Notes
P23
Test request variability in primary care in Catalonia
Ibarz M*1, Perich C2, Llopis MA3, Alvarez V4, Domenech MV5, Pastor R6, Busquets G7, LLovet MI8, Simon M9, Biosca C10
1Laboratori Clínic Hospital Arnau de Vilanova, Lleida, Spain
2Laboratori Clínic Bon Pastor, Barcelona, Spain
3Laboratori Clínic Barcelonès Nord i Vallès Oriental, Badalona, Spain
4Laboratori Clínic L’Hospitalet, L’Hospitalet de Llobregat, Barcelona, Spain
5Laboratori Clínic Manso, Barcelona, Spain
6Laboratori Clínic Hospital Joan XXIII, Tarragona, Spain
7Laboratori Clínic Hospital Dr. Josep Trueta, Girona, Spain
8Laboratori Clínic Hospital Verge de la Cinta, Tortosa, Spain
9Consorci de Laboratori Intercomarcal Anoia, Penedés i Garraf, Vilafranca del Penedés, Spain
10Servei Bioquímica Clínica Hospital Germans Trias i Pujol, Badalona, Spain
Background: Test request variabilityis an important performance indicator for preanalytical quality and laboratory efficiency. To know this variability is the first step to improve the use of clinical tests.
The aim of this study was to describe and evaluate the differences and possible determining factors of test request variability in primary care in a group of laboratories belonging to the public network of the Catalonian Public Health Service.
Materials and methods: The group is formed by ten laboratories, four provide services exclusively to primary care and six to hospitalized and primary care patients. Statistics of 2011 were obtained and the number of tests was set to the number of primary requests for every laboratory. The median and percentiles ratios were calculated. Possible determining factors were studied.
Results: 2.815.563 primary care requests were processed with a total of 32.706.843 analytical tests. The 15 most frequently ordered tests accounted for 60% of the laboratory activity and they were in descending order of frequency (depending on the median of the ratios of the different laboratories): cell blood count, glucose, cholesterol, creatinine, alanine aminotransferase, triglycerides, HDL-cholesterol, gamma-glutamiyltranspeptidase, uric acid, electrolyte profile (sodium and potassium), urinalysis, bilirrubine, thyrotropin, aspartate aminotransferase and alkaline phosphatase. Median ratios for all laboratories were within the interquartile range for the 7 first ones of the list and for thyrotropin as well.
Conclusions: These results show thatthe test ordering is fairly homogeneous in primary care in Catalonia, surely as consequence of the application of agreed protocols from 1997 for several pathologies and health states. The discrepant results could be due to diverse adaptations of protocols and/or diagnostic algorithms in laboratories and different specialists’ integration in primary care.
Key words: quality; preanalytical-phase; test-requiring
P24
Assessment of the preanalytical phase for molecular biology testing through participation in EU SPIDIA project
Kardum Paro MM*, Šiftar Z, Flegar- Meštrić Z
Department of Medical Biochemistry and Laboratory Medicine, „Merkur“ University Hospital, Zagreb, Croatia
Background: In molecular biology testing definite gene changes are identified or ruled out directly or indirectly on nucleic acids level. Because of variation of molecular biology testing across Europe, inter-laboratory comparability is possible through methodological proficiency testing (PT) programs intended for monitoring and controlling the quality of the preanalytical phase or External Quality Assurance (EQA) programs intended for relatively frequently tested diseases. To monitor and evaluate the laboratory preanalytical phase of molecular biology testing, we joined to the SPIDIA-DNA trial of the European Union (EU) project Standardisation and improvement of generic pre-analytical tools and procedures for in-vitro diagnostics (SPIDIA).
Materials and methods: In the 2nd SPIDIA- DNA trial the same whole blood sample was send to 174 participants to perform the DNA extraction using their own protocol and reagents. The participants were asked to make the DNA UV analysis and calculations, report results and additional information on the web site and to send back extracted DNA for further SPIDIA DNA analysis. The laboratory preanalytical phase of DNA extraction SPIDIA evaluated through DNA Quality and DNA Quantity by spectrophotometric method, DNA Integrity by ad-hocalgorithm combining by-eye evaluation of a panel of experts and Image J software through three categories (high, intermediate and low) and PCR kinetics interferences by real-time PCR using RnaseP and Kineret software. Final SPIDIA- DNA report included statistical analysis and the results interpretation in a form of a score which included the comparison of single laboratory performance with that of other participants.
Results: Statistical results of the DNA Quality and DNA Quantity (N = 174) by spectrophotometric method were 1,810 (lab) vs. 1,857 (SPIDIA) and 27,820 (ng/μL; lab) vs. 27,200 (ng/μL; SPIDIA). Results were scored and interpreted as „in control“. Only DNA Integrity (N = 158) was scored as of „intermediate integrity“. PCR kinetics evaluation (N = 173) showed value under the 5,99 SPIDIA threshold and score interpreted as „in control“.
Conclusions: Overall SPIDIA- DNA statistics objectively provides deep monitoring of the preanalytical phase of molecular biology testing. Results interpretation reflects state of the art in laboratory performance so that special emphasis could be placed on laboratories with sub- optimal performance.
Key words: SPIDIA; preanalytical phase; molecular biology testing
P25
Frequency of lipemia and its preanalytical impact on iron determination in pediatric population - our experience
Lenicek Krleza J*, Rajcic A, Grzunov A
Children’s Hospital Zagreb, Department of Laboratory Diagnostics, Zagreb, Croatia
Introduction: Samples for iron measurement should be taken in the morning from patients in a fasting state and there can be significant interference from lipemia. As most of our hospital clinics operate until the late afternoon hours, requests for iron determination pour in throughout all day.
Material and methods: The aim of this study was to determine the amount of lipemic samples accepted for iron determination in our laboratory over a period of three months. All samples were visually inspected for turbidity. For 41 lipemic samples time of sampling and patient age were noted. Iron concentration was determined two times in all lipemic sera. The first measurements from the original samples, and the second after the same samples were diluted 1:2 with normal saline. The iron level was determined on Olympus AU400 with a method for serum iron using TPTZ{2,4,6-Tri-(2-pyridyl)-5-triazine} as the chromogen. To compare the values of iron measured before and after dilution with saline nonparametric Wilcoxon test for paired data was used.
Results: 13% of samples were found visibly turbid. For 44% of lipemic samples sampling time was after 10 a.m. 71% of patients whose samples were lipemic were under 2 years of age. For 37% of lipemic sera a decision was made that results obtained from diluted samples are to be issued. Wilcoxon test did not show a significant difference between the two measures (p = 0.414).
Conclusion: Turbidity was observed in a significant number of samples. In a sizable number of cases, the results obtained from diluted samples were issued. Almost half of the lipemic samples were taken after 10 a.m., and it is very likely that the cause of lipemia is food consummation. The actual iron level in these patients is questionable, since iron values decrease by 30% during the course of the day.
Key words: child; iron; lipemia; sampling time; interference
P26
Preanalytical variability of the measurement of alpha fetoprotein in amniotic fluid: case report
Brescia V*, Scolozzi S, Circhetta S
U. O. C. Medicina di Laboratorio, A.O. Cardinale G. Panico Tricase, Lecce, Italy
Background: To evaluate a case of increased alpha-fetoprotein in amniotic fluid (AF-AFP) without ultrasonographic evidence of open neural tube defects (NTD) or other fetal abnormalities.
Materials and methods: A 38-year-old (age-specific risk) pregnant woman referred mild blood loss at the end of the third month of pregnancy; genetic amniocentesis and AFP-AP assay were performed at the 16th week. AF showed slight blood contamination. A primary aliquot of 2 mL AF was centrifuged at 1200 rpm within 3 hrs of withdrawal. AFP-AF assay was done with an automatic DxC880i Beckman Coulter (Chemiluminescent Immunoassay). The corpuscular part of the AF was washed with physiological solution, recentrifuged and lysed with redistilled water for HBF assay in high-performance liquid chromatography HPLC (Variant II BIORAD). The second ultrasound screening was negative for NTD or other fetal abnormalities (18th week).
Results: AFP-AP was 32,4 µg/mL (> 2.0 < 2.5 MoM), HBF on lysed AF blood was positive for fetal bleeding (16%). The AFP-AF findings were integrated with the warning “contamination by fetal blood: AFP-AF value not specific for NTD”. The pregnancy outcome was physiological.
Conclusion: The AFP-AF test is useful for screening for NTD or other fetal abnormalities. There is a high AFP concentration gradient between fetal plasma and AF. The presence of fetal bleeding in the AF can cause false positive increases in AFP-AF. The fetal hemoglobin assay in HPLC we adopted was shown to be reliable and not influenced by a different matrix. Therefore, in our laboratory all AF specimens testing positive for AFP (> 2.0 MoM) will also be tested for the presence of HBF to minimize errors in NTD risk estimation.
Key words: alpha-fetoprotein; amniotic fluid; open neural tube defects
P27
Comparison between 6-hour and 24-hour cupriuria in monitoring of therapy with D-penicillamine in patients with Wilson disease
Ivanova I*1, Krastev Z1, Petkova T1, Dragneva S2
1Department of Clinical Laboratory, University Hospital St. I. Rilski, Medical University, Sofia, Bulgaria
2Clinic of Gastroenterology, University Hospital St. I. Rilski, Medical University, Sofia, Bulgaria
Background: Adequacy of treatment in patients with Wilson Disease (WD) was monitored by measuring 24-hour urinary copper excretion while on treatment with D-penicillamine. The major effect of the drug is to promote the urinary excretion of copper. Values between 3-8 µmol/24h (200-500 µg) per day cupriuria means adequate therapy. The problems of measuring 24-hour copper excretion include: incomplete urine collection and copper contamination of the collection device. Therefore, by reducing time for testing we will also reduce errors in preanalytical stage of examination.
Materials and methods: The aim of the study was to assess and compare the 6-hour and 24-hour copper urinary excretion in patients with WD treated with D-penicillamine. The data used ware from 42 measurements in 32 patients with proven WD in University Hospital St. Ivan Rilski, Medical University-Sofia. The average dose D-penicillamine was 800-1000 mg per day. Urine was collected in two portions – 6-hour and 18-hour. The concentration of copper in two portions ware calculated to get 24-hour cupriuria. Copper analysis was made with atomic absorption spectrophotometry in Analyst 400 device.
Results: The copper urine excretion in 6-hour, 18-hour and in 24-hour urine was respectively x = 4.12 ± 2.26 µmol / 6h, x = 5.75 ± 3.42 µmol / 18h and x = 9.89 ± 4.7 µmol / 24h. Copper excretion in 6-hour urine was 41.7% of 24-hour cupriuria.
