods in use for evaluating the quality and appropriateness of measurement procedures within systematic reviews of IVDs. We suggest that this may limit the ability of systematic reviewers and health technology assessors to fully evaluate risk and model uncertainty. This has been highlighted in several recent NICE diagnostic assessment programme reports.

    To address this issue we have identified key parameters for consideration by systematic reviewers and developed a framework for the quality assessment of measurement procedures using IVDs. Herein we present a case study applying this framework to articles assessing the utility of neutrophil gelatinase-associated lipocalin in acute kidney injury. We identify several measurement parameters that present a high risk of irreproducibility and inapplicability, and highlight several uncertain measurement issues which may present a risk of bias. We noted that reporting is often poor, particularly for the reference standard.

    Further work is still required to refine the parameters and signalling questions for inclusion within the framework, develop guidance for users, and validate its utility more widely.

  • P139 Comparison of total bile acid assays in serum to an improved ID-GCMS method for cholic acid, chenodeoxycholic acid and deoxycholic acid

    David H Ducroq (WEQAS, Cardiff, United Kingdom), Stephen Thompson (WEQAS, Cardiff, United Kingdom), Gareth Davies (WEQAS, Cardiff, United Kingdom), Mary A Thomas (WEQAS, Cardiff, United Kingdom)

    Total bile acids are routinely measured by non-specific enzymatic methods resulting in analytical differences between methods. The preferred comparison method of returned EQA results is to the SI unit utilising a reference target, ensuring the transfer of accuracy from gold standard methods to routine methods. An ID-GCMS method for bile acids previously developed has been used to compare returns for total bile acids within the WEQAS EQA programme. This method was found to underestimate the amount of conjugated bile acids in the base material and the method was reworked. Linear serum pools containing cholic acid and deoxycholic acid, reflecting levels observed in obstructive cholestasis distributed to participants were analysed using the modified method.

    Target values were assigned to the EQA material utilising the previously published ID-GCMS method, with the addition of a heat treatment stage to release conjugated bile acids in the serum. Deviations from the ID-GCMS result for main analytical groups were plotted in the form of bias plots (Bland–Altman plots).

    Under recovery of conjugated bile acids in the base material for the original ID-GCMS method was in the order of 4 µmol/L when compared to the modified method. Observation of the bias plots for all methods showed a slight shift in the profile compared to previous distributions as expected. All methods over recovered below concentrations of 10 µmol/L to a level of approximately 30%. The Randox Thio-NADH method under recovered at concentrations above this level, trending to 10% under recovery at higher concentrations. Both the Enz-Thio-NADH and Enz-Formazan methods showed a positive bias throughout the range with the former showing slightly better agreement with the ID-GCMS target value.

    Conclusion: Reference method targeted data is useful as an accuracy target in EQA Schemes. The improved ID-GCMS method will provide a more accurate comparison of routine bile acid methods.

  • P140 Where are my results?: using A3 methodology to reduce turn-around times and improve patient care

    Stamatina Agalou (Viapath, St Thomas' Hospital NHS Trust, London, United Kingdom), Monica Arenas-Hernandez (Viapath, St Thomas' Hospital NHS Trust, London, United Kingdom), Rachel Carling (Viapath, London, United Kingdom)

    The process for reporting a referral test was ambiguous resulting in delays to patient results, impacting negatively on multiple patient pathways.

    Data on the current state was collected over a period of a month. The pathway walk was complex and was separated into three stages: pre-analytical, analytical and post analytical. The fishbone and root cause analysis identified many types of waste such as transport and overproduction. In particular, the post analytical stage was contributing significantly to the delays, with the validation of results taking anything from <1 day to 97 days, median = 8 days, mean = 20. Therefore, this stage of the pathway became the focus for improvement and our goal was to reduce validation time from 8 days to 3 days for >90% of the results.

    To achieve this, the counter measure implemented was formalising the rota for validating referrals, clearly tasking a key group of individuals with the responsibility. Additional training and competency assessments were provided. The initial trial period lasted one month and 98% of results were validated within the target time. Two years later this improvement has been sustained demonstrating the success of the A3 project. On average referral tests are validated within 2.4 days (median = 2.0) and 71% of all tests are validated within 3 days. We continue to work towards our goal of 90% within 3 days.

  • P141 Serum separation and stability: an audit into the effect of barrier gels on therapeutic drugs monitoring

    Abigail J Clark (Sandwell and West Birmingham Hospitals NHS Trust, West Bromwich, United Kingdom), Nicola Barlow (Sandwell and West Birmingham Hospitals NHS Trust, West Bromwich, United Kingdom)

    The Pathology Department at Sandwell and West Birmingham Hospitals NHS Trust routinely analyses the blood samples of patients taking therapeutic drugs, in order to monitor the circulating concentration and ensure the prescribed dose keeps the patient within a therapeutic range where the drug will be effective. Current guidelines require therapeutic drug samples to be received and analysed in the lab within two hours of blood collection, due to published research that highlights the tendency of these therapeutic drugs to migrate from the patient serum into the barrier gel found within serum separator blood collection tubes (SSTs). SWBH Pathology Department receives samples from a large area within the West Midlands and, as a consequence, a large number of blood samples requiring therapeutic drugs monitoring (TDM) tests do not make it to the laboratory within this two hour window; a decision was made to investigate the appropriateness of this time limit. The majority of TDM samples were found to be stable on the gel up to 2-3 days, some for much longer. While carbamazepine and phenytoin showed the largest variation in percentage change away from the original test, when absolute concentration values were considered this would be unlikely to translate into clinically significant changes. Of the 81 samples analysed, only 30% arrived and were analysed within two hours; 91% of samples arrived and were analysed within 48 hours. The oldest sample was 4 days and 8 hours old, the freshest was just 19 minutes old. The data shows that through extending current TDM time limit to 24 hours, or even 48 hours, and we would be able to accept many more samples, and the tracking data suggests that this would not have a significant impact on the stability of therapeutic drugs in serum separator tubes.

  • P142 Urgent and critical results: auditing the communication process at Central Manchester University Hospitals NHS Foundation Trust

    Anna F Robson (Central Manchester University Hospitals NHS Foundation Trust, Manchester, United Kingdom), Chris Chaloner (Central Manchester University Hospitals NHS Foundation Trust, Manchester, United Kingdom), Fiona Ivison (Central Manchester University Hospitals NHS Foundation Trust, Manchester, United Kingdom), Anne Shenton (Central Manchester University Hospitals NHS Foundation Trust, Manchester, United Kingdom), Lesley Tetlow (Central Manchester University Hospitals NHS Foundation Trust, Manchester, United Kingdom)

    Appropriate communication of urgent and critical results is a crucial part of providing a clinical biochemistry service. Whilst failure to communicate urgent and critical results can have a direct impact on patient care, inappropriate communication of non-urgent results can overburden clinicians and potentially desensitise them to test results that do require urgent action.

    At Central Manchester University Hospitals NHS Foundation Trust (CMFT) we have audited the efficiency of laboratory result communication. This was carried out in preparation for revision of the escalation policy for communication of urgent and critical results within the Trust. Data was gathered on a selection of chemistry analytes (ammonia, phenytoin, theophylline, sodium, potassium, glucose and gentamicin) for all test requests made in March 2016 and current CMFT action limits were compared to the new RCPath recommendations for the communication of critical and unexpected pathology results (July 2016). Analysis of these data showed that 179 results required communication, of which 166 (92.7%) were telephoned to a clinician. Although some results that should have been communicated were not, an additional 70 results were communicated to clinicians that may not have required telephoning. Spurious potassium results accounted for 62 the 70 potentially inappropriate result communications. Of the results that were communicated 71.0% were telephoned within 2 hours and 87.9% within 4 hours. Data from the audit also showed that adopting the new recommendations from RCPath would impact most on sodium and potassium result communication. This audit has provided a useful benchmark from which to assess the anticipated quality improvement gained by updating the Trust escalation policy for communication of urgent and critical results.

  • P143 A four-way comparison between two blood tube types from two manufacturers for a range of common analytes

    Joseph M Taylor (Royal Liverpool University Hospital, Liverpool, United Kingdom), Andrew Davison (Liverpool Clinical Laboratories, Liverpool, United Kingdom)

    Background: In preparation for changing blood tube manufacturer we undertook a comparison study between different blood tube types and manufacturers. Comparison of new instrumentation or consumable to those in current use is important for quality control and to identify any significant differences before entry into routine use.

    Methods: Venous samples were collected from 11 healthy volunteers using four different serum or serum gel tubes from two manufacturers. Analyses of 27 analytes were performed using Roche COBAS 8000 series automated analysers; 7 were performed using LC-MS/MS and 4 ICP-MS. Statistical analysis was performed.

    Results: Comparing serum gel with plain serum tubes from manufacturer 1 there was a mean difference of 0.46% between (range -4.29% to 3.80%). There were statistically significant differences (p<0.05) between sodium, chloride, androstenedione and vitamin D3 results.

    Comparing serum gel with plain serum tubes from manufacturer 2 there was a mean difference of 0.17% (range -2.78% to 9.17%). There were statistically significant differences between potassium, bicarbonate, androstenedione and vitamin D3 results.

    Comparing plain serum tubes from manufacturer 1 against 2 there was a mean difference of 0.19% (range -2.54% to 7.14%). There was a statistically significant difference between androstenedione results.

    Comparing serum gel tubes from manufacturer 1 against 2 there was a mean difference of 0.41% (range -8.14% to 3.84%). There were statistically significant differences between sodium, potassium, bicarbonate, phosphate and androstenedione results.

    Conclusions: Statistically significant differences between some analytes measured in serum from different tubes were likely a combination of small sample size, limited range of results and normal assay imprecision. When comparing results it is important to assess if observed differences will impact on clinical decision-making. In this comparison no differences were deemed clinically significant. This comparison shows that results obtained for the analytes examined are comparable across the tube types and manufacturers tested.

  • P144 Towards a reference material for the standardization of calprotectin (S100A8/A9; MRP8/14) immunoassays

    Thomas Vogl (Westfaelische Wilhelms University, Muenster, Germany), Michael Schneider (BÜHLMANN Laboratories AG, Schoenenbuch, Switzerland), Tom Nilsen (Uppsala University, Uppsala, Sweden), Christian Niederberger (BÜHLMANN Laboratories AG, Schoenenbuch, Switzerland), Jakob Weber (BÜHLMANN Laboratories AG, Schoenenbuch, Switzerland)

     

    Conclusions: We were able to prepare highly purified human Cp reference material in reproducible quality and quantity. The effective Cp concentrations could be consistently determined by UV/VIS spectroscopy and confirmed by Biuret protein assay. To support determination of appropriate standardization of Cp immunoassays and, consequently, to measure accurate Cp concentrations, possible interferences of the biological matrix must be carefully considered and may need further investigation.

     

  • P145 Audit of pseudohyperkalaemia in GP samples

    Charlotte E Harborow (Liverpool Clinical Laboratories, Liverpool, United Kingdom), Lisa M Bailey (Liverpool Clinical Laboratories, Liverpool, United Kingdom)

    Serum potassium concentration (reference interval 3.5-5.3 mmol/L) is used to assess renal function, monitor hypo- or hyperkalaemia, and calculate the anion gap. Hyperkalaemia is a metabolic abnormality that can have life-threatening consequences and therefore requires urgent attention. Common causes of hyperkalaemia are renal failure and medications including ACE-inhibitors and potassium-sparing diuretics. Spuriously high potassium (pseudohyperkalaemia) can be due to haemolysis, K-EDTA contamination or delayed separation. Delayed separation is a problem associated with GP samples; incorrect storage and temperature fluctuation during transit may also contribute to pseudohyperkalaemia.

    Our aim was to assess our current policy for telephoning high potassium results on GP samples. We also estimated the frequency of pseudohyperkalaemia. Patients were selected retrospectively from GP potassium requests from 01/01/15 - 28/02/15 (n=80) and 01/09/15 – 31/10/15 (n=54) with potassium ≥6.2 mmol/L as per our telephone limit. Standards applied to the data examined documentation of collection times, whether these were recorded in the LIMS, whether all results ≥6.2 mmol/L were telephoned to the requester, and whether tests were repeated within 24 hours.

    Neither of the standards relating to documentation of collection times was met; it is important that these are recorded so judgement can be made on the likelihood of delayed separation. All potassium results ≥6.2 mmol/L were telephoned to GPs. A repeat test within 24 hours took place in 51% of patients, however, the decision to ask a patient to attend for a repeat is down to their clinician. Alternatively, the patient may not wish to attend for another test. Pseudohyperkalaemia was estimated at 53% in samples with a subsequent repeat within 24 hours.

    Analysis of all 2015 data showed more high potassium results from GPs in winter than in summer. For this reason, temperature during transit will be looked at in further detail as a probable cause of the high frequency of pseudohyperkalaemia. Implementation of temperature-controlled vehicles has been discussed and will be evaluated.

  • P146 Non-targeted metabolomic evaluation of patients with alkaptonuria compared to healthy subjects

    Andrew S Davison (Royal Liverpool and Broadgreen University Hospitals Trust, Liverpool, United Kingdom), Anna Milan (Liverpool Clinical Laboratories, Royal Liverpool and Broadgreen University Hospitals, Liverpool, United Kingdom), Andrew Hughes (Royal Liverpool and Broadgreen University Hospitals Trust, Liverpool, United Kingdom), Gordon Ross (Agilent Technologies, Cheshire, United Kingdom), James Gallagher (Ageing and Chronic Disease, Liverpool, United Kingdom), Lakshaminarayan Ranganath (Royal Liverpool and Broadgreen University Hospitals Trust, Liverpool, United Kingdom)

    Background: Alkaptonuria (AKU) is a rare autosomal recessive disorder of tyrosine metabolism. It results from a deficiency in homogentisate-1,2-dioxygenase. The consequence of which is accumulation of homogentisic acid (HGA) in the blood. HGA is central to the pathophysiology of the disease and the observed consequences.

    The aim of this work was to use a non-targeted approach to evaluate the metabolome of patients with AKU (n=16, mean age (±SD) 47.5±14.5 y, 9 male) compared to Healthy subjects (n=22, mean age (±SD) 43.2±13.5 y, 12 male).

    Methods: Urine samples were analysed using the Agilent Infinity 1290 liquid chromatography system coupled to a 6550 quadrupole time of flight mass spectrometer in positive and negative ionisation modes. Chromatographic separation was achieved following a 2 µL injection onto a Zorbax Eclipse Plus C18 column (2.1x100 mm, 1.8 µ). Mobile phase composition was (A) water and (B) methanol (both contained 0.1% formic acid and 5 mmol/L ammonium formate). Gradient started at 5% B, increasing to 100% by 12 minutes.

    Data were processed using Molecular Feature Extraction, which interrogates a 3D data space extracting chromatographic peaks representing compounds and the associated mass spectra. Mass Hunter Profinder software was then used to align and clean-up data. ‘Found’ entities were compared using Mass Profiler Professional, which identified up and down regulated entities.

    Results: Three hundred and fifty one compounds were aligned across all samples in positive ionisation mode. When data were filtered by sample variability (CV <25%) 14 and 9 compounds were confirmed to be up and down regulated respectively (two-fold, p<0.05). In negative ionisation mode 201 compounds were aligned across all samples; 16 and 6 compounds were confirmed to be up and down regulated, respectively (two-fold, p<0.05).

    Conclusions: Analysis of serum from AKU and healthy subjects shows a significant difference in abundance and presence of several entities. Further experiments are being undertaken to identify these compounds.

  • P147 Validation of a tandem mass spectrometry method to quantify tyrosine in phenylketonuria patient bloodspot samples

    Emma L Smith (Clinical Biochemistry, Southmead Hospital, Severn Pathology, Bristol, United Kingdom), Fiona Davidson (Southmead Hospital, Severn Pathology, Bristol, United Kingdom), Anny Brown (Southmead Hospital, Severn Pathology, Bristol, United Kingdom), Leila Cornes (Southmead Hospital, Severn Pathology, Bristol, United Kingdom)

    Background: Phenylketonuria (PKU) is an inborn error of metabolism in which phenylalanine hydroxylase deficiency leads to accumulation of phenylalanine and low concentrations of tyrosine in blood and tissues. Monitoring treatment for patients with PKU is currently undertaken in our laboratory by measuring phenylalanine in bloodspot samples. However, maintenance of adequate tyrosine levels is also critical. This study therefore aims to improve treatment monitoring by validating a tandem mass spectrometry method for measuring tyrosine concentrations in bloodspots.

    Methods: Tyrosine and phenylalanine were extracted from bloodspots in methanol and analysed using a Waters Acquity TQD tandem mass spectrometer in positive ion mode using flow injection analysis and multiple reaction monitoring. The method was validated for linearity, carryover, accuracy and precision. Bloodspot tyrosine concentrations were compared to paired plasma samples. Ninety-five percent reference ranges were determined from 120 normal bloodspot samples and concentrations compared with those measured in patients with PKU.

    Results: Measurement of tyrosine was linear up to 1146 μmol/L (r2=0.9958) with minimal carryover. The assay achieved a lower limit of quantification of 18.8 μmol/L. Intra-assay and inter-assay %CVs were <10% across the range 18.8-740 μmol/L. Accuracy was determined by analysing certified quality control samples and statistical analysis showed no significant differences between target and measured tyrosine (p=0.107). Bloodspot tyrosine correlated with plasma tyrosine (r2=0.8847). Reference ranges were calculated: bloodspot tyrosine 18-132 μmol/L, and phenylalanine:tyrosine ratio 0.4-3.1. Bloodspot tyrosine concentrations were lower in PKU patients than non-PKU patients, although not significantly (range of 17-140 μmol/L vs. 16-219 μmol/L; p=0.2469). Phenylalanine:tyrosine ratios were significantly higher in PKU patients (range of 1.6-29.2 vs. 0.2-3.6 in non-PKU patients; p=0.0004).

    Conclusions: These results confirm that tyrosine can be reliably and accurately measured at lower concentrations to aid in monitoring treatment in patients with PKU.

  • P148 Accelerating amino acid analysis the long way: improving an analytical pathway using A3 methodology

    Luisa Beltran (Viapath, London, United Kingdom), Jonathan Daw (Viapath, London, United Kingdom), Ali Lubulwa (Viapath, London, United Kingdom), Farzana Ghoni (Viapath, London, United Kingdom), Rachel Carling (Viapath, London, United Kingdom)

    Amino acid analysis turnaround times (TAT) and sample backlog suggested that the analytical pathway was not flowing smoothly, resulting in delays at multiple stages of the pathway and staff pressures. Ultimately this caused delays in the issue of patient reports and thus appropriate clinical management.

    Data collected during January – March 2016 showed that the mean TATs for plasma, CSF and urine amino acid analysis were 8.0, 7.3 and 12.3 days respectively. Whilst these figures are within the target TAT of 14 days, it was noted that just 43% of plasma samples were reported within 7 days. Furthermore, waste steps that caused avoidable delays in the analytical pathway were identified.

    Further data collection and root cause analysis identified 2 key areas for improvement. Phase 1 action plan, implemented in August 2016, introduced changes in day to day organisation of the section to standardise practice and improve the workflow with the aim of reducing avoidable delays. Phase 2 action plan, implemented in November 2016, involved specialist, manufacturer-led analytical training and a programme of cross-training throughout the section with the aim of reducing instrument downtime.

    Data collected during August – October 2016, following implementation of phase 1, showed that the mean TATs for plasma, CSF and urine amino acid analysis were 6.4, 6.8 and 9.9 days respectively. Importantly, the percentage of samples reported within 7 days had increased to 67% of plasma samples, reflecting a more consistent process. Furthermore, improvements in TAT were observed for every stage of the analytical process within the laboratory. These TAT reductions represent an improvement in working practice and ultimately deliver more timely results to a greater number of patients.

  • P149 Incidental finding of MATI/III deficiency from newborn screening

    Caroline Griffith (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Robert R Barski (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Timothy Williams (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Mick Henderson (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), John Walter (Central Manchester University Hospitals NHS Foundation Trust, Manchester, United Kingdom)

    Four additional metabolic conditions were introduced to the UK newborn screening programme in January 2015. This followed the success of a pilot implementation study of five conditions commencing in July 2012. One of the conditions included for screening was pyridoxine non-responsive classical homocystinuria (enzyme defect: cystathionine β-synthase). This is identified by measurement of methionine in dried bloodspots using tandem mass spectrometry in the first instance followed by a second tier test for bloodspot total homocysteine.

    Our laboratory reports an infant who was screened and had a markedly elevated methionine of 327 µmol/L (normal <50 µmol/L). Total homocysteine was 6 µmol/L (<15 µmol/L). Further investigation indicated that an older sibling is also affected. The sibling had a methionine of 320 µmol/L on screening.

    Part of the methylation pathway includes the conversion of methionine to S-adenosyl methionine. These two patients are homozygous for a previously unreported mutation in a highly conserved region of the methionine adenosyl transferase (MAT1A) gene c.895C>T.

    Currently aged 11 and 3, these siblings are apparently developmentally normal (with a normal MRI scan for the elder sibling) but have methionine concentrations >1000 µmol/L. Neither patient is on any specific treatment.

    A recent paper suggests that patients with methionine >800 µmol/L are more likely to have CNS abnormalities. Methionine restricted diets have been used in patients with this condition but there is no clear evidence of either benefit or lack of harm in this approach.

    These cases illustrate the potential of incidental identification of conditions of uncertain prognosis as a result of newborn screening and the difficulties that management of such patients may pose for the clinicians involved.

  • P150 A summary of newborn screening for MCADD in West Yorkshire: have we learnt anything in the last 7 years?

    Robert R Barski (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Caroline Griffith (St James's University Hospital, Leeds, United Kingdom), Daniel Herrera (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Tim Williams (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Janet Mitchell (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Ian Sherratt (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Elisabeth Jameson (Central Manchester University Hospitals NHS Foundation Trust, Manchester, United Kingdom), Mick Henderson (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom)

    Screening for MCADD became universal in England from April 2009. Between 2009 and 2016 50 patients from West Yorkshire have either had positive MCADD newborn screening (39) or been diagnosed with MCADD pre-screening (11) by our lab. Data from 46 of these patients is presented here. Two of the patients identified through newborn screening were subsequently found to be non-affected heterozygotes. The incidence of MCADD in this population was ~1 in 7000. Just under a quarter of diagnoses were made prior to newborn screening either because they presented clinically, or early testing was initiated as a result of previously diagnosed siblings. Four patients (9%) died early in the neonatal period before the screening samples were taken. Affected siblings of two of these patients were subsequently born and diagnosed pre-screening. Both siblings survived. On average C8 concentration was higher on screening/initial presentation (mean C8 = 5.81 µmol/L (range: 0.54-51.0)) than on follow up samples (mean C8 = 1.94 µmol/L (range: 0.17-10.7)). In contrast C8:C10 ratio remained much more consistent (mean C8:C10 at presentation = 7.5 (0.7-15.5), follow up = 7.2 (0.9-13.2)). As expected the most frequently observed mutation was the common p.(Lys329Glu), c.985A>G mutation. A number of patients were identified with mutations either not known to be pathogenic or that have only been seen in screening populations. Such patients often produce a clinical dilemma as it is not clear whether they have clinically relevant disease. In our cohort patients with these mutations had lower C8 and C8/C10 ratios compared with patients who had known pathogenic mutations. Follow up of patients with urinary organic acid analysis is useful in confirmation of diagnosis and detection of acyl glycines also appears to discriminate between heterozygote and homozygote status.

  • P151 Evaluation of the Mitra microsampling device for collection of patient samples in alkaptonuria

    Joseph M Taylor (Royal Liverpool University Hospital, Liverpool, United Kingdom), Anna M Milan (Liverpool Clinical Laboratories, Royal Liverpool and Broadgreen University Hospitals, Liverpool, United Kingdom), Andrew Davison (Liverpool Clinical Laboratories, Liverpool, United Kingdom), Andrew T Hughes (Liverpool Clinical Laboratories, Liverpool, United Kingdom), James Rudge (Neoteryx Europe, RD Geleen, Netherlands), Lakshminarayan Ranganath (Liverpool Clinical Laboratories, Liverpool, United Kingdom)

    Background: Alkaptonuria (AKU) is a rare disorder of tyrosine metabolism due to deficiency of homogentisate dioxygenase causing accumulation of homogentisic acid (HGA), arthropathy and multi-organ dysfunction. As a reference centre for AKU, Liverpool receives samples globally. This study evaluated the Mitra Microsampling Device (Neoteryx, CA) for AKU metabolites in urine and whole blood.