Conclusion: About half of copper urine excretion was up to 6-hour diuresis. Reducing the time of urine collection could reduce the errors in preanalytical stage of examination. It is necessary the 6-hour urine copper excretion to be standardized and validated to use it as criteria of adequacy of therapy.
Key words: cupriuria; Wilson Disease; D-penicillamine
P28
The simultaneous effect of closed system urine sampling and the urine stabilizing additive on the outcome of urine sample evaluation
Lakatos A*, Nagy Cs Zs, Rékási Zs, Magyarlaki T, Kovács LG
Clinical Center of the University of Pécs, Laboratory Medicine Institute, Pécs
Background: The preanalytical factors have major importance in the evaluation of routine urinalysis and urine sediment examinations as well. The time factor of the preanalytical processes significantly influences the number of urine components of the urine sediment.
Materials and methods: We compared the number of red blood cells, white blood cells and bacteria in samples stored on room temperature, taken by BDVacutainer® tubes containing urine stabilizing agents (AD), with closed sampling; and on the other hand in samples taken in non-sterile, additive-free plastic tubes (Native).
Results: According to our examinations RBC number significantly decreased after 24 and 48 hours both in the Native tubes and in the BDVacutainer® evacuated additive containing additive tubes as well. The decrease of WBC number was nearly significant in the Native tubes (P = 0.06; 0.18), while there were essentially no change in the BDVacutainer® tubes (P = 0.69; 0.80). The BACT number increased significantly only in the Nativesamples.
Conclusions: Our results indicate that preanalytical errors can be lessened with the use of BDVacutainer® tubes: the use of this device prevents the change of the number of WBC and BACT as well.
Key words: urine sediment; stabilization; preanalytical
P29
Validation of a pneumatic tube system. Quantification of acceleration forces related to haemolysis.
Gómez-Rioja R*, Fernández-Calle P, Alcaide MJ, Iturzaeta JM, Oliver P, Buño A
Laboratory Medicine Department, La Paz University Hospital, Madrid, Spain
Background: Pneumatic tube system (PTS) has been described as a cause of in vitro haemolysis. Streichert described a method to measure the total sum of sudden accelerations changes produced during the PTS transport using data-loggers sent along with samples. To validate our PTS, we investigate the quantitative relationship between sudden accelerations changes and their haemolytic effect in healthy volunteers’ samples.
Materials and methods: Five plasma samples (lithium-heparin, BD) in the same blood-drawn were obtained from 20 healthy volunteers. One sample was hand transported and considered as the baseline. The rest were sent through the PTS to the laboratory along with the data-logger (MSR 145) repeatedly. One sample was removed prior to each subsequent sending, there being four sendings in total. Once finished, all samples were centrifuged and duplicate measurements of Haemolytic index (H), potassium, AST and LD were performed in a Dimension Vista (Siemens HD). As an index of stress caused by PTS, the accumulative sum of sudden acceleration changes recorded for each sample were used.
Results: A statistically significant relationship was found between the accumulative sum of sudden acceleration changes and haemolysis degree, evidenced by haemolytic index (H) as well as by plasma increase of intraeritrocitic compounds, being respective Pearson coefficients r = 0.65 (H), r = 0.53 (potassium), r = 0.73 (LD) and r = 0.56 (AST). The critical threshold to establish a “safe transport “was defined using the slope of the regression equation. The criteria to define a significant change in the test concentration were established following the recommendations of Spanish Society of Clinical Chemistry. Safe threshold were set at 0.69 x 103 G (LD), 1.44 x 103 G (AST) and 5.08 x 103G (potassium).
Conclusion: These results confirm the reliability of Streichert method to validate a PTS. The relation between haemolysis degree and sudden acceleration changes allows the definition of a safe threshold for each test.
P30
The cost of poor sample quality: assessing the financial impact of sample rejection and recollection in healthcare institutions
Chait G1, Schlueter K2, Baginska E3, Scraba K3, Flynn L4, Church S*5
1 Whythawk, Oxford, UK
2 BD Diagnostics, Preanalytical Systems, Heidelberg, Germany
3 BD, Mississauga, Canada
4 BD Diagnostics, Preanalytical Systems, Franklin Lakes, USA
5 BD Diagnostics, Preanalytical Systems, Oxford, UK
Background: To model the financial impact of specimen rejection due to preanalytical errors, using institution specific data.
Materials and methods: As part of BD Laboratory Consulting Services™ a cost of poor quality model was developed to estimate the opportunity cost associated with poor sample quality. Critical data such as overall budget, number of beds, patients of different types/year, sample rejections/year, as well as qualitative data, eg. probable impact of sample rejection were collected by interviewing institution staff. The data were then entered into the model to calculate the possible financial impact of laboratory sample rejections. The model separates patients into different groups according to the likely effects of having a sample rejected, the probability that the rejection would have a low, medium or high impact and the consequences on institution time and resources. The overall consequence, or opportunity cost, was expressed as patient treatment time and financial cost.
Results: The size of the institutions (N = 10) ranged from 326 to 1,200 beds, with total operating costs between €41 million and €1.1 billion. The number of blood tests per month was between 37,000 and 458,000 and of these between 0.09% and 2% were rejected (mean 0.93%). The total cost of specimen rejection ranged from €22 K to €5.9 Million per annum (mean €1.9 Million), equating to a percentage of total operating costs from 0.1% to 1.2% (mean 0.4%). The estimated cost per patient for a sample rejection was from €29 to €349 (mean €171).
Conclusions: A customisable model which allows institutions to estimate, using local data and assumptions, the effect of sample rejection on patient treatment and cost. Results can be compared to those from other institutions to benchmark performance and assumptions. The results demonstrated that a reduction in the number of rejected samples due to preanalytical errors could lead to significant cost savings for most institutions.
P31
Using BD Laboratory Consulting Services™ to understand the impact of the preanalytical phase on sample quality and safety, a multi country perspective
Schlueter K*1, Church S2
1BD Diagnostics, Preanalytical Systems, Heidelberg, Germany
2BD Diagnostics, Preanalytical Systems, Oxford, UK
Background: The complexity of the preanalytical (PA) phase has precluded standardisation of PA processes, despite its impact on sample quality, laboratory efficiency, or patient & healthcare worker safety. The BD Laboratory Consulting Services™ Preanalytical Review audits PA procedures and practices in hospitals in different countries. Processes were assessed from storage of blood collection materials through specimen collection, transportation, processing of the samples and the resulting sample quality. By following the samples through the complete process, it was possible to link specific PA attributes to sample quality deficiencies.
Materials and methods: A consistent method and data collection form were used for audits (N = 48) of all blood collection systems. Data were collected by observation of the PA phase. Sample quality was assessed for laboratory samples of the same type.
Results: The PA phase was observed for 3597 blood collection tubes over 1350 collections. Sample quality was assessed for 8016 chemistry and 3532 coagulation tubes. For collections that resulted in hemolysed samples, 48% had prolonged use of tourniquet, 31% used catheters and for 38% the disinfectant was not allowed to dry. For serum samples with fibrin where the PA process had been observed, 26% had less than 30 minutes between collection and centrifugation and 81% had not been mixed. The following list gives the percentage of collections where a particular behaviour was observed, incorrect patient identification procedure, 56%; tubes labelled prior to collection 61%; coagulation tubes filled to less than 90% of tube volume 7%; gloves not worn 37%; incorrect activation of needle safety device 19%.
Conclusions: The BD Preanalytical Review standardised audit methodology allows comparison of results between departments and institutions. The prospective nature of the reviews permits identification of issues based on more data than from rejected samples alone and therefore affords a more complete understanding for those involved in the PA phase.
P32
How to mix blood collection tubes, a comparative evaluation of the GME Labsystems T-Swing automated tube mixer vs. manual mixing during phlebotomy
Ford K*, Church S
BD Diagnostics, Preanalytical Systems, Oxford, UK
Background: Mixing of blood collection tubes is an important process which if not competed correctly has the potential to induce platelet clumping, haemolysis, and other sample artifacts. Recommendations for mixing are poorly adhered to. The GME Labsystems T-Swing is an automated sample mixer. This study evaluated the T-Swing mixer for effects on sample quality & key analytes.
Materials and methods: The performance of the mixer was compared against recommended manual mixing by the phlebotomist at the time of sample collection. Samples were collected from apparently healthy individuals in the correct order (one discard tube followed by twelve BD Vacutainer® Blood collection tubes, Citrate (3), EDTA (3), BD SST™II (3), BD PST™II (3). One set was mixed by the phlebotomist according to the manufacturer’s recommendations and the other two sets using the T-Swing mixer. With one of these sets the tubes were mixed either 4 or 8 times (whichever was closest to Manufacturer’s recommendations) and all tubes in the third set were mixed 4 times. The appropriate samples were analysed for routine coagulation (PT & aPTT), haematology (FBC) and biochemistry tests. The results were compared for each sample set, along with visual observations of the samples.
Results: Equivalent performance was demonstrated for all visual observations and analyte comparisons. No haemolysis was observed in the serum (BD SST™II) or plasma (BD PST™II) tubes, analytes biases were for potassium (BD SST™II <0.5%, BD PST™II <3.0%) and LD (BD SST™II <1.0%, BD PST™II <7.0%). All EDTA tubes showed no platelet clumping (bias < 1.0% for 4 or 8 mixes) indicating effective mixing.
Conclusions: Equivalent performance was demonstrated between the automated methods and the manual mixing method, with no increase in the incidence of haemolysis or platelet clumping. The data suggest that four automated inversions is equivalent to recommended number of manual inversions.
P33
Accuracy evaluation of two blood glucose monitoring systems according to prediction-error grid analysis
Isgrò MA, Morelli R, Colacicco L, Zuppi C,Scribano D*
Dipartimento di Medicina di Laboratorio, Cattedra di Biochimica Clinica e Biologia Molecolare Clinica, U.O.C. Analisi I, Policlinico “A. Gemelli”, Roma, Italia
Background: The accuracy of blood glucose (BG) self monitoring systems is mandatory, as reliable results are prerequisite for therapeutic decisions.
Materials and methods: We evaluated two BG monitoring systems: Menarini StatStrip Xpress and Bayer Breeze®2 for system accuracy according to ISO 15197:2003, AACC/NACB guidelines and Clarke error grids. BG measurements of 74 patients were performed on both systems and then compared with results obtained using hexokinase reaction on ROCHE Cobas 8000.
Results: We compared glucose results using t-test, Mann-Whitney test, Passing-Bablok linear regression and Bland-Altman plots. T-test showed a statistical difference (P < 0.03) of Menarini BG systems: mean glucose = 126.6 mg/dL (range 78-301), SD = 37.6 mg/dL, CV% = 29.7 vs. laboratory values: mean glucose = 112.9 mg/dL (range 72-295), SD = 37.5 mg/dL, CV% = 33.2, while it wasn’t found any statistical difference (P < 0.12) for Bayer BG systems: mean glucose results = 117.4 mg/dL (range 71-293), DS = 36.6 mg/dL, CV% = 31.1. Mann–Whitney test further confirmed the statistical difference (P < 0.003) of Menarini BG vs. laboratory values. Passing-Bablok linear regression and Bland-Altman plots showed that both BG monitoring systems fulfilled ISO 15197 requirements (the percentage of results showing the minimum acceptable accuracy was 100.0% for either StatStrip Xpress or Breeze®2). Nevertheless, only Breeze®2 meets the AACC/NACB guidelines, but not StatStrip Xpress (95.9% vs. 86.5% of results showed the minimum acceptable accuracy, respectively). According to Clarke error grids, both StatStrip Xpress and Breeze®2 had 100.0% of their results in error zones A and B, considered as “clinically uncritical”.
Conclusions Both BG systems fulfilled the minimal accuracy ISO standards’ requirements and didn’t present clinically critical errors, but only Breeze®2 met the more stringent AACC/NACB guidelines. Because inaccurate systems bear the risk of false therapeutic decisions and subsequent possible severe health injury, regular and standardized evaluation of BG meters and test strips should be performed by manufacturers in order to ensure adherence to more recent guidelines.
Key words: blood glucose monitoring systems; accuracy evaluation; prediction-error grid analysis
P34
Lipemia as nonconformity of blood donor samples
Miletić Lovrić M*, Vuk T, Stojić Vidović M, Mihaljević I
Croatian Institute of Transfusion Medicine (CITM), Zagreb, Croatia
Background: Mandatory screening of blood donors for transfusion-transmitted diseases in Croatia includes HBsAg, anti-HCV and HCV Ag, HIV Ag/Ab and antibodies to Treponema pallidum (TP). For that purpose we use the most sensitive and specific immunoassays. The part of preanalytical phase of testing is also organoleptic examination of the samples for pronounced lipemia, hemolysis and icterus. According to the most manufactures of serology immunoassays there is potential interference from elevated levels of triglycerides (> 34 mmol/L), hemoglobin (> 5 g/L) and bilirubin (> 342 µmol/L) and it is usually 10-20% difference in the test unit (ratio = optical density/cut-off). And also there is possibility of invalid test result. We evaluate these samples as unsuitable or nonconforming due to our guidelines for management of nonconforming samples.
Materials and methods: In the period between January 1st and October 31, 2012 we have collected 84,512 blood donations and so much blood samples we organoleptic examined for lipemia, hemolysis and icterus, and tested for mentioned markers: HBsAg, anti-HCV and HCV Ag, HIV Ag/Ab and anti-TP, used Abbott Prism and Architect tests and Bio-Rad Ultra tests.
Results: 20 of 84,512 samples we evaluated as pronounced lipemic (0.024%) and only 1 as pronounced hemolytic (0.0012%).
Conclusion: The most common cause of nonconformities in our Institute are lipemic samples, compare to other nonconformities together: incomplete clotting, hemolysis, labeling errors, icterus, inadequate quantity and others. There is notable reduction of samples nonconfirmties, primarily because of reduced frequency of lipemic samples. That is due to systemic education of blood donors on the favorable pre-donation dietery habits which has been performed for many years in our Institute.
Key words: nonconformity; lipemic samples; blood donors
P35
Effect of hemolysis on routine coagulation assays
Milevoj Kopcinovic L*, Bronic A, Hreljac Sevcenko M, Pavic M
University Department of Chemistry, University Hospital of Traumatology, Medical School University Hospital Sestre Milosrdnice, Zagreb, Croatia
Background: Hemolysis might affect the reliability of laboratory testing and is reported to be the leading source of unsuitable specimens. Our aim was to investigate the effect of in vitro hemolysis on routine coagulation parameters.
Materials and methods: We studied 50 pairs of hemolyzed and subsequently re-collected non-hemolyzed plasma samples within individual patients. The time interval between the two collections was recorded. Hemolysis and approximate hemoglobin concentration were assessed by visual inspection, comparing samples against a standard color scale. According to hemoglobin concentration, hemolyzed samples were divided into 4 subgroups: slightly (0.5 g/L), mildly (1 g/L), moderately (2 g/L) and severely (> 3.0 g/L). PT, APTT and fibrinogen were determined in all samples immediately after collection, using an automated analyzer. Differences between hemolyzed and corresponding non-hemolyzed specimens were evaluated. The effect of hemolysis was expressed as interfering bias and compared to analytical quality specifications and reference change values to define analytically and clinically relevant variations.
Results: Statistically significant overestimation of PT values was observed in all investigated subgroups. APTT was significantly shortened only in the severely hemolyzed subgroup (23.9 (22.0-27.8) s vs. 26.5 (24.0-29.1) s, P = 0.002). Likewise, fibrinogen concentrations were overestimated in the severe hemolysis subgroup (4.5 (3.7–5.5) g/L vs. 4.4 (3.5-5.5) g/L, P = 0.005). The mean interfering biases were heterogeneous for all assayed parameters, independent of hemolysis degree. When compared to desirable specifications, investigated parameters showed no analytically or clinically relevant variation.
Conclusions: Although the differences between hemolyzed and non hemolyzed specimens observed in this study were not analytically nor clinically relevant, they were heterogeneous and unpredictable, which emphasizes the need to thoroughly investigate this issue and eventually revise, standardize and harmonize procedures for management of unsuitable samples in coagulation testing.
Key words: hemolysis; coagulation testing; unsuitable specimens
P36
Design of a high quality and controlled process for a centralized phlebotomy unit
Kot A, de Jonge Niels*
Centraal Klinisch Chemisch Laboratorium (CKCL), Leiden University Medical Center, Leiden, The Netherlands
Background: The error rate in the preanalytical phase is still relatively high. Main root causes are an insufficient control mechanism for quality and the presence of waste. In our hospital, phlebotomy used to be a decentralised responsibility, i.e. each ward and outpatient clinic was responsible for the phlebotomy in their patients. Recently, it was decided to bring phlebotomy under the responsibility of the laboratory and to centralise phlebotomy for outpatients in a new facility. Our objective was to design a high quality and controlled phlebotomy process. Total turnaround time should be minimal in order to facilitate clinical pathways.
Materials and methods: Our hospital is a medium-sized academic hospital, performing over 130.000 phlebotomy procedures for outpatients (approximately 500-700 patients per day). We applied the Delft Systems Approach operational research to analyse the current situation and to redesign the process. Lean-tools were used to eliminate waste and to increase the productivity. Finally, an adequate capacity was calculated to guarantee a high service level and minimize waiting times, using Kendall’s Queueing Theory.
Results: The process redesign for the new phlebotomy process includes a pragmatic model to meet the defined requirements for high patient safety, patient satisfaction and productivity. With the elimination of waste a 30% reduction of labour can be gained. The calculation of the capacity was validated for one of the current facilities. The capacity was calculated to guarantee a maximal average waiting time of 4 minutes for at least 90% of the patients. The overall average waiting time is estimated to be 30 seconds and total average turnaround time less than 7 minutes
Conclusions: The Delft Systems Approach is a valuable methodology for both process analysis and redesign. Kendall’s Queueing Theory was validated as a tool to calculate the capacity and forecast waiting and service times in a phlebotomy facility.
P37
A case report of EDTA contamination and high potassium result in non-hemolyzed serum
Banys V
Vilnius University Faculty of Medicine, Department of Physiology, Biochemistry, Microbiology and Laboratory Medicine, Vilnius, Lithuania
Background: EDTA contamination of serum samples is one of pre-analytical errors occurring in daily work. EDTA may be present in serum due to improper order of tubes (EDTA tube being drawn before serum tube – possibility of backflow of EDTA into the collection device and carryover of additive into serum tube) or manual transfusion of EDTA blood into serum tube. EDTA causes hypocalcemia, hypomagnesiemia (due to chelation) and subsequently decreased activity of alkaline phosphatase. A case of high potassium value in non-hemolyzed serum due to EDTA contamination is reported.
Materials and methods: 61 years old male presented to hospital with hip fracture, serum potassium 10.44 mmol/L, normal sodium and chloride results, and calcium result below detection limit of the method. Suspecting pre-analytical error two subsequent samples were drawn in 2 and 4 hours. All samples were retested for potassium, sodium and chloride using three different methods: Beckman Coulter Synchron CX9 (USA), Abbott Architect ci8200 (USA), Radiometer ABL800 (Denmark). Total calcium concentration and alkaline phosphatase activity were measured by first two methods. Hospital administration organized audit on work of phlebotomists.
Results: Mean of first sample (elevated potassium) was 10.31 mmol/L (SD = 0.27, N = 3). Means of other samples (normal potassium) were 5.14 mmol/L (SD = 0.07, N = 3) and 4.85 mmol/L (SD = 0.04, N = 3) respectively. Alkaline phosphatase activity (58 vs. 87 IU/L Architect; 43 vs. 69 IU/L Synchron) and total calcium concentration (0.63 vs. 1.92 mmol/L Architect; 1.11 vs. 2.09 mmol/L Synchron) of first sample was decreased significantly comparing to other samples. Sodium and chloride results were uniform and normal. Audit revealed a nurse who manually transferred part of EDTA blood to serum tube.
Conclusions: Irresponsible behavior of nurses may produce erroneous biochemical data, lead to incorrect diagnosis and unnecessary treatment. Education of medical and nursing staff regarding proper blood collection techniques is needed to prevent sample contamination.
Key words: EDTA; potassium; contamination
P38
Performance of aprotinin tubes for routine ACTH assay in the Hospital Unit with manual transport system
Pollak J*, Senterkiewicz L, Sypniewska G
Department of Laboratory Medicine, Collegium Medicum in Bydgoszcz, Nicolaus Copernicus University in Torun
Background: Preanalytical factors such as proteolytic degradation can affect reliability of hormone assay results. Adrenocorticotropic hormone (ACTH) in blood is considered highly unstable. Manual transport system of the tubes, especially at night, led to discordant ACTH results between night and morning collection. The aim of the study was to investigate the effect of different anticoagulants on the stability of ACTH in test tubes transported with the use of manual transport system.
Materials and methods: Blood was collected from 26 patients twice: at 8 am and 11 pm into 2 different types of EDTA-containing tubes, with or without the presence of aprotinin. Both tubes from each patient were brought on ice to the laboratory within 1 hour, centrifuged and assayed immediately using the Roche-ACTH-electrochemiluminescence immunoassay on fully automated Cobas e411 analyser.