    Methods: Urine analysis was performed using a previously validated LC-MS/MS method. Whole blood LC-MS/MS measurement of tyrosine (TYR), phenylalanine (PHE), HGA and nitisinone (NTBC) was validated against published performance criteria. Sampling accuracy, imprecision and drying time were assessed gravimetrically. For recovery, multiple elution protocols were tested. Analyte stability was assessed for 28 days at 20°C (light and dark), 4°C and 37°C (light protected). Patient comparisons were performed using AKU patients and healthy volunteers.

    Results: Mitra displayed excellent sampling accuracy (mean 103%, n=10) and imprecision (4.1%, n=10) with urine. Mean elution recoveries (15 mins agitation, n=6) were 104%, 101% and 99% for TYR, PHE and HGA respectively. Imprecision was 4.9% (n=18).

    Urine PHE was stable for 28 days under all conditions. Urine TYR was stable at 4°C for 28 days, 21 days at 20°C and 1 day at 37°C. Urine HGA rapidly degraded except at 4°C. Acidification did not improve overall stability.

    Patient comparisons showed a concentration dependent positive bias (approximately 15%) across all analytes, minimised by using eluted calibrators.

    Whole blood assay performance was acceptable against published criteria. Further whole blood evaluation is ongoing.

    Conclusion: Mitra may be suitable for use in AKU and potentially phenylketonuria and hereditary tyrosinaemia. Though naturally unstable, HGA stability improved with refrigeration and using Mitra at 4°C surpasses that seen in native specimens.

    Microsampling offers potential benefits including improved patient experience, simple sample collection and transport and cost reduction. As such, device merit should not solely be based on technical performance.

  • P152 Verification of the DiaSorin Liaison Oestradiol II Gen assay for use in paediatrics

    Amy E Dunne (Birmingham Children's Hospital, Birmingham, United Kingdom)

    Background: Following a decrease in the reported sensitivity of the Roche assay used to measure oestradiol in samples referred from Birmingham Children’s Hospital (BCH), we investigated the possibility of bringing the assay in-house, utilising a DiaSorin Liaison analyser. The Liaison Oestradiol II Gen assay quotes a lower limit of quantitation (LLOQ) of 16.2 pg/mL, equating to approximately 60 pmol/L. This is lower than the quoted Roche LLOQ of 100 pmol/L, which is unsuitably high for a paediatric population. Currently within the UK, no laboratories are using the DiaSorin Liaison Oestradiol II Gen assay.

    Objective: To assess the performance of the Liaison Oestradiol II Gen assay for use in paediatrics (limited by the number of available tests and small sample volumes available).

    Materials and Methods: QC material and patient pooled samples were used to assess the precision, the quoted LLOQ, and to perform recovery experiments. EQA samples were tested to assess the performance of the assay against other instruments and methods.

    Results: Precision was shown to be good at concentrations assessed above 160 pmol/L (CV of 6.9% at mean 167 pmol/L; CV of 4.1% at mean 449 pmol/L). The precision at the stated LLOQ was however greater than the 20% level required (CV of 33.6% at mean 62.7 pmol/L). Recovery experiments demonstrated the recovery to be within ±15% for concentrations <217 pmol/L, but became >20% at a higher concentration. Comparison of EQA samples against the All Laboratory Trimmed Mean quoted by NEQAS demonstrated a positive bias of the Liaison of almost 30%.

    Conclusion: This assay was considered not to be suitable for use in assessment of oestradiol concentration in paediatric populations as the LLOQ of 60 pmol/L could not be verified. It may however be suitable for use in adults where a higher limit of sensitivity may be acceptable.

  • P154 A case report of late onset cystinosis: diagnostic challenges and pregnancy management

    Daniel Juan Herrera (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Emma Ferriman (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Albert Ong (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Janet Mitchell (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Ian Sherratt (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Mampreet Thandi (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Helena Russon (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Alex Broderick (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Claire Smith (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Tracy Charlesworth (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Mick Henderson (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Godfrey Gillet (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom)

    Cystinosis is an autosomal recessive disorder characterized by intra-lysosomal accumulation of cystine as a result of a defective cystine carrier protein (cystinosin) located in the lysosomal membrane. Three clinical forms of the disease have been described in the literature based on the severity of the symptoms and age of onset: infantile cystinosis with natural progression to end stage renal failure in the first decade of life; a juvenile form characterized by a slower rate of progression and an adult form with ocular abnormalities and no renal involvement. A 19 year old woman was referred to the ophthalmologist by her local optician after identifying refractive corneal crystals. The patient had history of headaches for several years and over the previous year she has developed an aversion for bright lights with increasing conjunctiva stinging and watering. Past medical history was unremarkable apart from nocturia and significant diurnal polyuria on-going for many years. Clinical and biochemical examination was unexceptional apart from proteinuria (urine protein 1.31 g/L, 117 mg/mmol creatinine). The diagnosis was confirmed by ophthalmological review through demonstration of cystine corneal crystals by slit lamp examination, biochemical confirmation of elevated cystine concentration in peripheral blood leukocytes (leucocyte cystine 3.23 nmol ½ cystine/mg protein, ref:≤0.1) and mutation analysis of the CTNS gene (c.416C>T:c.971-12G>A). A renal biopsy showed a mildly scarred kidney with patchy fibrosis and on electron microscopy cystine crystals were visible within tubule epithelial cells and macrophages. The patient has been treated with topical and oral mercaptamine that was stopped during her pregnancy. Unfortunately, she developed pre-eclampsia and miscarried at 27 weeks of gestation. Cystine crystals were not apparent on histology of the placenta.

  • P155 Advancing newborn blood spot sickle cell screening: validation of the SpOtOn diagnostics kit assay on a Waters TQS Micro

    Anna F Robson (Central Manchester University Hospitals NHS Foundation Trust, Manchester, United Kingdom), Heather A Brown (Waters Corporation, Wilmslow, United Kingdom), R Neil Dalton (Guy's & St Thomas' NHS Foundation Trust, London, United Kingdom), Charles Turner (Guy's & St Thomas' NHS Foundation Trust, London, United Kingdom), Beverly Hird (Central Manchester University Hospitals NHS Foundation Trust, Manchester, United Kingdom), Lesley Tetlow (Central Manchester University Hospitals NHS Foundation Trust, Manchester, United Kingdom)

    Newborn blood spot screening for sickle cell disease (SCD) has traditionally been carried out using two different separation techniques: high performance liquid chromatography (HPLC) and isoelectric focusing (IEF). HPLC and IEF use different principles for separation and are able to accurately identify the most common haemoglobin variants when used in combination as first and second line methodologies. Mass spectrometry (MS) is an emerging tool in the clinical laboratory that has demonstrated its utility in a number of different disciplines including newborn screening for metabolic disorders. A commercial kit is now available (SpOtOn diagnostics) for targeted screening of SCD and other haemoglobinopathies using MS. This method works by targeted detection of specific peptide fragments, diagnostic for different haemoglobin variants, from a tryptic digestion of the haemoglobin molecule. Different haemoglobin variants are then identified by calculating the ratio of variant peptide fragments to fragments from normal adult haemoglobin.

    The SpOtOn MS technique for haemoglobinopathy screening is currently in routine use in in three of the 16 UK screening laboratories, all using different MS instrumentation. Here we present a preliminary validation of the SpOtOn method using a Waters TQS Micro. One hundred and seventy blood spots were selected, containing a range of different haemoglobinopathies and transfusion states, and digested samples were analysed on two different MS platforms: Waters TQS Micro and Sciex API5000. Haemoglobin variants were correctly identified in all samples by both MS instruments, with the exception of one sample previously reported as containing haemoglobin variant E (HbE). Absence of the HbE peptide in this sample demonstrates the limitations in HPLC and IEF and the advantages of moving to a targeted technique such as MS. Data from this study show that the TQS Micro is a viable MS platform for haemoglobinopathy screening in dried blood spot samples.

  • P156 Urinary galactitol quantitation by gas chromatography mass spectrometry for the diagnosis and monitoring of galactosaemia

    Yuh Luan Choo (University of Manchester, Manchester, United Kingdom), Hoi Yee Wu (Willink Biochemical Genetics Laboratory, Manchester, United Kingdom), Jacqueline Till (Willink Biochemical Genetics Laboratory, Manchester, United Kingdom), Mick Henderson (Willink Biochemical Genetics Laboratory, Manchester, United Kingdom)

    Introduction: Galactosaemia is an inborn error of metabolism caused by the deficiency of galactose-1-phosphate uridyltransferase, galactokinase or galactose-4-epimerase. Urinary galactitol has been proposed as the first line test to aid the diagnosis of galactosaemia in transfused patients, and to monitor dietary compliance in affected patients.

    Aims: This project aims to adapt and validate a gas chromatography mass spectrometry (GCMS) method for urinary galactitol quantitation in the Willink Biochemical Genetic Laboratory, evaluate the key analytical validation components, establish the age-related reference ranges, and to study the relationship between urinary galactitol excretion and hepatic dysfunctions.

    Methods: Acetate derivatives were prepared and perseitol was used as the internal standard for GCMS. Plain urine samples from known galactosaemic patients, non-galactosaemic patients with suspected hepatic dysfunctions and normal individuals were randomly selected from samples sent for the urinary metabolic screen in the laboratory.

    Results: Our method showed good agreement with established methods by Bristol laboratory and ERNDIM participants. A small positive bias was observed in our method but not statistically significant. Intra- and inter-assay precisions were 1.41-6.22% and 2.54-31.51%, respectively. The method was linear from 2.5 µmol/L to 330 µmol/L. The LoD and LoQ were 3 µmol/L and 9 µmol/L. Urinary galactitol were stable up to 7 days under storage at -20°C, 4°C and room temperature. The age-related reference ranges were ≤85, ≤68, ≤29, ≤23, ≤9 and ≤4 µmol/mmol creatinine for the 0-3 months, 4-11 months, 1-2 years, 3-6 years, 7-15 years and >15 years age groups. Galactosaemic patients excreted 9-fold to ≥800-fold while non-galactosaemic patients with suspected hepatic dysfunctions excreted 3-fold more urinary galactitol than the age-matched control group.

    Conclusion: This study has compiled a set of preliminary data to show that our method is acceptable for the diagnosis and monitoring of galactosaemia. The data is promising but further work is necessary to complete the method validation before introducing this assay into routine use.

  • P157 Pilot study for screening Anderson-Fabry disease in hospital patients

    Monika D Kohli (Royal Free Hospital, London, United Kingdom), Devaki Nair (Royal Free Hospital, London, United Kingdom), Derralyn Hughes (Royal Free Hospital, London, United Kingdom)

    Introduction: Anderson-Fabry disease is an X-linked glycosphingolipid disorder caused by deficiency of lysosomal enzyme alpha-galactosidase A which result in accumulation of globotriaosylceramide in the body causing multi-systemic pathology. Enzyme replacement therapy has shown significant clinical benefits. Definitive diagnosis is made by determining alpha-galactosidase A activity in plasma and leukocytes and mutations in the α-Gal A gene. This was a pilot study to provide an idea about the number of hospital patients to screen and to assess the feasibility of screening for Anderson-Fabry disease using 96 well alpha-galactosidase A activity assay.

    Method: Discarded full blood count EDTA samples from 383 Royal Free Hospital patients were included in the study. Plasma was used to measure α-Gal A enzyme activity by a 96 well plate fluorimetric assay and the cells were frozen for DNA analysis. DNA was extracted from the cells of patients with deficient enzyme activity. DNA was analysed by high resolution melt curve analysis to screen for potential mutations. Sequencing was carried out to confirm mutations. The 96 well α-Gal A assay was evaluated by performing replication experiment and by comparison of results with the well-established 24 well assay.

    Results: All positive controls from known Anderson-Fabry patients showed α-Gal A enzyme activities <0.9 nmol/h/mL by the 96 well assay and the 96 well assay identified all diseased patients correctly. Both intra-assay precision and inter-assay precision were excellent (1.16% and 3.8%, respectively). The reference range for 96 well enzyme activity was derived to be 3.2 to 26.4 nmol/h/mL. Three patients with reduced α-Gal A enzyme activity had non-pathogenic -10C>T and c.640-16A>G intronic mutations. The prevalence of Anderson-Fabry disease was 0% in our study group. The corresponding 95% CI was 0 to 0.96% (calculated using the exact binomial method).

    Conclusion: The 96 well α-Gal A enzyme activity assay can be used to screen hospital patients. Compared with the 24 well plate assay it is simpler, faster, less expensive and more amenable to screening as it uses 80% less sample.

  • P158 Variation in paediatric creatinine practice across Scotland

    Chris Stockdale (Queen Elizabeth University Hospital, Glasgow, United Kingdom), Jane McNeilly (Queen Elizabeth University Hospital, Glasgow, United Kingdom), Peter Galloway (Queen Elizabeth University Hospital, Glasgow, United Kingdom)

    Appropriate plasma creatinine reference ranges are of particular importance in children as paediatric equations for eGFR are not widely used and physiological increases in plasma creatinine accompany growth. Differences in analytical bias, particularly between enzymatic and Jaffe methods, present additional complications for this analyte: these may prevent harmonisation of reference ranges and also complicate monitoring of a patient’s renal function across different laboratories. In Scotland a managed clinical network for paediatric nephrology means that patients may have creatinine measured both locally and by the Health Board hosting the network.

    The aim of this work was to assess paediatric creatinine practice across Scotland, with emphasis on reference ranges and analytical bias, by means of a questionnaire. Replies were received from 11 laboratories, representing 7 Health Boards which cover 8% of Scotland’s population.

    The survey results indicate a roughly equal split between enzymatic or Jaffe methods. The spread of results observed for UKNEQAS samples is wider if all methods are considered together rather than split into enzymatic or Jaffe sub-groups. Fourteen µmol/L is the widest spread seen for a human serum sample when both method types are considered. The survey also revealed variety in upper reference limits of up to 46 µmol/L: beyond what appears justifiable from the spread of EQA results. In addition there is inconsistency in the use of gender-specific ranges, which are quoted by 5 of 11 surveyed laboratories.

    The results of the questionnaire suggest that laboratories should review, and harmonise where possible, their paediatric creatinine reference ranges. Adoption of the ranges endorsed by BAPN (British Association of Paediatric Nephrology) and PaLMNet (Paediatric Laboratory Medicine Network) for enzymatic methods may help to improve identification of renal impairment in children. Adopting enzymatic creatinine for all paediatric samples may help to improve comparability of results produced by different laboratories.

  • P159 Incorporation of emergency department and paediatric samples into automation line to improve workflow and turnaround time

    Firdaus B Mohd Abu Bucker (National University Health System, Singapore, Republic of Singapore), Sharon Saw (National University Health System, Singapore, Republic of Singapore), Sunil Sethi (National University Health System, Singapore, Republic of Singapore)

    Before the changeover, all emergency department (EMD) and paediatric samples were part of a cell workflow, where the samples were spun on an external centrifuge and thereafter run on the standalone analyser, Vitros 5600 (Ortho Clinical Diagnosis); a wholly manual process. To improve the turnaround time and to streamline the process into a single automated workflow, incorporation of all EMD and paediatric samples into the automation system (Beckman Coulter) was done. Incorporation of samples into automation would ensure that there is improved safety and performance, eliminate human error and reduce manpower.

    Troponin and renal panel tests were selected as the tests to examine to see whether incorporation of the EMD and paediatric samples into the automation system was able to improve the turnaround time (TAT). The turnaround time for the results for the samples received was analysed for a period of about one month before and after the changeover.

    The EMD and urgent troponin TAT has not improved significantly. However the mean and median for urgent troponin have slightly dropped. The mean and median for both EMD renal panel and urgent renal panel have risen and the TAT for EMD renal panel has dropped post changeover.

    The drop in TAT could be attributed to the poor quality of the barcode labels that EMD samples were labelled with but this has since been corrected. EMD samples are usually hard to obtain from patients and thus a good number of them will have low volume of sample that requires manual intervention and front-loading into the analysers, raising the TAT.

    The turnaround time will continue to be monitored on a weekly basis, as the findings are not significantly conclusive due to the prematurity of the data being collected.

  • P160 Cascade testing of primary care blood samples with elevated serum ferritin identifies subjects with iron overload

    Robert Desborough (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Mary Moulding (Leeds Teaching Hospitals Trust, Leeds, United Kingdom), Tim Degg (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Emma Baddams (Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom), Julian Barth (Leeds Yeaching Hospitals Trust, Leeds, United Kingdom)

    Introduction: Serum ferritin is a common request from Primary Care for investigation of iron status. A raised ferritin is not an unusual finding but is relatively non-specific and can typically be found in patients with iron overload, liver disease, malignancy and inflammation. This study sought to determine whether those patients from primary care with high ferritins had either hereditary haemochromatosis or porphyria cutanea tarda (PCT).

    Methods: Consecutive, redundant serum samples identified with high ferritin (>500 µg/L) received from primary care were analysed for serum iron, iron-binding capacity and for porphyrins by fluorescence scanning.

    Results: There were a total of 284 samples (102 females, 182 males) collected over a 12 week period. Serum iron was 18.4 (13.3) µmol/L (median ([QR]), TIBC 52.2 (13.0) µmol/L and transferrin saturation 34.6 (22.7)%. There were 82/284 (29%) with transferrin saturations ≥45% (50 males, 32 females). Of the samples 3/284 (1%) were positive for porphyrins by spectrofluorimetry and 1/3 had total porphyrins >11.2 nmol/L. There was insufficient sample volume to assess further by HPLC.

    Discussion: This study confirms a previous study carried out in this laboratory. It demonstrates the feasibility of cascading tests using laboratory protocols and identifying potential cases of iron overload. A further study is planned to assess for HH genotype and urine and stool porphyrin analysis to confirm the diagnosis in these patients.

  • P161 Urine metabolomics using liquid chromatography quadrupole time-of-flight mass spectrometry indicates common markers of disease in alkaptonuria and idiopathic osteoarthritis

    Brendan Norman (IACD, University of Liverpool, Liverpool, United Kingdom), Andrew Davison (Royal Liverpool & Broadgreen University Hospitals Trust, Liverpool, United Kingdom), Peter Wilson (IACD, University of Liverpool, Liverpool, United Kingdom), Gordon Ross (Agilent Technologies, Cheshire, United Kingdom), Anna M Milan (Liverpool Clinical Laboratories, Royal Liverpool and Broadgreen University Hospitals, Liverpool, United Kingdom), Norman Roberts (Royal Liverpool & Broadgreen University Hospitals Trust, Liverpool, United Kingdom), Lakshaminarayan Ranganath (Royal Liverpool and Broadgreen University Hospitals Trust, Liverpool, United Kingdom), James Gallagher (IACD, University of Liverpool, Liverpool, United Kingdom)

    Osteoarthritis (OA) is associated with destruction of cartilage and is widespread in alkaptonuria (AKU), a serious genetioarthritic disease characterised by increased levels of homogentisic acid (2,5-dihydroxyphenylacetic acid; HGA) and resultant pigmentation, particularly of cartilagenous tissue. AKU is a severe, ultra-rare osteoarthropathy but there is considerable interest in using it as a model system to study pathological changes in common idiopathic OA (iOA). In support of this, some of the striking pathological features characteristic of AKU-associated cartilage destruction are observed at lower frequency in iOA. These include thinning of the subchondral plate, loss of collagen fibril integrity, high-density mineralised protrusions and aberrant bone remodelling. It was predicted that shared underlying biochemical processes underlie these common features of disease, and that these are detectable in urine.

    A non-targeted metabolomic profiling experiment using liquid chromatography quadrupole Time-of-Flight mass spectrometry (LC-QTOF-MS) investigated this hypothesis using urine from human AKU (n=15), iOA (n=7) patients and age-matched healthy controls (n=30). A total of 109 and 275 and metabolic entities (molecular weight <1700 Da) passed quality-control filtering and were aligned across all sample groups in positive and negative ESI respectively. Interrogation of the variance across all datasets using principal components analysis (PCA) showed clear separation between the urine metabolic profiles of AKU, iOA and controls. However, a number of entities were identified as up or down regulated in both AKU and iOA compared with controls indicating common mechanisms of disease in these two conditions. These entities span a relatively broad range in mass (100-800 Da) and retention time, suggesting diverse chemical identities. Putative identities for these entities can be derived based on accurate mass and retention time using web-based databases (e.g. METLIN), supporting further investigation using a more-targeted metabolomics approach. This study demonstrates the utility of non-targeted metabolomic profiling using LC-MS for biomarker discovery in diseases such as OA. Findings support further studies investigating the chemical identity of these potential biomarkers using approaches such as targeted MS/MS to yield valuable structural information.

  • P162 Profiling key disease markers in the genetic disease alkaptonuria using liquid chromatography quadrupole time-of-flight mass spectrometry

    Brendan Norman (IACD, University of Liverpool, Liverpool, United Kingdom), Andrew Davison (Royal Liverpool & Broadgreen University Hospitals Trust, Liverpool, United Kingdom), Peter Wilson (IACD, University of Liverpool, Liverpool, United Kingdom), Gordon Ross (Agilent Technologies, Cheshire, United Kingdom), Anna M Milan (Liverpool Clinical Laboratories, Royal Liverpool and Broadgreen University Hospitals, Liverpool, United Kingdom), Norman Roberts (Royal Liverpool & Broadgreen University Hospitals Trust, Liverpool, United Kingdom), Lakshaminarayan Ranganath (Royal Liverpool and Broadgreen University Hospitals Trust, Liverpool, United Kingdom), James Gallagher (IACD, University of Liverpool, Liverpool, United Kingdom)

    Alkaptonuria (AKU) is a debilitating genetic disease characterised by increased circulating homogentisic acid (2,5-dihydroxyphenylacetic acid; HGA) causing severe early-onset osteoarthropathy. Quantitative assays using tandem liquid chromatography mass spectrometry are successfully employed for monitoring HGA and other relevant metabolites of the tyrosine catabolic pathway in serum and urine in AKU with sensitivity and specificity. This study investigated the potential of liquid chromatography quadrupole Time-of-Flight mass spectrometry (LC-QTOF-MS) using reverse phase high performance liquid chromatography for metabolomics analysis of urine from human AKU patients. Metabolomics aims to profile low molecular weight compounds in a non-targeted approach to explore new biochemical markers for diagnosis, monitoring and understanding of disease.

    Using aqueous HGA and tyrosine standard solutions, semi-quantitative data was obtained with linearity at 0.1-10 µM. As this method relies upon ionisation of analytes through electrospray ionisation (ESI), we examined potential ‘ion suppression’ effects caused by biological matrix; human urine. The signal for reference ions, delivered isocratically from a separate source throughout the analytical run for accurate mass calibration, was monitored following injection of matrix-free ‘blanks’ (water or methanol) compared with urine diluted 1 in 4 with water. Compared with blank injections, urine did not cause excessive suppression of reference ion signal and the degree of suppression did not compromise accurate mass calibration.

    This study shows the potential of this LC-QTOF-MS method for semi-quantitative profiling of new and established disease-related compounds in AKU. Signal at 0.1µM, the lowest concentration tested, demonstrates the potential of this method for detecting low abundance reliable markers of disease. The dynamic range demonstrated combined with analysis of matrix effects on ESI has important implications for planning urine metabolomics experiments; a 1 in 4 dilution may be sufficient to avoid some detrimental ion suppression, but further dilution may be required to retain quantitative linearity and accurate mass calculation of high-abundance analytes.