Results: ACTH concentrations were: 1–731 pg/mL in EDTA tubes and 1.61-992 pg/mL in EDTA-aprotinine tubes taken at 8 am; 1-60 pg/mL in EDTA tubes and 1-101 pg/mL in EDTA-aprotinin tubes taken at 11 pm. Results were compared between two different types of tubes and expressed as mean percentage of difference. Results showed that plasma ACTH concentration in EDTA-aprotinin tubes was always higher than in EDTA-tubes, respectively: 15% in samples collected at 8 am (P < 0.05) and 9.35% in samples collected at 11 pm (P < 0.05).
Conclusions: The addition of aprotinin to EDTA-tubes showed an improvement of ACTH stability in the blood that is of essential importance in the hospitals without pneumatic tube transport system.
Key words: ACTH; aprotinin; hormone stability
P39
The effect of standardization of sampling procedure to the proportion of haemolytic and clotted samples
Reimann K*, Osi K
Central Laboratory, East Tallinn Central Hospital, Tallinn, Estonia
Background: Clotting and haemolysis are the most frequent problems in blood sampling. In the East-Tallinn Central Hospital (ETCH), standardization of sampling procedure was carried out in December 2009 in attempt to reduce the number of unusable samples. New sampling sets were implemented and training for the whole hospital nursing staff was performed. Current overview evaluates the effect of these standardization measures on the proportion of clotted and haemolytic samples.
Materials and methods: Proportion of samples, considered clotted or haemolytic by visual inspection, was calculated for hospital’s surgery, internal diseases’, and women’s clinics in 2009, 2010 and 2011. Statistical significance of the yearly differences was assessed using Chi-square test.
Results: The total proportion of clotted samples decreased from 0.04% in 2009 to 0.03% in 2010 (P = 0.004); and this reduction was maintained in 2011. The total proportion of haemolytic samples was reduced from 2009 to 2010 (respectively 0.19 and 0.11%, P < 0.001); the reduction continued in 2011, when the proportion of haemolytic samples was 0.09%. Regarding yearly changes in the clinics, the proportion of clotted and haemolytic samples decreased statistically significantly in all clinics between 2009 and 2010 with the exception of clotted samples in women’s clinic where the reduction was achieved only in 2011. In women’s and internal diseases’ clinics the reduction in proportion of haemolytic samples continued in 2011. The reduction, achieved in 2010 in proportions of haemolytic samples in surgery clinic, was maintained in 2011. The proportion of clotted samples in surgery and internal diseases’ clinics in 2011, however, returned to the level of 2009.
Conclusions: The standardization of blood sampling in the ETCH was associated with reduction in proportions in both clotted and haemolytic samples. Yet for maintaining the reduction, repeated staff trainings may be necessary.
Key words: blood sampling; standardization; clotting; haemolysis
P40
Preparing the patient as an imperative for obtaining adequate sample for qualitative urinalysis
Sikirica M*, Radišić Biljak V, Flegar-Meštrić Z
Department of Medical Biochemistry and Laboratory Medicine, University Hospital Merkur, Zagreb, Croatia
Background: Urine is an analytical matrix that provides information on numerous physiological processes. It is accessible and it can be collected noninvasively what is of great importance. Urinalysis is one of the most frequently requested analyses at all levels of health care. Providing the adequate preanalytical conditions for obtaining an appropriate sample is the basic precondition for reliable results and appropriate clinical interpretation afterwards. It can be obtained in a form of oral or written recommendations.
Materials and methods: Anonymous survey in a form of questionnaire was conducted in 78 randomly selected patients who were instructed to the Department of Medical Biochemistry and Laboratory Medicine for laboratory analysis. Our aim was to gain insight into the awareness of patients regarding preanalytical conditions affecting urine sampling. The main question was whether patients received any kind (oral or written) of recommendations how to prepare for giving urine sample either form the physician, nurse or laboratory staff.
Results: Only 13/78 patients received partial verbal instructions (5 from the doctor, 8 from the nurse) prior urine sampling, while 65/78 patients had never received any form of recommendations for administering the single sample of urine in the practice of primary care. However, almost 80% of selected patients (59/78) received written and/or oral instructions when they arrived in their primary laboratory.
Conclusion: These results indicate that patients are poorly instructed for urine sampling in their primary care office, while these instructions are available in most primary laboratories. Informing patients is one of the basic conditions of obtaining a good sample. International recommendations a continuous education of health teams leads to overall reduction of preanalytical errors. The effectiveness of these basic conditions is defined by the quality indicators that objectify all steps in the collection, transport and storage of samples.
P41
The preanalytical phase starts before blood collection
Attard R*, Farrugia R, Bezzina Wettinger S
Department of Applied Biomedical Science, Faculty of Health Sciences, University of Malta Msida, Malta
Background: Large biological collections are being set up for biobanking, epidemiological studies and in the search of biomarkers. A number of variables could play an important role in determining levels of analytes in blood even before blood is collected and can result in contradictory outcomes of studies, particularly when levels are compared between groups of individuals. There is a lack of guidelines and standardised instructions that need to be followed prior to phlebotomy.
Materials and methods: A literature search was conducted to determine the variables that can influence results prior to blood collection in a research setting and to draft instructions for participants to follow in the week preceding blood collection.
Results: In addition to fasting, several variables are reported to influence levels of analytes in blood. Depending on the scope of the collection it may be necessary to adopt measures to limit the effect of these variables. These may include: (a) simple recording of the variable (e.g. use of medications, day of menstrual cycle, pregnancy, time and date of phlebotomy since season may also influence variables, time of last food intake, perceived psychological stress levels); (b) postponement of blood draw (e.g. for medical and dental interventions, infection and vaccination, minor injuries in the previous week, blood transfusions and donation); (c) giving instructions to the research subject (e.g. to fast, avoid alcohol, tea, coffee and milk, avoid smoking 12 hours before blood draw and if not possible record time of last cigarette, maintain normal physical activity and dietary habits, avoid strenuous exercise, avoid use of chewing gum before phlebotomy, determine instructions on use of vitamins and supplements, drink water to avoid dehydration, empty bladder before blood draw). Clear explanations which may include a leaflet must be given to the research subject.
Conclusion: There are several factors influencing blood analytes even prior to blood draw and they need to be given considerable attention.
Key words: instructions; prior to blood draw; research; biobanking
P42
The analysis of performance by medical nurses of blood sample collection procedure for laboratory tests
Kovalevskaya SN*, Khorovskaya LA, Petrova NG
State Medical University n. a. I. P. Pavlov, St.-Petersburg, Russian Federation
Background: Three distinct steps can be identified in the measurement procedure: preanalytical, analytical and postanalytical. It has been reported that a majority of errors (60-70%) occurs during the preanalytical step, which is heavily depending on human factors. Phlebotomy is a most important part of the preanalytical step. In the Russian Federation blood samples are collected by medical nurses employed by clinical divisions of the Health Care System. The aim of the present study was to analyze skills and knowledge of the nurses.
Materials and methods: A questionnaire comprising 24 questions was developed and distributed among 153 nurses in State and private outpatient clinics and hospitals in 3 cities in different regions of the Russian Federation; St.-Petersburg (North-West Region), Khanty-Mansiysk and Noyabrsk (Siberia). The questionnaire addressed the demographics of respondents, their professional experience and particularly phlebotomy training using evacuated blood collection systems (EBCS), dealing with difficulties encountered during phlebotomy, transport/transfer of blood samples and safety issues/information support. Answers were de-identified and then evaluated using general statistical methods.
Results: The investigation revealed that about 20% of nursing staff was not trained in EBCS systems. Of those that were trained 55% attended courses given by manufactures of EBCS and 45% were trained by their colleagues. Half of respondents (50%) had difficulties to answer to the questions concerning method for obtaining samples with EBCS. 86% of phlebotomists answered questions about rules for application of a tourniquet incorrectly and 75% were unaware of the volumes of blood required for different types of evacuated tubes. More than 50% of the respondents reported difficulties during venipuncture. Most commonly are: causes to recollect of sample collection (80%), difficulties to locate a vein (40%), hazard of injury by used or unused equipment (28%), venipuncture of neonates and children less than one year old (10%) were encountered. The most frequent reasons for duplication of sample collection were: hemolysis of samples (80%), errors in sample labeling (16%) and broken evacuated tubes (4%). The survey revealed that 7% of nurses refused to use EBCS and preferred to work with traditional methods.
Conclusion: The study indicates that there is an insufficient training of phlebotomists in using EBCS. This creates a potential risk for patients and needs attention in formulating the curriculum for medical nurses within the Russian Federation Health Care System.
Key words: phlebotomy; preanalytical stage; medical nurses
P43
Frequencies of hemolysis and lipemia in whole blood samples in pediatric population
Grzunov A*, Jagetic R, Lenicek Krleza J
Children’s Hospital Zagreb, Department of Laboratory Diagnostics, Zagreb, Croatia
Introduction: The aim of this study was to assess prevalence of interference from hemolysis and lipemia in whole blood samples in pediatric population.
Materials and methods: 3903 consecutive samples collected during one month period were visually monitored for turbidity and hemolysis and patients age were noted. The statistical significance of differences between relative frequencies of interferences was calculated with χ2test.
Results: 7% of all samples were identified to have hemolysis. Hemolysis in samples drawn from vein occurred more frequently than hemolysis in capillary samples (P < 0.001). 18% of samples were identified to have lipemia. Both lipemia and hemolysis were observed in < 1% of samples. Interferences were more frequent in population up to 2 years old (P < 0.001).
Conclusion: Due to presence of interferences in significant number of samples, particularly in population up to 2 years old, a considerable number of whole blood samples may be unreliable. Continuous education and training of healthcare personnel in phlebotomy and patient preparation prior to sample collection is needed.
Key words: whole blood sample; hemolysis; lipemia; interference; child
P44
National guideline for phlebotomy in the Netherlands
Eppens E1, Roelofs-Thijssen M2, de Jonge N3, van Dongen-Lases E*4
1Department of Clinical Chemistry, Gelderse Vallei Hospital, Ede, the Netherlands
2Department of Clinical Chemistry, Rijnstate Hospital, Arnhem, the Netherlands
3Department of Clinical Chemistry,Leiden University Medical Center, Leiden, the Netherlands
4Department of Clinical Chemistry, Academic Medical Center, Amsterdam, the Netherlands
Background: The Working Group Preanalytical Phase of the Netherlands Society for Clinical Chemistry and Laboratory Medicine addresses all elements of the preanalytical phase of clinical chemistry laboratory testing in the Netherlands. One of its major tasks is to develop national guidelines for the preanalytical phase. Therefore, the Dutch Working Group Preanalytical Phase recently developed the national guideline for phlebotomy. The aim of this guideline was to summarize the best practices in phlebotomy in order to improve the safety of patients and phlebotomists and the quality of laboratory testing.