  • P163 Development of an LC-QTOF method for measurement of catecholamines and related metabolites in alkaptonuria patients treated with nitisinone

    Andrew S Davison (Royal Liverpool and Broadgreen University Hospitals Trust, Liverpool, United Kingdom), Anna Milan (Liverpool Clinical Laboratories, Royal Liverpool and Broadgreen University Hospitals, Liverpool, United Kingdom), Andrew Hughes (Royal Liverpool and Broadgreen University Hospitals Trust, Liverpool, United Kingdom), Gordon Ross (Agilent Technologies, Cheshire, United Kingdom), James Gallagher (Ageing and Chronic Disease, Liverpool, United Kingdom), Lakshaminarayan Ranganath (Royal Liverpool and Broadgreen University Hospitals Trust, Liverpool, United Kingdom)

    Background: Alkaptonuria (AKU) is a rare autosomal recessive disorder of tyrosine metabolism resulting from a deficiency in homogentisate-1,2-dioxygenase. Nitisinone is being evaluated off licence for the treatment of the disease. One of the consequences of this treatment is that circulating tyrosine increases 10-12 fold. This is also a documented consequence of nitisinone treatment in hereditary tyrosinaemia-1. It is proposed that tyrosine may alter catecholamine metabolism and lead to cognitive impairment. In order to study the impact of tyrosine on catecholamine metabolism in AKU a liquid chromatography quadrupole time of flight (LC-QTOF) mass spectrometer method and spectral library were developed.

    Methods: Samples were analysed using the Agilent Infinity 1290 LC system coupled to a 6550 QTOF mass spectrometer. Chromatographic separation of adrenaline, noradrenaline, dopamine, tyrosine, metadrenaline, normetadrenaline, 3-methoxytyramine, vanillylmandelic acid and their respective internal standards was achieved following a 1 µL injection onto a BEH Amide column (3.0x150 mm, 1.7 µL). Mobile phase composition was (A) water and (B) acetonitrile (both contained 0.1% formic acid). Gradient started at 99% (B), decreasing to 70% by 12 minutes.

    Results: All compounds eluted between 3.5-5.8 minutes, during which the signal for the low m/z reference ion 121.0509 ranged from 92.7-43.6% of that at starting conditions. Optimal signal was achieved using a fragmentor voltage of 150 V (assessed 100-350 V).

    The linear range for all compounds was between 0.1-10 µmol/L. No signal was observed below 0.1 µmol/L and above 10 µmol/L the detector signal was saturated.

    In generating MS/MS data, collision energies of 0-40 V and signal threshold for triggering fragmentation were assessed; collision energies between 0-10 V and a signal threshold of 5000 counts were found to be optimal.

    Conclusions: An LC-QTOF method has been developed for the analysis of catecholamine metabolites to assess if nitisinone treatment alters catecholamine metabolism. Furthermore a spectral library has been generated that enables targeted assessment of catecholamine metabolism.

  • P164 Lipaemia can lead to spuriously low caeruloplasmin measurements with potential for misdiagnosis of Wilson’s disease

    Helen O Cordy (University Hospital of Wales, Cardiff, United Kingdom), Ian McDowell (University Hospital of Wales, Cardiff, United Kingdom), Paul Bramhall (University Hospital of Wales, Cardiff, United Kingdom), Jo Rogers (University Hospital of Wales, Cardiff, United Kingdom)

    We have observed markedly low serum caeruloplasmin results in patients being investigated for Wilson’s disease with normal serum copper values. These samples were subsequently found to be lipaemic. It was thought likely that low caeruloplasmin results were spurious and could potentially lead to a misdiagnosis of Wilson’s disease. In this study we sought to assess this further.

    Serum samples were collected in three groups: endogenously lipaemic samples, normal control samples, and a serum pool spiked with varying degrees of exogenous lipid. Samples were split in to two aliquots, one of which was spun by ultracentrifuge (Beckman Coulter Optima-MAX TL, Beckman Coulter Inc, California, USA) at 107000 x g for 15 minutes. Both the lipoprotein-reduced infranatant and the paired un-spun sample were analysed for serum caeruloplasmin using an immunoturbidometric method (Abbott Diagnostics, Maidenhead, UK) and copper by inductively coupled mass spectrometry (Agilent Technologies, Berkshire, UK).

    Twenty-four lipaemic serum samples, six control samples and ten spiked samples were analysed. Results demonstrated a clinically and statistically significant degree of lipaemic interference on serum caeruloplasmin for both endogenous and spiked samples (p<0.05). There was also a statistically significant difference between pre- and post-centrifugation of endogenous and spiked lipaemic samples for copper, however the mean percentage difference falls under the 10% cut-off for clinical relevance.

    Lipaemia is known to cause interference in some laboratory assays but this is not widely appreciated for caeruloplasmin. Mechanisms of interference include light scattering which can affect several methods employed in routine biochemistry analysis. This study demonstrated a significant degree of interference of lipaemia on serum caeruloplasmin; suggesting that for lipaemic samples under investigation for Wilson’s disease, serum copper measurements are more reliable.

  • P165 Analysis of mifepristone using ultra-rapid LC-MS/MS

    Victoria Treasure (Viapath Analytics, King's College Hospital, London, United Kingdom), Lewis Couchman (Viapath Analytics, King's College Hospital, London, United Kingdom), Tracy Dew (Viapath Analytics, King's College Hospital, London, United Kingdom), Roy Sherwood (Viapath Analytics, King's College Hospital, London, United Kingdom)

    An ultra-rapid method for the analysis of mifepristone in human plasma and serum samples using liquid chromatography-tandem mass spectrometry (LC-MS/MS) has been developed and validated to support a clinical study to investigate the neural basis of mifepristone treatment in bipolar disorder following single-dose administration.

    An Agilent 1290-6460 LC-MS/MS system was used with electrospray ionisation, in multiple reaction monitoring (MRM) mode. Chromatographic analysis was performed using a RaptorTM biphenyl column (5.0 x 3.0 mm, 2.7 µm particle size) fitted directly to the MS/MS nebuliser. System eluents were: eluent A, 0.1 % (v/v) formic acid in deionised water; eluent B, 0.1 % (v/v) formic acid in acetonitrile:methanol (1+1). The flow-rate was 0.75 mL/min, and the total analysis time (injection-to-injection) was just 60 seconds. Calibrators and internal quality control samples were prepared in analyte-free human plasma. Plasma and serum samples (50 µL) were prepared by precipitation with eluent B containing mifepristone-D3 (150 µL), and supernatants (2 µL) were directly injected for analysis.

    Typical retention times for mifepristone, mifepristone-D3 and normifepristone were 30.3, 29.8 and 29.1 seconds, respectively. Intra-batch and inter-batch precision (% RSD) and accuracy (% nominal concentrations) were <9 % and 98-104%. Carryover was <0.02 % on both autosampler needles. The assay was linear over the calibration range (20-2,000 µg/L). Mean (SD) matrix effects across five matrices at two concentrations were 14.6 (3.4) and 10.8 (3.1)% for mifepristone and mifepristone-D3, respectively. Mifepristone was analysed in 67 samples. In those samples where mifepristone was detected (>20 µg/L, n=13), the mean (range) concentration was 1085 (573-1918) µg/L.

    This high-throughput assay has been validated and is suitable for the analysis of mifepristone in human plasma or serum following administration of single doses to support ongoing clinical trials.

  • P166 UPLC-MS/MS analysis of busulfan in plasma for clinical research

    Stephen Balloch (Waters Corporation, Wilmslow, United Kingdom), Gareth Hammond (Waters Corporation, Wilmslow, United Kingdom)

    An analytically sensitive UPLC-MS/MS method for clinical research of busulfan pharmacokinetics in plasma using the Waters ACQUITY UPLC® I-Class and XEVO® TQD mass spectrometer has been developed.

    Samples (50 µL) were deproteinised with busulfan-2H8 internal standard in methanol. Elution was achieved within 2.5 minutes using a HSS-T3 C18 UPLC column (2.1x50 mm, 1.8 µm). Matrix-matched calibrators (0.025-5 µg/mL) and quality control samples (0.05, 0.75, 1.5 and 3.5 µg/mL) were prepared by gravimetric weighings of independent stocks of busulfan. Due to known instability of busulfan in plasma, aliquots of in-house calibrators and quality control samples were stored frozen and thawed prior to use for each assay.

    Analytical sensitivity was calculated to be 0.020 µg/mL (n=10 extractions, 5 occasions, ≤16.0% CV). Linearity was demonstrated over the concentration range 0.0175-6.51 µg/mL and system carryover was absent in samples ≤10 µg/mL. Precision studies (n=5, 5 occasions) demonstrated repeatability and total precision ≤7.3%. A method comparison was made by analysing anonymised plasma samples (n=40, range 0.20-2.28 µg/mL) against an independently validated UPLC-MS/MS method; good agreement was evident from the Deming equation y=1.01x+0.04. The mean recovery for busulfan pooled plasma samples (n=3, 0.05 and 3.5 µg/mL) was between 85.1-106.1% in the presence of high concentrations of endogenous compounds albumin, bilirubin, cholesterol, intralipid, triglycerides and uric acid and exogenous Intralipid®. The mean recovery was between 93.2-103.0% in the presence of acetaminophen, fluconazole, ketoconazole, itraconazole, phenytoin, posaconazole and voriconazole. Negligible matrix effects were observed at low (0.05 µg/mL) and high (3.5 µg/mL) concentrations as indicated by respective mean internal standard adjusted matrix factors of 0.99 (range 0.97-1.02) and 0.98 (range 0.95-1.00).

    The clinical research method developed for the quantification of busulfan demonstrated good linearity, analytical sensitivity and precision with negligible matrix effects.

    For Research Use Only. Not for use in diagnostic procedures.

  • P167 A UPLC-MS/MS method for the quantitative analysis of urinary free cortisol for clinical research

    Rachel Sanig (Waters Corporation, Wilmslow, United Kingdom), Gareth Hammond (Waters Corporation, Wilmslow, United Kingdom), Lisa Calton (Waters Corporation, Wilmslow, United Kingdom)

    Background: A simple and cost effective high throughput UPLC-MS/MS method for the quantitative analysis of urinary free cortisol (UFC), for clinical research investigations, has been developed.

    Methods: In-house calibrators were created using phosphate-buffered saline and cortisol reference material from Cerilliant (Round Rock, TX). The sample and internal standard (cortisol - 2H3, Sigma-Aldrich (Dorset, UK)) were diluted with aqueous trichloroacetic acid (0.5%), mixed, centrifuged and injected directly from a 96-well plate onto the Waters® ACQUITY UPLC® I-Class system. The samples were separated on a Waters ACQUITY UPLC CSH C18 column using a water/methanol/ammonium acetate/formic acid gradient with a run time of less than 3 minutes. The analytes were quantified using a Waters Xevo® TQD mass spectrometer.

    Results: The method demonstrated no interferences from analogous steroids, minimal matrix effects and no significant carryover. The assay was shown to be linear from 4.0 to 1340.0 nmol/L and total precision and repeatability coefficients of variation (CV) for four QC levels (20.7 nmol/L, 41.5 nmol/L, 206.9 nmol/L and 827.7 nmol/L) were all <10 % (n=5, days = 5). Analytical sensitivity of the method allows for cortisol quantification at 1.38 nmol/L (≤20%, inter-assay CV) and cortisol recovery was shown to be between 90 and 110%. The accuracy of the method was determined by analysing samples (years 2015-2016, n=48) from the UK NEQAS Scheme (Birmingham, UK); good agreement was obtained with a mean method bias of ≤4.1% for all samples (Altman-Bland).

    Conclusions: We have successfully developed a rapid and simple UPLC-MS/MS procedure for the determination of UFC concentrations for clinical research. The method has shown good analytical sensitivity, specificity, linearity, precision and accuracy.

    For Research Use Only, Not for use in diagnostic procedures.

  • P168 LC-MS/MS analysis of a serum steroid panel using solid phase extraction for clinical research

    Dominic Foley (Waters Corporation, Wilmslow, United Kingdom), Lisa Calton (Waters Corporation, Wilmslow, United Kingdom)

    Background: Here we evaluate an offline automated extraction method for the measurement of serum steroids; testosterone, androstenedione and dehydroepiandrosterone sulfate (DHEAS), 17-hydroxyprogesterone (17-OHP), cortisol, 11-deoxycortisol and 21-deoxycortisol, enabling steroid profiling for the investigation of metabolic dysfunction biomarkers for clinical research.