Materials and methods: Information on phlebotomy was collected and summarized by the Dutch Working Group Preanalytical Phase. Information was gathered from many resources such as national legislation e.g. the Individual Healthcare Professions Act, national guidelines e.g. guidelines concerning hygiene from the Dutch Working Group Infection Prevention, international guidelines e.g. guidelines from the Clinical and Laboratory Standards Institute and literature. In April 2013, the national guideline for phlebotomy will be approved by the Netherlands Society for Clinical Chemistry and Laboratory Medicine.
Results: The national guideline for phlebotomy in the Netherlands has been divided in chapters and subdivided in sections, each presenting background information on the specific aspects of phlebotomy, e.g. factors that affect laboratory values, facilities, supplies, venepuncture procedure, venepuncture in children and difficult collections, competence of phlebotomists, hygiene and safety. Each section ends with minimum standards and/or recommendations.
Conclusions: The national guideline for phlebotomy in the Netherlands describes the best practices in phlebotomy. Compliance with this guideline will improve the safety of patients and phlebotomists and the quality of laboratory testing.
Key words: guideline; national; phlebotomy
P45
Reference ranges for coagulation screening tests and coagulation inhibitors in newborn’s cord blood on ACL TOP 500
Trampuš Bakija A*1, Benedik Dolničar M2, Mencin D1, Prelec A3, Kitanovski L2
1UMC Ljubljana, University Children’s Hospital, Unit of Special Laboratory Diagnostics, Ljubljana, Slovenia
2UMC Ljubljana, University Children’s Hospital, Department of Oncology and Hematology, Ljubljana, Slovenia
3UMC Ljubljana, Division of Gynecology and Obstetric, Department of Perinatology, Ljubljana, Slovenia
Background: The concentration and function of coagulation factors and natural inhibitors depend on gestational and postnatal age of the infant. The concept of developmental haemostasis is broadly accepted and age related reference values should be concerned in interpretation of laboratory results. It is also recommended that absolute values of reference ranges for coagulation assays should be established for reagent/analyzer system. Our study was designed to determine reference ranges for newborns for activated partial thromboplastin time (APTT), prothrombin time (PT), thrombin time (TT), fibrinogen, antithrombin (AT), protein C activity and free protein S.
Materials and methods: Plasma samples were obtained from umbilical vein from 50 healthy full term neonates born in University Medical Centre Ljubljana. All tests were performed on ACL TOP 500 coagulation analyzer (Instrumentation Laboratory) using reagents from Instrumentation Laboratory. Reference ranges and t-test for statistical differences were calculated using MedCalc software.
Results: Reference values for tests are: APTT: 32 s (23-41), PT [s]: 12.2 s (10.0–14.3), PT [%]: 87% (62–109), TT: 18 s (15–21), fibrinogen: 1.72 g/L (1.1–2.4), AT: 61% (37-85), protein C activity: 36% (20-52) and free protein S: 49% (28-71). Values for all tests are significantly different (P < 0.001) from those for adults or manufacturer. Compared to scarce published literature, the directions of age-related changes are consistent, but the magnitude is specific for reagent/analyzer combination.
Conclusions: Our data confirm previously reported influence of lower thrombin generation potential and age related changes of inhibitors in neonates that results in prolongation of screening tests and marked reduction of inhibitor activity. Adult reference ranges for coagulation screening tests, especially PT and APTT and inhibitor activity tests cannot be applied to newborns. Age related test specific reference values are needed for accurate diagnosis and management of children with suspected bleeding disorder.
P46
Troponin T high sensitivity assay: serum or lithium heparin as specimen type?
Mulliez S, Stove V, Delanghe J*
Department of Clinical chemistry, microbiology and immunology, Ghent University Hospital, Belgium
Background: The manufacturer of the TroponinT high sensitivity (TnT hs) assay states in its reagent leaflet that serum or K2-EDTA, Li-heparin and Na-heparin plasma are acceptable specimen types for analysis. However, heparin has been described to bind some isoforms of TroponinT, especially the forms distributed in circulation in early phases o
References
P47
Transportation of blood samples through adjacent phlebotomist cabins to the laboratory by a simple conveyer belt reduces turn around time for the samples and waiting time for the patients and may make recommended hand inversions of samples with additives superfluous
Stender S*, Mandir IA
Department of Clinical Biochemistry, Copenhagen University Hospital, Gentofte, Denmark
Background: In our out-patient clinic blood samples are drawn by phlebotomists in 5 separate but adjacent cabins placed as an elongation of the laboratory. In order to reduce turn around time (TAT) we installed a conveyer belt and the phlebotomists placed the samples on the belt during or immediately after each phlebotomy (~4500/month).
Materials and methods: The belt runs 18 m with a constant speed (15 cm/second) below the windowsill through all 5 cabins and into the laboratory and delivers the blood samples close to the centrifugation unit. During passage of a fire wall the samples are slightly tilted and for the last 6 m within the laboratory, the belt has a slight upwards slope. The samples make a 180 degree inversion when they leave the belt and end in a tray.
Results: We observed a reduction from 17 to 14 minutes from the hematology samples were drawn until the results were available, when comparing the median values obtained for months before with values obtained for months after installation of the belt. We also observed a reduction from 15 to 10 minutes in median waiting time for the patients from they arrived to the out-patient clinic until they entered the phlebotomist cabins. In contrast to the procedures before the conveyer belt, now the phlebotomists do not have to leave the cabins except for their planned intermissions. We found the same number of thrombocytes in EDTA-coated tubes with blood drawn from 507 consecutive patients placed on the belt without any inversions compared to samples drawn from the same venipuncture in similar tubes, inverted 8-10 times as recommended in various guidelines before they were placed on the belt.
Conclusion: The conveyer belt in our setting reduces TAT, increases the number of phlebotomies / hour / phlebotomist and may make recommended inversions of blood tubes with additives superfluous.
Key words: conveyer belt; phlebotomy; TAT; inversion of blood tubes
P48
Blood samples with additives. Stirred - not shaken. Is that enough for sufficient mixing?
Stender S
Department of Clinical Biochemistry, Copenhagen University Hospital, Gentofte, Denmark
Background: Guidelines recommend gentle mixing i.e. 4-10 times inversions of blood samples with additives immediately after phlebotomy. We wanted to find the minimum number of inversions that ensures a sufficient mixing of the blood with the additives.
Materials and methods: Two samples of blood were drawn in 2 identical EDTA-tubes from the same venipuncture from consecutive patients visiting the hospitals out-patient clinic for phlebotomy. One EDTA-tube was inverted 8-10 times as recommended in the guidelines immediately after the venipuncture, whereas the experimental tube was gently placed in a rack in room temperature and remained there for at least half an hour. Both EDTA-tubes were analyzed on the same equipment. The same type of experiments was conducted with samples collected from bedridden patients by phlebotomists at their rounds in the various clinical departments. In the first experiment the experimental tubes received only the mixing that occurred during the walking from the bed to the laboratory. In the second experiment the experimental tubes were inverted only once immediately after phlebotomy. Based on double analysis of the number of thrombocytes in the same EDTA-tube, a more than 20 % drop in thenumber of thrombocytes in the experimental EDTA-tube, was taken as a sign of coagulation due to insufficient mixing of blood and EDTA.
Results: Based on 604EDTA-tubes placed in a rack immediately after phlebotomy without any inversions for half an hour, 2.8 % were insufficiently mixed. Based on 1009 EDTA-tubes without any planned inversions after phlebotomy of bedridden patients, 0.6% was insufficiently mixed, while 1032 EDTA-tubes inverted only once, a maximum of 0.4% were insufficiently mixed.
Conclusion: Insufficient mixing of blood and EDTA in an EDTA-tube is nowadays a rare event even without any inversion of the tube. Guidelines presently recommend up to 10 inversions although fewer may be sufficient.
Key words: guidelines; EDTA-tubes; inversion of blood tubes; phlebotomy; sufficient mixing
P49
Risk of hemolysis in the Emergency Department depending on phlebotomy practices
Prieto Alvarez N*1, Gómez-Rioja R2, Alcaide Martín MJ2, Barba Yegro MR1, Fernandez–Calle P2, Torres Hidalgo A1
1Emergency Department, La Paz University Hospital, Madrid, Spain
2Emergency Laboratory, La Paz University Hospital, Madrid, Spain
Background: In the University Hospital of “La Paz” Emergency Department, the prevalence of hemolyzed specimens is estimated around 14.4%. This rate of hemolysis is higher as compared with other services and can be related to phlebotomy practices, mainly because of the frequent use of catheter. The objective of this study was to analyze the relation between current practices of phlebotomy in our department and the quality of the samples.
Materials and methods: For one month, six nurses from the Emergency department registered the collection procedures used in every patients, categorized as: straight needle + vacutainer, butterfly needle + vacutainer, straight needle + syringe, catheter and arterial blood collection instead of phlebotomy. Hemolysis was quantified as serum index in a Dimension Vista (Siemens HD) and considered as significant by the laboratory.
Results: 435 samples were included. Catheter-drawn was the most frequently used (55% of patients), followed by Butterfly needle (20%), straight needle + vacutainer (13%), straight needle + syringe (10%) and arterial-drawn (2%). Significant differences were observed in the prevalence of hemolyzed samples depending on phlebotomy practice. 50% of the arterial samples, 30.2% of syringe-drawn samples, 29.5% of the catheter-drawn samples, 12.6% of butterfly needle and 10.3% of the straight needle + vacutainer were severely hemolyzed. Risk of hemolisys compared to straight needle + vacutainer drawn was calculated using logistic regression. No significant differences were observed between the straight needle + vacutainer and butterfly needle + vacutainer drawn. In the other hand, the use of catheter, syringe or arterial drawn imply an increasing risk (Odds ratio) of 3.6, 3.7 and 8.6 respectively (P < 0.05).
Conclusions: We find a clear relation between hemolysis and phlebotomy practices. Catheter-drawn or syringe use, involving blood transfer to the tubes, suppose a 3.7-fold increase in the risk of hemolysis. These practices are used in 65% of patients in our Emergency Department. To avoid the high prevalence of hemolysis in the Service, it would be necessary to change phlebotomy practices.
Key words: hemolysis; emergency medicine; phlebotomy; catheter
P50
Implementing safe sharps practices in Europe, what do I need to do?
De Carli G*1, Abiteboul D2, Bouvet E2, De Schryver A3, Mazón Cuadrado L4, Ncube F5, Puro V1, Wittmann A6, Ippolito G1 for the Sharps Safety in the European Union Group
1SIROH, National Institute for Infectious Diseases L. Spallanzani, Rome, Italy
2GERES, Faculté de Médecine X, BICHAT, Paris, France
3IDEWE-Universiteit Antwerpen, Leuven, Belgium
4Servicio De Prevención Hospital De Fuenlabrada, Madrid, Spain
5Bloodborne Viruses Section, National Infectious Disease Surveillance Centre, Health Protection Agency, London, UK
6Technischer Infektionsschutz, Bergische Universitaet Wuppertal, Wuppertal, Germany
Background: The Directive 2010/32/EU “Prevention from sharp injuries in the hospital and healthcare sector” was issued to protect healthcare workers from the risk of needlestick/sharps injuries and subsequent infection with bloodborne pathogens, and should be adopted within May 2013.
Materials and methods: The Sharps Safety in the European Union Group gathered experts to develop a consensus statement and practical recommendations on the main preventive issues included in the Directive to help in this transition.
Results: Consensus statement: Raise awareness among workers to create a ‘need for safety’ and overcome resistance to change procedures and devices. Implement an active surveillance of needlestick/sharps injuries and provide feedback to workers. Incorporate Standard Precautions and Safe Injection Practices within education to be delivered at employment and regularly thereafter, particularly when changing technologies or procedures. Link prevention of needlestick/sharps injuries and bloodborne infections through a safe and appropriate use of sharps/needle devices to prevention of healthcare associated infections, to be included in national health plans. Explain rationale and procedures for pre- and post-exposure management, including Hepatitis B vaccination and antiretroviral prophylaxis, which must be always available. Organize work to decrease use of unnecessary needles: concentrate blood tests, switch parenteral therapies to oral as soon as safe for patients, re-evaluate the need for peripheral catheters daily. Place sharps containers where procedures are performed; monitor their use and contents to discourage recapping. Perform task analysis for every procedure involving sharps, especially in surgery. Prioritize procedures involving access to vein/artery in implementing safety-engineered devices incorporating protection mechanisms, followed by all other instances in which an injury may occur regardless of the setting.
Conclusions: Experts must have a proactive role in disseminating the contents and suggesting the best approach to implement preventive measures in Directive 2010/32/EU, helping employers and workers to optimize use of resources.
Key words: prevention; bloodborne infections; occupational injury
P51
Handling of pseudohyponatremia in the routine laboratory workflow
Siska A*, Seres E, Nagy A
University of Szeged, Department of Laboratory Medicine, Szeged, Hungary
Background: When sodium is measured by some laboratory methods falsely low sodium values can be measured in cases of extreme hyperlipidemia or hyperproteinemia. This phenomenon is called pseudohyponatremia. Our aim was to determine the degree of the lipemia and hyperproteinemia where the direct potetiometry provides more adequate sodium concentration than the indirect one.
Materials and methods: The direct potentiometry equipment was an AVL 9180 and the indirect one was a Roche Modular ISE 900. The lipemia index is automatically measured in our laboratory due to the total laboratory automation.
Results: In the examined range of both lipemia index and triglyceride concentration there was no significant difference between the sodium levels, measured by the above mentioned methods. However, a weak relationship could be observed between triglyceride concentration and lipemia index in patient samples. In the case of hyperproteinemia significant difference was found between sodium concentrations obtained by the two types of electrode measurement. Above 100 g/L of total protein level the sodium measurement without sample dilution (i.e. direct potentiometry) provided clinically more relevant result than the methods with sample dilution (indirect potentiometry).
Conclusion: In accordance with these results the value of sodium in samples with lipemia index below 2000 is acceptable. Above this value a comment should be attached about the possibility of pseudohyponatremia in the laboratory findings. In case of total protein level higher than 100 g/L the sodium concentration should be determined by direct potentiometry and the clinicians have to be notified about the used methodology.
Key words: pseudohyponatremia; lipemia; hyperproteinemia; potentiometry
P52
Case report: HAMA interference in determinating CEA concentration in non-Hodgkin`s lymphoma patient
Stancin N1, Derek L*2, Unic A2, Romic Z2, Banek T1, Ajdukovic Stojisavljevic R3
1Center of Nuclear Medicine, Dubrava University Hospital, Zagreb, Croatia
2Clinical Department for Laboratory Diagnostics, Dubrava University Hospital, Zagreb, Croatia
3Department of Hematology, Dubrava University Hospital, Zagreb, Croatia
Background: A 47 year old female was treated for non-Hodgkin`s lymphoma CS IV B/B-LCL with immunochemotherapy (R-EPOCH protocol) from 2008-2009. After completing the treatments, follow-up exams included the determination of carcinoembryonic antigen (CEA) concentration. From 2009-2010 CEA concentration repeatedly increased, but using imaging techniques (PET-CT), relapse of the disease was not detected. Due to the suspicion of the possible human anti-mouse antibodies (HAMA) interference, CEA was determined using different immunoassay.
Materials and methods: CEA was determined using chemiluminiscence assay on Vitros ECiQ (Ortho Clinical Diagnostics, Johnson and Johnson, Buckinghamshire, UK), and electrochemiluminiscence assay on Cobas e411 (Hitachi High Technologies Corporation, Tokyo, Japan). Besides the retesting of CEA on different analyzers, dilution test was also performed.
Results: Patient CEA concentrations, determined on Vitros ECiQ vs. Cobas e411 were as follows (3.5 vs. 0.9 µg/L on May 4, 2009; 43.0 µg/L vs. 0.2 µg/L on November 18, 2009; 47.4 µg/L vs. 0.9 µg/L on November 23, 2009 and 60.2 µg/L vs. 1.1 µg/L on May 24, 2010). Dilution test with 1:2, 1:5 and 1:10 sample dilutions was also performed on Vitros ECiQ, and CEA concentrations were not linear (20.6 µg/L, 6.3 µg/L and 0.5 µg/L respectively).
Conclusions: Increased CEA concentrations on Vitros EciQ that were not in accordance with the clinical status, the deviation from linearity in the dilution test, as well as a significant difference in the concentration of CEA on Cobas e411, point to HAMA interference that was probably a consequence of a immunochemotherapy. Noncompetitive methods that use murine antibodies are most susceptible to interference because of the mice immunoglobulins commonly used in diagnostic and therapeutic procedures. Cases like this can result in a misdiagnosis, errors in patient treatment and monitoring, as well as unnecessary additional testings, so it is necessary be aware of the possible interferences, and collaborate with clinicians so that the diagnostic and clinical disagreement could be clarified as soon as possible.
Key words: human anti-mouse antibodies; carcinoembryonic antigen; immunoassay
P53
Appropriate time of thrombophilia testing related to oral anticoagulant therapy: are recommended guidelines followed
Margetić S*, Getaldić B, Ćelap I, Vrkić N
Clinical Institute of Chemistry, Medical School Clinical Hospital Center Sestre Milosrdnice, Zagreb, Croatia
Background: The appropriate time of testing related to oral anticoagulant therapy (OAT) is one of the crucial preanalytical variables for thrombophilia investigation since many of these tests are affected by this therapy. In order to gain insight into physicians ordering practices, we performed a retrospective analysis of consecutive thrombophilia tests ordered during the 10-month period.
Materials and methods: The study included lupus anticoagulant (LA) investigation and functional assays for protein C (PC), protein (PS) and resistance to activated protein C (APCR), using commercially available coagulation methods (Siemens, Germany).
Results: A total number of tests performed was as follows: 853 (LA), 608 (PC), 443 (PS) and 540 (APCR.) Proportions of positive results for each test were: LA = 35/853 (4.1%), PC = 28/608 (4.6%), PS = 50/443 (11.3%) and APCR = 52/540 (9.6%). Of all positive test results, 10/35 (28.6%) for LA, 20/28 (71.4%) for PC, 22/50 (44%) for PS and 3/52 (5.8%) for APCR were attributable to OAT, thus representing false-positive results that could have considerable implications for patients.
Conclusions: Our findings suggest the appropriate time of testing related to OAT to be simply overlooked by clinicians managing patients with thrombosis, thus resulting with a huge waste of resources with concomitant high likelihood of false-positive test results. Our findings indicate that thrombophilia testing in patients on OAT still occurs frequently in clinical practice despite recommended guidelines. The net effect of such testing is likely to be more detrimental than beneficial for the patient. On the basis of our results, it is obvious that laboratories need to take a more substantial role in the thrombophilia investigation process than just providing doctors with numbers. The corrective actions should include ongoing education of clinicians and reject testing in patients on OAT since it would be more rational, cost-effective and beneficial for patients.
Key words: thrombophilia testing, oral anticoagulant therapy, preanalytical variable
P54
Application of Sample Reader inSIGHT2 device for automated reading of barcoded samples at sample reception unit
Fuček M*, Alpeza Viman I, Zekušić M, Herceg A, Sertić J
Department of Laboratory Diagnostics, University Hospital Centre Zagreb, Croatia
Background: Systemic control of entire laboratory process and monitoring and management of nonconformities are mandatory procedures for every medical laboratory with established quality control system. Along with the most common nonconformities related to sample quality (hemolysis, clot...), a significant number of the nonconformity entitled “sample not delivered” was recorded at Department of Laboratory Diagnostics in 2011. The number of such cases was considerably reduced after introduction of Sample Reader inSIGHT2 device for automated reading of barcoded samples.
Materials and methods: Prior to introduction of Sample Reader inSIGHT2 device, there was no possibility for individual receipt of requests for laboratory management in the laboratory information system (LIS), regardless of whether all samples associated with an individual request have been delivered to the laboratory. The Sample reader inSIGHT2 accepts requests only for the samples delivered to the laboratory and barcoded with the label generated from the hospital information system (HIS) or LIS. Communication with the user takes place via a touch sensitive screen. The device informs the user if a sample has already been received and recorded in LIS and if it is acceptable with regard to the sampling time. The files with accreditation-related data (sample receipt time, identification of the user who has received a sample, identification numbers of received samples) are kept for each received sample while, for the purpose of subsequent forensic management, each sample or test-tube rack is photographed.
Results: After application of Sample Reader inSIGHT2, the nonconformity “sample not delivered” was recorded in 154 cases, as compared to 1092 cases in 2011 (for the periods of three months), while the average receipt time has been reduced significantly.
Conclusion: The application of the device for automated reading of barcoded samples at sample receipt unit of the Department of Laboratory Diagnostics, has allowed rapid and accurate reading of barcoded samples and also a reduced number of the cases of nonconformity entitled “sample not delivered”.
Key words: automated reading of barcoded samples; nonconformities
P55
Acceptance limits for human blood sample stability: a review
Oddoze C1, Portugal H1, Avelland T2, Gianolli JM3, Letondel M4, Lalaurie E4, Coauch I4. GEFAP: Groupe d’Etude Français d’Amélioration de la phase Préanalytique.