    Methods: One hundred µL serum samples were pre-treated with internal standard, methanol and water. SPE was performed using a Waters® Oasis® PRiME HLB µElution plate, allowing direct analysis of the SPE eluate. Sample extraction was automated using a Tecan® Freedom Evo 100. Using an ACQUITY UPLC® I-Class system, samples were injected onto a 2.1 x 50 mm Waters ACQUITY UPLC HSS T3 column with a pre-column T3 VanGuard™ using a water/methanol/ammonium acetate/formic acid gradient and quantified with a Waters Xevo® TQ-S micro mass spectrometer.

    Results: The method was shown to be linear for all of the steroids evaluated. Coefficients of variation (CV) for total precision and repeatability on five separate days for low, mid and high QC samples were ≤7.6% (n=30) for all analytes. Analytical sensitivity investigations demonstrate a CV <20% at 0.10 nmol/L for testosterone, 0.09 nmol/L for androstenedione, 0.19 nmol/L for 17-OHP, 0.011 µmol/L for DHEAS, 0.35 nmol/L for cortisol, and 0.72 nmol/L for 11-deoxycortisol and 21-deoxycortisol. Matrix Factor experiments demonstrate minimal matrix ion suppression (0.93–1.00) for all steroids. Excellent agreement between this analytical method and the EQA LC-MS mean values have been demonstrated with mean method bias of -0.1%, -5.1%, 5.2%, -7.1% and -1.4% for testosterone, androstenedione, 17-OHP, DHEAS and cortisol, respectively.

    Conclusions: We have developed an analytical method to quantify serum androgens and corticosteroids using SPE with LC-MS/MS for clinical research. This offline automated method demonstrates excellent linearity, analytical sensitivity, selectivity, precision and accuracy, while providing high sample throughput capabilities.

    For Research Use Only, Not for use in diagnostic procedures.

  • P169 Sample extraction optimization for the LC-MS/MS analysis of vitamin D metabolites in clinical research

    Robert Wardle (Waters Corporation, Wilmslow, United Kingdom), Dominic Foley (Waters Corporation, Wilmslow, United Kingdom), Lisa Calton (Waters Corporation, Wilmslow, United Kingdom)

    Background: Minimizing matrix interference, in particular from lysophosphatidylcholines (LysoPCs), is a challenge when analyzing vitamin D metabolites by LC-MS/MS. Using Waters® Oasis® HLB and Oasis PRiME HLB µElution solid phase extraction (SPE) plates, testing was carried out to assess the reduction of such interferences compared to protein precipitation. Performance testing of the optimized method was performed to show the potential for a clinical research method to be developed.

    Method and Results: Chromatographic separation of 25OHD2, 25OHD3, 24,25(OH)2D3 and C3-epi-25OHD3 enabled peak area profiles to be compared with targeted LysoPCs for the sample extraction methods. The Oasis PRiME HLB µElution plates provided the highest LysoPC depletion when using high concentrations (90-100%) of acetonitrile, with >99% of the targeted LysoPCs being removed when compared to Oasis HLB and protein precipitation. This resulted in >5x increase in peak area for all vitamin D metabolites combined when compared to Oasis HLB and >10x increase when compared to protein precipitation.

    The optimised method utilized a simple 3-step SPE protocol; load, wash and elute, with only one reagent being added to perform protein precipitation. Within-run and total precision of ≤8.2% CV was demonstrated and accuracy assessments showed that the mean bias was within ±10% bias for 25OHD3, 25OHD2 and C3-epi-25OHD3. Analytical sensitivity was determined at 0.5 nmol/L for 24,25(OH)2D3 and C3-epi-25OHD3 and 1nmol/L for 25OHD2 and 25OHD3 (%CV of <20% and a signal to noise ratio of >10:1).

    Conclusion: Elution profiling provides a useful and simple approach to SPE method development. An optimized SPE method using Oasis PRiME HLB has been developed for the LC-MS/MS analysis of serum vitamin D metabolites for clinical research. The method has excellent precision, accuracy, and analytical sensitivity with a simple, cleaner and faster sample preparation workflow.

    For Research Use Only, Not for use in diagnostic procedures.

  • P170 Metrological traceability of the new Waters MassTrak™ vitamin D assay

    Norma Breen (Waters Technologies, Wexford, Ireland), Leanne Davey (Waters Technologies, Wexford, Ireland), Robert Wardle (Waters Corporation, Wilmslow, United Kingdom), Lisa Calton (Waters Corporation, Wilmslow, United Kingdom)

    Background: The Waters MassTrak™ Vitamin D kit* is designed for the quantitative determination of serum or plasma 25-hydroxyvitamin D3 (25OHD3) and 25-hydroxyvitamin D2 (25OHD2), which in combination provide the total 25-hydroxyvitamin D concentration as an aid in the assessment of vitamin D sufficiency. Metrological traceability of the kit has been incorporated into the design, development and manufacture to meet the requirements of ISO 17511: 2003 which aids laboratories with their compliance to ISO 15189: 2012.

    Metrological Traceability: The MassTrak Vitamin D kit calibrators and quality control sets are traceable to NIST SRM 2972 through an unbroken chain of calibrations. Primary calibrators were prepared using certified reference materials and assigned 25OHD2 and 25OHD3 concentrations using reference measurement procedures at the Laboratory of Analytical Chemistry, University of Ghent, directed by Prof Dr Linda Thienpont. All MassTrak Vitamin D kit calibrator and quality control lots are value assigned using the primary calibrators. The concentrations assigned to the MassTrak Vitamin D kit are verified in a value confirmation process using independently assigned QC materials.

    To assess the accuracy and precision of the MassTrak Vitamin D kit assay, Waters enrols in the CDC Vitamin D Standardisation Certification Program (VDSCP) for total serum 25OHD. The programme assesses bias and precision of assays relative to reference measurement procedures. The MassTrak Vitamin D assay achieved a mean bias of 0.6% from the VDSCP reference values and a mean imprecision of 4.9%, meeting the certification performance criteria.

    Conclusion: The MassTrak Vitamin D kit calibrators are traceable to NIST SRM2972 via a documented unbroken chain of calibrations. The accuracy of the kit has been verified through participation in the VDSCP.

    *MassTrak™ Vitamin D kit is for in vitro diagnostic use. Not available for sale in all countries.

  • P171 Urinalysis: to acidify, or not to acidify, that is the question

    David J Marshall (University Hospital South Manchester, Manchester, United Kingdom), Brian Keevil (University Hospital South Manchester, Manchester, United Kingdom)

    Overview: Urolithiasis is characterised as the formation of stone deposits in the urinary tract. Urinalysis is an important tool for monitoring and identification of potential stone formers and is routinely carried out throughout the country.

    Aim: Using 6 years’ worth of laboratory data we aim to identify if acidification of 24 hour urine bottles prior to collection is required and how this affects stone screen analysis.

    Methods: Currently, 24 hour urine samples are collected in plain bottles and then acidified within the laboratory. To assess the effectiveness of this the patients at the Urology department at the University Hospital of South Manchester kindly agreed to take 2 separate urine collections. Paired 24 hour urine samples were sent to the laboratory, of which one was pre-acidified with HCL (prior to collection) and the other acidified within the laboratory (post-collection) following a well-defined SOP. Prior to the acidification step in the laboratory, aliquots were taken for analysis for a number of different analytes which included calcium, phosphate, protein, cystine and urate. These tests would have been unsuitable post acidification. Oxalate and citrate were quantified separately using an in house LC/MS/MS method for comparison. Calcium, phosphate and magnesium were analysed on the same Abbott c16000 Chemistry analyser using the manufacturer’s method. Results were sorted using clinical details including whether the sample was pre-acidified or acidified in the laboratory. Where no clinical details were stated the sample type was determined using the collection volumes recorded at analysis.

    Results: Comparison of pre and post acidified samples showed poor agreement for the analysis of calcium, phosphate and oxalate.

  • P172 Carcinoid: taken to a 5-higher (HIAA) level

    Leanne Evans (Liverpool Clinical Laboratories, Liverpool, United Kingdom), Andrew Hughes (Royal Liverpool and Broadgreen University Hospitals Trust, Liverpool, United Kingdom), Andrew Davison (Liverpool Clinical Laboratories, Liverpool, United Kingdom), Anna Milan (Liverpool Clinical Laboratories, Royal Liverpool and Broadgreen University Hospitals, Liverpool, United Kingdom)

    Introduction: Over the last decade the measurement of serotonin metabolite 5-hydroxyindole acetic acid (5-HIAA) by liquid chromatography-tandem mass spectrometry (LC-MS/MS) has become common practice for the diagnosis and monitoring of carcinoid disease. Despite the wide-spread utilisation of the assay there are relatively few commercial calibrators available. Chromsystems offer a single point calibrator which is diluted in-house to produce a 3-point calibration curve. Despite good EQA results indicating adequate assay performance, this is not good practice. Poor stability currently necessitates fresh calibrator use with each patient batch. Evaluation of a new 7-point Chromsystems calibrator set was undertaken driven by the requirement to meet ISO 15189 standards, a need for an increased calibration range and improved stability.

    Method: The 7-point calibrator set was included on routine patient sample batches over a period of two months. Analysis by reverse phase chromatography was performed using a Waters Atlantis® C18 column and Waters Quattro PremierTM mass spectrometer. Identification and quantification was based upon multiple reaction monitoring transitions m/z 191.85>146.4 for 5-HIAA and 196.9>150.4 for d5-5-HIAA respectively.

    Discussion: Good correlation upon patient sample comparison using current calibration and new 7 point calibration (r2=0.9998, n=80). Spiked samples processed using the two calibrator sets showed good correlation (r2=1) and accuracy using EQA (r2=0.99). There was an increased stability of calibrators to 2 months when frozen (~4.3% difference on average) reducing the need for reconstitution with every patient batch. Introduction of the new 7-point calibrator set into routine use will therefore lead to significantly reduced calibrator costs, improved inter-batch precision and the extended calibrator linearity will reduce need for sample dilution. Inter-lab variability would be improved with LC-MS/MS users adopting this multi-point stable calibrator set.

  • P173 How a change in reagent led to an increase in blood product use

    Emma J Lewis (Countess of Chester NHS Trust, Chester, United Kingdom), Clare Barnard (Countess of Chester NHS Trust, Chester, United Kingdom), Nicola Swarbrick (Countess of Chester NHS Trust, Chester, United Kingdom)

    In February 2016 the laboratory was notified that the supplier of our albumin method had gone out of business and no longer able to supply us with reagent. This left us with 2 weeks to source a new supplier, verify the new method and inform all our users of the change. Verification of the new method showed a drop in values and we changed our albumin and globulin reference ranges to compensate for this.

    Subsequently we heard that the transfusion department were having a noticeable increase in their human albumin solution usage. At first this was attributed to a particularly ill patient but as the increase continued for several months a new explanation was sought. Further investigation showed the main source of the increased usage in albumin was the intensive care unit (ICU).

    Six months after changing the albumin reagent and the reference range we were again notified that we would have to change supplier of albumin reagent with a similar short lead in time. Analysis of the comparison data showed that the reference range again needed to be changed, this time there was an increase. Following this second change in reagent the usage of albumin dropped.

    Whilst the laboratory changed its quoted reference range to reflect the change in levels due to the reagent this change was not made in ICU documentation where albumin is transfused at a fixed level. Whilst users were notified of the changes in reference ranges this had clearly not led to a change in documentation in use in the ICU.

    This shows that the laboratory needs to be aware of the impact small changes can make on other services.

  • P174 Prevalence of folate ‘deficiency’ increases 3-fold after re-standardisation of method

    Claudia R Tomkins (Kettering General Hospital, Kettering, United Kingdom), Richard J Harris (Kettering General Hospital, Kettering, United Kingdom), Gwyn McCreanor (Kettering General Hospital, Kettering, United Kingdom)

    Background: Roche Diagnostics re-standardised their Elecsys folate III assay and published a revised reference range of 3.9-26.8 µg/L. At Kettering General Hospital we verified this new version against the previous Folate III, and implemented it in September 2016 with reference range 3.9-20.0 (upper limit of measuring range being 20 µg/L). After implementation it was noted that a higher proportion of patient results were below the reference range compared to previously, when reference range was 4.6-20 µg/L.