1UTH Timone, Marseille, France
2LABM Analys, Marignane, France
3LABM des Charpennes, Villeurbanne, France
4BD PAS, Pont-de-Claix, France
Background: The preanalytical phase accounts for up to 70% of the errors occurring in laboratory medicine. Some of these errors occur due to instability of the analyte in the blood matrix which is not only linked to degradation of the sample, but also to external environmental factors such as time delays to sample processing and temperature fluctuations, both of which can artificially accelerate degradation. ISO 15189 ‘Medical laboratories — Particular requirements for quality and competence’ and country specific accreditation schemes require that the laboratory ‘develop and document criteria for acceptance or rejection of primary samples’. To achieve this, clinical laboratories need to refer to high quality sources of information (guidelines, recommendations and publications) to help determine their acceptance limits for processing time and temperature extremes for specific analytes.
Materials and methods: The objective of this review is to understand and compare published acceptance limits. The initial focus was on acceptance limits for whole blood stored at room temperature. Stability data and acceptation limits (%) were reviewed from more than 20 publications (post 2000) by compiling several criteria such as number of patients and statistical analysis methodology. To illustrate this process acceptance limits were compared for three sensitive parameters for chemistry (potassium, phosphate and glucose) and one for endocrinology (PTH).
Results: Significant discrepancies in acceptance limits for potassium and the other parameters were found. Inconsistent statistical analyses and small number of patients enrolled can explain some of these differences. However, they do not explain differences in acceptance limits for potassium that were published (4.8% vs. 10%). This could result in a significantly different interpretation of what the maximum allowable transportation time should be for whole blood.
Conclusion: There is a need for better definition of acceptance limits and these limits should be standardised internationally.
P56
Metabolomics technology validated quality markers for biobank plasma samples
Kamlage B*1, Schmitz O1, Schatz P2
1Metanomics GmbH, Berlin, Germany
2Metanomics Health GmbH, Berlin, Germany
Background: Research in the healthcare area like identification and validation of new diagnostic biomarkers, drug target discovery and treatment monitoring approaches often starts with the analysis of existing biobank samples. The quality of these biobank samples can be impaired by various pre-analytical sample processing steps that will confound the analytical results and decrease the value of research if not identified and addressed properly. Metabolite profiling, also known as “metabolomics”, is a well-suited technology to support the identification of technical biomarkers for the quality assessment of biobank samples due to its high sensitivity plus the broad coverage of physiological and chemical processes.
Materials and methods: Human EDTA plasma samples obtained after applying defined pre-analytical confounding factors were subjected to mass-spectrometry based metabolomics including selected targeted platforms MxP™ Broad profiling, MxP™ Eicosanoids, MxP™ Catecholamines and MxP™ Lipids.
Results: Metabolomics data sets were analyzed by simple and mixed-effect linear models. Various pre-analytical processes resulted in significant and reproducible changes of the human plasma metabolome. Several metabolites suited as Quality Markers were identified and validated in independent data sets after Bonferroni-Holm correction of the false-positive rate with P-values being <0.001.
Conclusions: The plasma metabolome is influenced by the pre-analytical phase. Reproducible and meaningful biomarker research demands standardized protocols for sample handling (quality assurance) as well as a quality control of samples. High-level result interpretation of metabolomics studies requires framework studies to understand the impact of the pre-analytical phase on the results and to elucidate the underlying physiological and chemical mechanisms.
Key words: metabolomics; quality control, pre-analytical variation
P57
MEPS method optimization for forensic toxicological analysis
Mayer M1,Romvári Zs1, Porpáczy Z1, Lajtai A2, Lakatos Á*2
1University of Pécs Medical School Department of Forensic Medicine, Pécs, Hungary
2University of Pécs Medical School Department of Laboratory Medicine, Pécs, Hungary
Background: The more sensitive procedures and instruments allow using smaller sample volumes for toxicological analysis. In addition, strict environmental guidelines, the more expensive chemicals and labor encourage promote the use of micro-techniques. We aimed to optimize the „micro-extraction by packed sorbent” (MEPS) method for phenylethylamine-type drugs form biological samples.
Materials and methods: We developed a rapid and easy-to-perform micro-extraction method using MEPS to extract stimulants from urine. We compared two reversed phase sorbent (C8 and C18) filled MEPS bin. We optimized the washing and eluting steps to enhance the recovery and to decrease the presence of endogenous substances. After evaporation of the eluate the dry residue was redissolved and analyzed by HPLC-DAD.
Results: The lower limit of detection of drugs of abuse was between 50 and 100 ng/mL in urine samples. The standard deviation (RSD%) of the repeatability was acceptable, the recoveries were between 57% and 92% for the examined compounds depending on their logP.
Conclusions: The C18 MEPS bin was found to be effective for sample preparation of drugs of abuse before forensic toxicological analysis. Using this method the time and solvent requirement was significantly reduced compared to our previously used SPE method.
Key words: urine; MEPS; phenylethylamine-type drugs of abuse
P58
The role of clinical signs in successful sweat testing in children
Matica J*, Aralica M
Clinical Hospital Centre Rijeka, Clinical Department of Laboratory Diagnostics, Rijeka, Croatia
Background: According to guidelines for the performance of sweat test for diagnosis of cystic fibrosis, patient’s clinical state is important factor for successful collection of sweat after pilocarpine iontophoresis. Aim of the study was to investigate the role of clinical signs such as poor hydration, fasting, elevated body temperature, acute illness, edema and skin lesions on successful sweat collection in children.
Materials and methods: Regarding the quantity of collected sweat participants were divided in two groups: sufficient sweat quantity group (SSQ) and non-sufficient sweat quantity group (NSSQ). There were 117 participants; 90 children (median age 30 (range 1-196) months) in SSQ group and 27 children (median age 25 (range 1-208) months) in NSSQ group. The questionnaire was consisted of five questions regarding clinical state of child (fasting, elevated body temperature, acute illness, edema, skin lesions) and two questions considering poor hydration. The questionnaire was presented to person accompanying tested child (parent or nurse), offering favorable (1 point) and unfavorable (0 points) answers, according to presence or absence of listed clinical signs.
The Fisher’s exact test (Medcalc statistical software, Mariakerke, Belgium) was used to calculate the difference between proportions of listed clinical signs in SSQ and NSSQ groups.
Results: Statistical analysis found significant difference between proportions of poor hydrated children in SSQ and NSSQ groups (P = 0.004 and P = 0.007). This was not the case for fasting, acute illness, elevated body temperature and skin lesions (P-values: 0.309; 0.638; 1.000 and 0.661 respectively) between two groups. There was no unfavorable answer related to edema in all tested children.
Conclusion: The study showed the importance of adequate hydration for successful collection of sweat sample, emphasizing crucial role of preanalytical phase.
Key words: sweat test; clinical sign; children; questionnaire; hydration
P59
Verification of claimed limit of detection for the Roche Troponin T hs STAT assay
Dukic L*, Simundic AM, Malogorski D
Medical School University Hospital Sestre Milosrdnice, Clinical Institute of Chemistry, Zagreb, Croatia
Background: Detection of cardiac biomarker troponin T (cTnT) together with other evidence of myocardial ischaemia makes important criteria for the diagnosis of myocardial infarction (MI). Limit of detection (LoD) is defined as the lowest amount of analyte in a sample that can be detected by given analytical procedure. Establishment of the limit of detection for cTnT has important role in assessment of patient suspected for MI. The aim of our study was to verify the claimed limit of detection of 0.005 μg/L for the cTnT assay (Roche Troponin T hs (high sensitive) STAT (Short Turn Around Time)) according the Clinical and Laboratory Standards Institute (CLSI) EP17-A protocol.
Materials and methods: Verification of the claimed LoD was performed on Cobas e 411 analyzer (Roche Diagnostics GmbH, Mannheim, Germany). Twenty repeated measurements of pooled plasma samples with concentration of cTnT 0.005 μg/L were performed.
Results: Claimed LoD for the Roche Troponin T hs STAT assay showed that observed proportion (20/20) is in accordance with the expected value (95%) at concentration of 0.005 μg/L.
Conclusions: Our data support the LoD claimed by Roche for Troponin T hs STAT assay.
Keywords: limit of detection; verification; troponin T assay
P60
New device reduce the pain but increase laboratory variability during diagnostic blood collection
Lima-Oliveira G*1,2,Campelo MDR3, Valentim CD3, Romano SJC3, Picheth G2, Guidi GC1,2
1Laboratory of Clinical Biochemistry, Department of Life and Reproduction Sciences, University of Verona, Italy
2Post-Graduate Program of Pharmaceutical Sciences, Department of Medical Pathology Federal University of Parana, Curitiba, Parana, Brazil
3Bioanalise Laboratory, Teresina, Piaui, Brazil
Background: A new device called Buzzy® promotes cold and vibration to relieve venipuncture pain during diagnostic blood specimen collection. The aim of our study was to evaluate the influence of Buzzy® device on routine clinical chemistry testing.
Materials and methods: Blood was collected from 15 volunteers by a single, expert phlebotomist. All patients were maintained seated for 15 minutes to eliminate possible interferences of blood distribution. After that, a vein was located on the left forearm using only a subcutaneous tissue transilluminator device (without tourniquet), and blood samples were collected using a 20-G straight needle (BD Vacuntainer) directly into 5mL SST II Advance® vacuum tubes with clot activator and acrylic gel separator. In sequence, an external cold and vibration by Buzzy® was applied on the right forearm for one minute before venipuncture and continued until the end of the same procedure realized in left forearm. The routine clinical biochemistry tests and haemolysis index (HI) were performed in duplicate immediately after centrifugation. The significance of the differences between samples, taking the values from samples collected using only the subcutaneous tissue transilluminator device as the reference ones was assessed by paired Student’s t-test after checking for normality.
Results: Statistical significant differences (P < 0.05) were observed for: total protein (P = 0.024), albumin (P = 0.014), urea (P < 0.001), creatine kinase (P = 0.004), gamma glutamyl transferase (P = 0.023), alkaline phosphatise (P = 0.023), amylase (P = 0.007), transferrin (P = 0.027), total cholesterol (P < 0.001), HDL-cholesterol (P = 0.008) and triglycerides (P = 0.010). No significant differences (P > 0.05) were observed for: glucose (P = 0.694), magnesium (P = 0.719), phosphate (P = 0.072), chloride (P = 0.276), potassium (P = 0.659), sodium (P = 1.000), iron (P = 0.682), creatinine (P = 0.776), C-reactive protein (P = 0.750), uric acid (P = 0.849), lipase (P = 0.248), lactate dehydrogenase (P = 0.113) and HI (P = 0.9885).
Conclusion: The Buzzy® device relieving venipuncture pain during blood specimen collection can be a new source of laboratory variability. It should be used only for laboratory tests that shown none significant differences. Moreover laboratory personnel should train their phlebotomists to verify the laboratory test request before to use Buzzy® during blood collection.