    Aims: To compare a month of patient serum results, with respect to the reference range, before and after the folate assay change.

    Methods: All patient folate results during August 2016 (previous assay) and October 2016 (re-standardised assay) were extracted from the laboratory information management system. Colleagues in other laboratories in the UK were contacted to compare reference ranges.

    Results: Median folate concentration decreased from 7.6 µg/L to 5.2 µg/L after re-standardisation. Percentage of patient results below the reference range increased from 10% to 30% while the percentage above the reference range remained at 7%. Folate workload remained constant (August 3552, October 3651). Despite the re-standardisation aiming to bring results in line with other methods, the lower limit of the Roche reference range is considerably higher than that of both Siemens (2.6 µg/L) and Abbott Diagnostics (3.1 µg/L).

    Discussion: Patient results decreased considerably after re-standardisation of Folate III, as expected from our verification data. However, the revised Roche reference range does not account for the magnitude of this change, and is not consistent with other manufacturers. The British Society of Haematology guidance recommends treatment with folic acid at serum concentrations <3 µg/L therefore we propose this as an appropriate lower reference limit. The current limit 3.9 µg/L is likely to cause confusion amongst non-specialists as to what action to take and possibly lead to inappropriate treatment and/or unnecessary repeat testing.

  • P175 Development of a functional thiamine assay to assess thiamine status and management of chronic alcoholic patients

    Alana J Burns (NHS GG&C, Glasgow, United Kingdom), Donogh Maguire (NHS GG&C, Glasgow, United Kingdom), Fiona Stefanowicz (NHS GG&C, Glasgow, United Kingdom), Dinesh Talwar (NHS GG&C, Glasgow, United Kingdom)

    Thiamine (vitamin B1) is important for the metabolism of carbohydrates, fats, alcohol and contributes to nervous system protection. Thiamine is required by the body in the form of thiamine diphosphate (TDP). TDP is a co-factor for transketolase activation in the pentose phosphate shunt to provide NADPH necessary for biosynthetic reactions.

    Thiamine stores are relatively small (~30mg) thus, deficiencies can develop quickly particularly in chronic alcoholics, the elderly and patients requiring parenteral nutrition. As a thiamine deficiency develops the levels of TDP in erythrocytes and brain tissue remain stable until stores become significantly deplete; around 90% of circulating TDP is found in erythrocytes. A decrease in thiamine levels in erythrocytes correlates with a decrease in the TDP in other tissues. Low thiamine status is implicated in the aetiology of Wernicke’s encephalopathy with cases confirmed in 2% of autopsies in western society. WE is characterised by the triad of symptoms: confusion, ataxia and ophthalmoplegia.

    Currently, nutritional status of thiamine may be assessed in the laboratory by direct measurement of TDP in whole blood using HPLC, or by measurement of the functional marker erythrocyte transketolase activity (ETKA) in the absence or presence of TDP in vitro. While direct measurement of thiamine in whole blood has been shown to be helpful in the initial assessment of thiamine status, there has been some questions over its utility in monitoring the efficacy of treatment. Historically, ETKA has been measured using a centrifugal analyser with spectrophotometric measurement of NADPH consumption. We have adapted, optimised and semi-automated an existing literature method for the measurement of ETKA for use with a BioTek Synergy H1 plate reader. ETKA will be used to assess thiamine status and efficacy of treatment in chronic alcoholic patients and compared to direct measurement of TDP, ETKA may be more useful monitoring patients who undergoing treatment.

  • P176 Effect of pH on an oxalate oxidase method for plasma oxalate

    Hayley Sharrod-Cole (Birmingham Children's Hospital, Birmingham, United Kingdom), Mary Anne Preece (Birmingham Children's Hospital, Birmingham, United Kingdom), Adam Gerrard (Birmingham Children's Hospital, Birmingham, United Kingdom)

    Background: Published methods for the measurement of oxalate in plasma recommend the dilution of plasma in acid, prior to an ultrafiltration step, to limit the spontaneous conversion of L-ascorbate to oxalate.

    Method validation revealed unacceptable recovery of spiked oxalate from plasma. It was hypothesised that sample pH after dilution and centrifugation was critical for the generation of accurate and unbiased results.

    Aim: The aim was to determine the effect of the diluent on the measured concentration of oxalate in plasma.

    Methods: Plasma samples were diluted 1:1 with 0.24 M HCl and filtered through a 30 kDa filter. Oxalate in the resulting ultra-filtrate was measured using the Trinity Biotech oxalate kit, which is based on an oxalate oxidase reaction.

    To assess accuracy, plasma samples were spiked with doubling concentrations of an oxalate certified reference material (CRM). The samples were prepared using one of 3 diluents of increasing pH (0.24 M HCl, 0.01 M HCl and H20). The pH values of the resulting extracts were 2.0, 5.5 and 7.0 respectively. Recovery was calculated.

    Bias was investigated by measuring oxalate CRM at 3 concentrations (5.7, 57 and 142 µmol/L). The pH of the CRM was modified by addition of 2 µL concentrated HCl, 0.01 M NaOH or 1M NaOH to 400 µL. Linearity, precision and stability were also assessed using various diluents.

    Summary: Bias and recovery were optimal using water as a diluent, and appeared to be inversely related to the pH of the diluent. It is suggested that this is due to protein precipitation prior to filtration resulting in loss of oxalate, and also due to suboptimal oxalate oxidase activity at acidic pH.

    Conclusions: Sample pH after dilution and centrifugation has a significant effect on the bias and accuracy of measured oxalate in plasma when using an oxalate oxidase method. Dilution of plasma samples in water is recommended.

  • P177 Tyrosine metabolites in alkaptonuria

    Andrew T Hughes (Liverpool Clinical Laboratories, Liverpool, United Kingdom), Anna Milan (Liverpool Clinical Laboratories, Royal Liverpool and Broadgreen University Hospitals, Liverpool, United Kingdom), Andrew Davison (Liverpool Clinical Laboratories, Liverpool, United Kingdom), James Gallagher (Ageing and Chronic Disease, Liverpool, United Kingdom), Lakshminarayan Ranganath (Liverpool Clinical Laboratories, Liverpool, United Kingdom)

    Introduction: Alkaptonuria (AKU) is a rare debilitating autosomal recessive disorder affecting tyrosine metabolism. Deficiency of homogentisate 1,2-dioxygenase results in increased circulating homogentisic acid (HGA) which is deposited as an ochronotic pigment, resulting in early onset arthritis and other complications including increased stone formation and cardiac complications. The National AKU Centre (NAC) at Liverpool is using 2mg nitisinone off-licence to treat patients with AKU where it is important to monitor tyrosine metabolites from a safety and efficacy point of view.

    Methods: The aim was to expand our previously published LC-MS/MS methods for both serum and urine to include hydroxyphenyllactate (HPLA), hydroxyphenylpyruvate (HPPA) and phenylalanine (Phe), with subsequent method re-validation. Analysis was performed on an Agilent 6490 QQQ (ESI-MS/MS). Mass detection was in negative mode for HPPA and HPLA and positive for Phe and d5Phe.

    Using a multiple analyte matrix-matched calibration curve patient samples were analysed pre and post-nitisinone to determine the metabolic profile of AKU patients upon treatment.

    Results: Validation demonstrated a robust assay with recovery, intra and inter-precision within acceptable limits; recovery in serum was 93.3-114.1% for HPLA, 93.1-112.6% for HPPA and 92.4-104.2 for Phe; in urine 91.7-115% (HPLA), 87.7-111.6% (HPPA) and 94.4-102.7% for Phe. Intra-assay precision (n=11) was <10.9%, <11.0% and <6.3% for serum HPLA, HPPA and Phe respectively; with equivalent CVs in urine of <9.4%, <7.2% and <6%. Inter-assay precision (n=20) was <10.6% (HPLA), <10.0% (HPPA) and <5.6% (Phe) in serum and <9.0% (HPPA), <9.9% (HPLA) and <8.1% (Phe) in urine. Phe accuracy was determined by performance in ERNDIM EQA scheme (serum only).

    Discussion: A multi-metabolite assay has been validated to quantitate tyrosine, phenylalanine, HGA, HPPA, HPLA and nitisinone. The assay has been applied to serum and urine of AKU patients both in the NAC and International DevelopAKUre Clinical trials.

  • P178 Validation of fluid analysis to conform to UKAS standard as outlined in ISO 15189 Louise

    J Ward (Viapath, Bedford Hospital NHS Trust, Bedford, United Kingdom), Nancy Wassef (Viapath, Bedford Hospital NHS Trust, Bedford, United Kingdom), Sally Brady (Viapath, St Thomas' Hospital, London, United Kingdom), Stephen Quayle (Viapath, Bedford Hospital NHS Trust, Bedford, United Kingdom), Janet Reeves (Viapath, King's College Hospital, London, United Kingdom), WS Wassif (Viapath, Bedford Hospital NHS Trust, Bedford, United Kingdom)

    Laboratories issue reports for assays performed on body fluids which are not validated by manufacturers e.g. pleural fluid albumin. To conform to UKAS standard, these assays require verification and validation as outlined in ISO 15189 standards 5.5.1.2 and 5.5.1.3. This study provided sufficient data to fulfil these criteria for albumin, glucose, total protein, triglyceride, amylase and creatinine assays using the Roche COBAS 8000.

    Validation: Within-batch precision assessed using neat fluids and sequential dilutions of fluid and serum to provide data across a range of assay concentrations. The linearity/recovery of the assays was studied using sequential dilutions. Between-batch assessment using 10 different fluids. IQC and EQA data were also examined.

    Verification: Performance of fluid assays studied using a set of 26 anonymised fluids analysed on three sites, Bedford (Roche COBAS 8000), St Thomas’ (Roche COBAS 8000) and King’s (Siemens ADVIA 2400).

    Validation results: Two neat fluids for each assay determined within batch precision (n=5). The experimental CV ranged 0.38-2.63% compared to 0.5-2.3% published by Roche. The between batch CV range across all assays for the 10 fluids was 0.6-9.8% compared to 0.9-3.7% published by Roche. IQC and EQA data were within target limits.

    Verification results: (mean±SD, Roche COBAS 8000 Bedford, Roche COBAS 8000 St Thomas’, Siemens ADVIA King’s; respectively) for albumin g/L (17±8, 17±8, 18±8), total protein g/L (29±13, 31±13, 30±13), amylase U/L (42±28, 42±27, 40±26), glucose mmol/L (6.1±3.9, 6.1±3.7, 5.9±3.6), triglyceride mmol/L (0.6±1.1, 0.5±0.8, 0.6±1.1) and creatinine µmol/L (65±31, 80±32, 61±30). No significant difference noted for all analytes (ANOVA, P between 0.1-0.98). As such ISO accreditation of fluid assays which are not validated by manufacturers is achievable. However, as EQA is unavailable for these sample types, regular fluid sample swaps is necessary as an alternative.

  • P179 Introduction and validation of an immunoturbidimetric calprotectin assay

    Kelly Foley (Cork University Hospital, Cork, Ireland)

    Introduction: Calprotectin has been widely accepted as a marker of inflammation and has gained widespread use in the differential diagnosis of IBD from IBS. Most methods in use for calprotectin measurement are ELISA based. These ELISA methods can be time consuming and sometimes require an extra piece of equipment. In the ever-increasing requirement for doing more with less in the health service, being able to utilise equipment already in situ provides for easier introduction of new tests. For this reason we chose to validate the Buhlmann immunoturbidimetric method (fCAL™ turbo) on the Beckman AU5800.

    Methods: The extraction method used was the Calex tube recommended and provided for by Buhlmann. The analytical performance specifications of the Buhlmann immunoturbidimetric calprotectin assay were assessed. Precision studies were done using low, medium and high level samples. Linearity was tested using a high calprotectin sample serial diluted and observed data plotted against expected. For method comparison, 110 samples that were requested for calprotectin were aliquoted. One sample was sent to an external reference laboratory for analysis (Buhlmann ELISA) and the other analysed in house using fCAL turbo.

    Results: Precision of the new assay was acceptable with CVs all below 20%. The assay was linear 2525 µg/g. The precision of the extraction method showed to be very good, although it was noted that mucus samples can show variation. Comparison studies showed acceptable correlation with (r=0.917). However, it was observed that certain discrepant results could be found when the sample was mucous like. Any discrepant results were further investigated for delay in transportation and sample consistency.

    Conclusion: The Buhlmann Immunoturb method is a satisfactory method for introduction into the busy biochemistry laboratory. Demand management measures will need to be put in place as well as certain acceptance criteria for sample types.

  • P180 A case of interference in Abbott Architect creatine kinase assay

    Sarah Darch (Imperial Healthcare NHS Trust, London, United Kingdom), Jaimini Cegla (Imperial Healthcare NHS Trust, London, United Kingdom)

    A 28 year-old male collapsed whilst running. On waking in the paramedics tent, he was found to be tachycardic and tachypnoeic. The patient had no previous medical or family history, exercised regularly, was a social smoker and consumed alcohol at weekends. He admitted to occasional cocaine use, having taken 1 g of cocaine two nights prior to the event.

    On admission, his ECG showed early transition with tall R-waves anteriorly, however his Troponin-I, ALT and AST all followed the rise and fall expected with myocardial damage and the lag time of the ALT and AST compared to troponin-I were characteristic. The high level of serum troponin-I (2303 ng/L) indicated significant myocardial injury. However, the CK measurements (within normal limits) were completely at odds with the other cardiac markers. Suspicions were raised that there may be an interfering factor affecting the assay, therefore linear dilutions of the sample were performed.

    The non-linear dilution results confirmed that the patient had a serum interferent. The Abbott Architect Creatinine Kinase (NAC (N-acetyl-L-cysteine) assay involves three successive reactions and creatine is an intermediary. The question as to whether the patient was taking creatine supplementation was raised with the clinical team. The patient admitted to regularly taking protein powder supplements. Protein powder supplements contain creatine and are very likely to be the cause of the interference observed.

    UE, troponin-I and LFTs all returned to normal. In-patient CT coronary angiogram revealed normal coronary arteries and coronary vasospasm was deemed the likely cause of troponin release. The patient was advised to avoid further recreational drug use. Follow up cardiac MRI 3 months later, showed no long-term cardiac damage. Users of the Abbott Architect CK assay should be aware that protein supplements containing creatine may cause interference with the assay.

  • P181 Plasma protein quantitation by tryptic peptide LC-MS/MS: what analytical goals can the clinical laboratory achieve?

    Craig McKibbin (Surrey Pathology Services, Guildford, United Kingdom)

    The use of targeted tryptic peptide LC-MS/MS assays to quantify specific proteins in biological fluids represents a largely unrealised opportunity for extending the services of specialist diagnostic testing laboratories and accelerating the pace of translational research.

    To explore the potential of instrumentation available to clinical laboratories for targeted protein assay development, selective reaction monitoring (SRM) peak intensity and reproducibility were determined for a published set of plasma protein transitions representing the top four orders of mass concentration abundance using a mid-range triple quadrupole mass spectrometer equipped with an HPLC system and standard bore diameter reverse phase columns. Samples were prepared by in-house digest method or commercial digest kit (with or without prior immunodepletion of albumin and immunoglobulins) followed by solid phase extraction.

    Conclusion: Tryptic peptide LC-MS/MS is a versatile and increasingly feasible methodology for the clinical laboratory interested in measuring specific protein targets using instruments and equipment many already use for drug-like molecule analysis. Plasma proteins in the 100 mg/L range are accessible with reasonable imprecision performance. Furthermore, tryptic peptide LC-MS/MS has much potential for the rapid translation of research medicine to the clinic, cost-effective development of plasma protein test panels, identification of genotypic and regulatory variants at the protein level, and the validation of immunoassay methods.

  • P183 Development of LC-MS/MS method for the quantification of red cell protoporphyrin in the diagnosis of erythropoietic protoporphyria and comparison with fluorimetric techniques

    Teng Teng To (Salford Royal NHS Foundation Trust, Salford, United Kingdom), Paul Reed (Salford Royal NHS Foundation Trust, Salford, United Kingdom)

    Introduction: Erythropoietic protoporphyria ( EPP) is an autosomal dominant inherited disorder with a defective ferrochelatase gene or aminolevulinic acid synthase-2-gene. At Salford Royal, EPP testing is carried out using a fluorimetric assay, with a second extraction carried out if total red cell protoporphyrins are raised (>2.0 µmol/L) to quantify free (PPIX) and zinc protoporphyrin (ZnPPIX).

    Aim: To develop a novel LC-MS/MS method to simultaneously quantify free and zinc protoporphyrin and compare the results from this and a synchronous fluorimetric method to our in-house fluorimetric assay.

    Method: Extraction procedures for protoporphyrins from red cells and isolation using liquid chromatographic parameters were explored. Tuning for free and zinc protoporphyrin was carried out on a tandem mass spectrometer. PPIX, ZnPPIX, coproporphyrin III hydrochloride, uroporphyrin I standards, EQA samples and patient samples (confirmed EPP patients and normal patients) were used to investigate carry-over, linearity, precision, standard stability, assay interference and quantification of PPIX and ZnPPIX in red cells.

    Results: A preliminary LC-MS/MS method for the quantification of PPIX and ZnPPIX was developed. Calibration curves for PPIX and ZnPPIX show good linearity, r2=0.947 and r2=0.936 respectively. Method comparison of in-house assay to LC-MS/MS method showed poor correlation for ZnPPIX measurements, r2=0.7309, with ZnPPIX being half in-house assay values. However, good correlation, r2=0.9604, for PPIX readings were obtained, but LC-MS/MS results were higher by a factor of 1.5. Method comparison of in-house fluorimetric technique to synchronous fluorimetric assay showed overall good correlation (up to r2=0.993), but a factor of up to 2.5 was observed in ZnPPIX results.

    Conclusion: Further work is needed to establish an appropriate internal standard and preparation/ storage of calibration standards, since these showed instability and required fresh preparation each time. Further work is needed to address the difference in PPIX and ZnPPIX concentrations obtained using the various techniques.

  • P184 Accelerating LC-MS/MS analysis of 25-OH-Vitamin D2 and D3 analysis in the clinical laboratory

    Nicola Gray (Shimadzu UK Ltd, Milton Keynes, United Kingdom), Christopher Titman (Shimadzu UK Ltd, Milton Keynes, United Kingdom), Stuart Phillips (Shimadzu UK Ltd, Milton Keynes, United Kingdom)

    Introduction: Vitamin D deficiency has been linked to several disease conditions and is widely tested in human serum in clinical laboratories. Vitamin D2 and D3levels are primarily measured using the 25-hydroxy-vitamin D metabolites due to higher circulating serum levels and longer half-lives. While the majority of clinical methodologies for measuring vitamin D in the UK remain immunoassays, liquid chromatography-tandem mass spectrometry (LC-MS/MS) was recently proposed as a reference method for vitamin D determination due to enhanced accuracy and the ability distinguish 25-OH-D3 and D2.

    Methods: Here we demonstrate a method for the quantification of 25-OH-Vitamin D3/D2 using the Nexera MX system for high-throughput analysis. The Nexera MX Dual Stream Technology (MX-DST) features ultra-high speed operation for high-throughput analysis with two separate flow streams and overlapping sample injections, combined with dedicated MX Solution software for simplified batch and method creation. The LCMS-8060 has a number of features which make it powerful tool for sensitive quantification of vitamin D to provide a facile implementation of the analytical method.

    Conclusions: The Nexera MX system offers reduced analysis time by multiplexing sample analysis for increased throughput, resulting in a 5-hour analysis time for a 96-well plate permitting same-day results. The specification of the ultra-fast LCMS-8060 enables sensitive, robust and reliable quantification of 25-OH-Vitamin D3/D2 in serum/plasma. Combined with the Nexera MX configuration, this system provides a simple solution for high-throughput routine quantification.

  • P186 Investigating the relationship between triglycerides and lipaemia in serum samples

    Laura Lewis (Heartlands Hospital, Birmingham, United Kingdom), Alexander Lawson (Heart of England Foundation Trust, Birmingham, United Kingdom), Craig Webster (Heartlands Hospital, Birmingham, United Kingdom)

    Lipaemia in serum samples can affect many tests within the laboratory leading to falsely high or low results. Most laboratories measure the extent of lipaemia in patient samples (lipaemic index) and apply cut-offs to prevent incorrect results from being reported. The lipaemic cut-offs are generally taken from the reagent kit inserts.

    An incidental finding was discovered on a patient who had high triglycerides (>50 mmol/L) and obvious lipaemia; all tests had been automatically reported as ‘unable to analyse’ with the exception of troponin. According to Abbott Architect kit insert triglycerides over 30 g/L (33.9 mmol/L) may cause interference in the troponin assay; hence the result should have been knocked-out. On further investigation it was discovered that some Abbott kit inserts state Intralipid values for the lipaemia cut-off, whereas others state triglyceride concentrations, meaning the values used for our cut-offs are not consistent in their basis. We sought to investigate the relationship between lipaemic index, Intralipid concentration and triglyceride concentration.

    Serum was spiked with Intralipid of varying concentrations before measuring lipaemic index and triglyceride concentration. Intralipid concentration had a reasonable agreement with lipaemic index (y=0.8x+0.2). Intralipid concentration had a linear relationship with triglyceride concentration; however triglycerides were approximately 2.6 times higher than intralipid concentration (y=2.6x+0.9). This meant that lipaemia cut-offs based on triglyceride concentration were approximately 2.6 times higher than they should be, allowing results to be reported on samples that potentially have significance interference from lipaemia.

    The result of this work allowed us to recalculate lipaemic index cut-offs for assays where only a triglyceride concentration cut-off was given, thereby ensuring better quality of results reported. It is also an important lesson on the care that must be taken during initial implementation of such limits.

  • P187 Developing the clinical biochemistry research leaders of the future: opportunities to work with The National Institute for Health Research

    Michael Messenger (Leeds Centre for Personalised Medicine and Health, University of Leeds, Leeds, United Kingdom; NIHR CRN West Midlands Lead for Laboratory Medicine, Birmingham, United Kingdom), Owen Driskell (NIHR CRN West Midlands Lead for Laboratory Medicine, Birmingham, United Kingdom), Catharine Sturgeon (Royal Infirmary of Edinburgh, Edinburgh, United Kingdom; NIHR Diagnostic Evidence Co-Operative Leeds)

    The National Institute for Health Research (NIHR) is committed to attracting, developing and retaining a highly skilled clinical laboratory workforce in order to undertake world-class research to improve the health and wealth of the nation. The purpose of this poster is to inspire the next generation of Clinical Biochemists and other laboratory professionals to participate in research by highlighting the numerous opportunities available to them through the NIHR.

    The NIHR supports approximately 5,000 trainees from a wide range of professional groups. The awards range from early phase Masters Studentships right through to Professorships, with funding provided for equipment & consumables, training courses and international conferences. In addition to the awards themselves, the NIHR also provides a range of development opportunities, including leadership courses and mentoring, encompassed in its training programmes.

    The NIHR strongly encourages working with experts across different disciplines, to bring in new perspectives. For example Clinical Biochemistry trainees might work with engineers, biologists, mathematicians, psychologists, data-scientists and economists. Many trainees also work closely with industry, developing insight, experience and skills that put them in a position to drive forward change within their profession and clinical service.

    The NIHR infrastructure is there to support the development of strong applications. The NIHR Diagnostic Evidence Co-Operatives (DECs) are specifically focused on delivering research on in vitro diagnostic (IVD) devices and are therefore ideal partners for Clinical Biochemists and other laboratory professions. The NIHR Research Design Service, Clinical Research Network and DECs can help with job planning, identifying academic and commercial partners, research methods, delivery and dissemination to ensure impact and adoption. NIHR trainees are prepared for a longer-term research career, and complete their training with more than just an academic qualification.

Updated 25 April 2017

© Copyright The Association for Clinical Biochemistry and Laboratory Medicine
The Association for Clinical Biochemistry and Laboratory Medicine, 130-132 Tooley Street LONDON SE1 2TU
Tel: 020 7403 8001 Fax: 020 7403 8006