Key words: blood collection; phlebotomist; phlebotomy; quality laboratory management; laboratory variability; preanalytical phase, preanalytical variability
P61
Evaluation of the thrombin tubes and processing by Statspin express for the need of urgent laboratory assays and impact on TAT
Meško Brguljan P
University Clinic for Respiratory Diseases and Allergy Golnik, Clinical Chemistry Department, Golnik, Slovenia
Background: The laboratory, in consultation with the users, shall establish turnaround times (TAT) for each of the examinations that reflect clinical needs. The laboratory shall periodically evaluate whether or not it is meeting the established TATs. In our laboratory, one hour has been defined as TAT for urgent examinations. The preanalytical phase has an important impact on TAT. In order to improve TATs we decided to evaluate the use of Rapid Serum Tube (RST) Plus vacutainer (BD) using StatSpin Express 3 centrifuge for processing (IRIS). The use of RST vacutainers for TDM examinations have not been evaluate till now. Theophylline and digoxin are examination on our urgent repertoire and our aim was to evaluate the use of RST vacutainers and StatSpin processing in the comparison to tubes a traditional centrifugation, used in daily routine.
Materials and methods: 39 samples have collected in parallel to standard gel tubes (SST II, BD) and RST (BD) tubes and allowed to clot for 30 and 5 minutes, respectively. SST II tubes were centrifuged in traditional centrifuge (Rotina 38, Hettich) at 25 °C for 10 minutes at 3500 RPM and RST tubes for 5 minutes at 5600 RPM in StatiSpin Express 3 centrifuge. The samples were analysed on Cobas 6000 (Roche Diagnostics) with standard procedures. Results were analysed using Medcalc statistical software. TAT (monthly median of urgent requests) was evaluated by LIS (Labis Statistics).
Results: Clinically acceptable performance was demonstrated for both tested parameters, the regression equation for digoxin was y = 0.04296 + 0.9566 x and for theophylline y = 0.6313 + 1.0082 x. TATs (time from sampling to report) were reduced from 55 min to 42 min.
Conclusions: The use of RST tubes and processing in StatSpin centrifuge is appropriate for the whole urgent repertoire at our clinic, including TDM. The change in preanalytical procedure has a great impact on TAT.
Key words: theophylline; digoxin; preanalytical processing; TAT
P62
New vacuum tube for erytrocyte sedimentation rate: a comparison with K2EDTA
Lima-Oliveira G*1,2, Dima F1, Salvagno GL1, Marcori M1, Mainenti L1, Guidi GC1,2
1Laboratory of Clinical Biochemistry, Department of Life and Reproduction Sciences, University of Verona, Italy
2Post-Graduate Program of Pharmaceutical Sciences, Department of Medical Pathology Federal University of Parana, Curitiba, Parana, Brazil
Background: The erythrocyte sedimentation rate (ESR) is a nonspecific marker of disease and is often used by clinicians in assisting diagnosis and follow-up of a variety of infectious and inflammatory disorders. Nevertheless, no information is available on the influence of different Vacuum Tubes on ERS analysis. The aim of the present investigation is to compare ESR results obtained on blood specimens collected with two different types of vacuum tubes.
Materials and methods: Blood samples from 20 volunteers were collected by a single, expert phlebotomist. All were maintained seated for 15 minutes to eliminate possible interferences of blood distribution. After this interval of time, a vein was located on the forearm using only a subcutaneous tissue transilluminator device (without tourniquet), and blood samples were collected using a 20-G straight needle (Terumo) directly into 2 different vacuum tubes: Tube I: 2 mL 4NC ESR Sodium Citrate Premium® (Greiner bio-one, GmbH, Kremsmunster, Austria) and Tube II: 3 mL Venosafe® 5.9 mg K2EDTA (Terumo, Europe, Leuven, Belgium). All samples were assayed for ESR on the TEST 1 YDL® (ALIFAX, Padova, Italy). Calibrations were performed according to the instructions provided by the manufacturer. Analytical imprecision, expressed as inter-assay coefficient of variation (CV) and calculated according to internal quality control is 0.8-2.2%. Data were analysed with the paired Student’s t-test after checking for normality.
Results: Our resultsare expressed as mean ± standard error of the mean (SEM), shown statistically significant difference between Tube I (16 ± 2 mm/h) and Tube II (28 ± 3 mm/h), P < 0.001.
Conclusions: This investigation clearly attests that the preanalytical variability might also affect ESR testing, in that the type of the tube can influence test results. Finally, laboratory personnel should validate reference ranges for this new kind of tube before chose to introduce it in routine laboratory.
Key words: blood collection; phlebotomy; quality laboratory management; laboratory variability; preanalytical phase;
P63
Comparing apples with apples: the need for consistent pre-analytical error classification
Glaysher J*, Watson I, Wootton A
Department of Clinical Biochemistry, Aintree University Hospitals NHS Foundation Trust, Liverpool, UK
Background: With improvements to analytical quality, the collection and recording of pre-analytical errors is receiving increasing focus. Our laboratory was the first in the UK to enrol in the Australian Key Incident Monitoring and Management Systems QA scheme (KIMMS) which defines a number of pre-analytical error categories. The IFCC Working Group ‘‘Laboratory Errors and Patient Safety’’ has now published a similar list of 16 pre-analytical error categories.
Materials and methods: We designed and implemented a quality system aligned to KIMMS criteria to capture error information directly from the Laboratory Information System. The data are grouped by location and type, before feedback to service users using email of results and face-to-face discussion of their significance.
Results: Since September 2011 our laboratory has recorded an average of 3712 errors/month. In November 2012 the major error was not receiving a required sample type (N = 134; 2.28% of total samples received), closely followed by haemolysis (1003; 2.01%) and collection errors (433; 0.87%). Some of these data are at variance from KIMMS overall rates, particularly haemolysis (0.3%).
Conclusions: Despite the apparently straightforward definition of haemolysis, different laboratories categorise results differently. We designate every sample with an un-reportable test, whereas some laboratories only record samples that are completely rejected. Furthermore, the number of samples used as denominator for calculations is also problematic, since some laboratories classify each sample type whilst others count collection events singly. The automated error recording described here is likely to be more efficient than manual counting; significant numbers of errors may go unrecorded with the latter technique.
Consistent recording of errors is required to allow comparison between laboratories. KIMMS and other pre-analytical QA schemes provide a valuable basis for categorisation and collection of laboratory error rates. International agreement is needed to specify inclusion within each error category to ensure comparability of data.
P64
Performance characteristics of a commercially available reagent and dietil-eter for removing lipoproteins
Marevic S*, Laskaj R, Sokolic B
University Hospital for Infectious Diseases ‘’Dr. Fran Mihaljevic’’, Department for Medical Biochemistry and Hematology, Zagreb, Croatia
Background: Lipemic samples are a common preanalytical problem and interference of lipids with measurement of routine clinical biochemical tests like enzymes, bilirubin and C-reactive protein (CRP) is significant. We made a short evaluation, of the performance characteristics of a commercially available lipemia clearing reagent and also evaluated the use of dietil-eter for the same purpose, with visibly non-lipemic samples.
Materials and methods: We treated 10 non-lipemic serum samples with a commercially available reagent for removing lipoproteins (LipoClear, StatSpin, Iris, USA). Also 10 other, non-lipemic serum samples were processed with dietil-eter (Kemika, Croatia) for lipid removal. Determination of glucose, urea, creatinine, bilirubin, AST, ALT, GGT, ALP, LD and CRP was done in all samples, before and after removing lipoproteins, on a biochemistry analyzer (Beckman Coulter AU640, USA).
Results: When compared samples before and after treatment with LipoClear (multiplied with 1.2 for dilution correction) we did not found a statistically significant difference for all biochemical tests (P > 0.10) except for CRP (P < 0.05). In samples before and after removal of lipids with dietil-eter, the unpaired Wilcoxon test for all measured parameters, showed that there was no statistically significant difference (P > 0.10).
Conclusions: LipoClear is easy to use and shows good performance characteristics for determination of all our measured parameters except CRP. Dietil-eter is more complicate to use and careful handling of the specimen after centrifugation is necessary because of pipetting the lower layer for analysis. Although we get comparable results for all our parameters and CRP in our non-lipemic samples treated with dietil-eter, one must be careful because we had one CRP result that was not correct, probably due to specimen handling.
Key words: lipemic; LipoClear; dietil-eter
P65
Noncognitive factors in preanalytical of CBC measurement: negligible impact or a disastrous mistake?
Vidranski V, Laškaj R, Kolovrat K,Kozić Dokmanović S,Marević S*, Sokolić B
University Hospital for Infectious Diseases ‘’Dr. Fran Mihaljevic’’, Department for Medical Biochemistry and Hematology, Zagreb, Croatia
Background: Complete blood count (CBC) is one of the most commonly ordered routine laboratory test. Making great efforts to eliminate cognitive factors still cannot affect those noncognitive ones, whose misinterpretation, if not recognized, may bring the patient into a considerable risk by diagnostic and therapeutic mistakes.
Materials and methods: A total of four patients admitted to our Hospital were enrolled in this case reports. Blood samples were collected with tripotassium EDTA and with sodium citrate respectively. UniCell DxH 800, Beckman-Coulter hematology analyzer was used to perform CBC. Blood smears of samples were made for microscope observation (Olympus BX51). Samples with lipemia were centrifuged on Hettich Universal 320R for 5 min / 20 °C / 4000 RPM. Samples with cold agglutination were heated in waterbath Rowa GmbH for 10 min / 37 °C.
Results: In the first case it was a transient phenomenon-presence of cold agglutination of red blood cells due to a diagnosis of bilateral pneumonia with pleural effusion in the left chest, during antibiotic therapy. The second case discussed the transient EDTA-dependent pseudothrombocytopenia due to antibiotic treatment of erysipelas. The third case was an influence of lipemia on the value of hemoglobin and erythrocyte constants in the individual with previously known hyperlipemia. The fourth case described interference of malaria parasites on results of differential blood cell count.
Conclusions: Misinterpreted results, due above described-noncognitive influences of errors, can give rise to erroneous diagnosis and inappropriate treatment. The present literature about influence of infectious diseases on preanalitical errors is scarce, so these errors should be recognized and described.
Key words: preanalytical; lipemia; pseudothrombocytopenia; cold agglutination; malaria
P66
Are patients well informed about proper collection of 24-hour urine sample?
Miler M*, Simundic AM
University Department of Chemistry, Sestre Milosrdnice University Hospital Center, Zagreb, Croatia
Background: Outpatients are usually not quite well informed about the importance of proper urine collection. The proper collection of 24-hour urine sample is important because results of all biochemical parameters depend on its quantity and quality. In this preliminary study we aimed to assess how well outpatients are informed about the policy and procedures for 24-hour urine sample collection in our laboratory.
Materials and methods: Outpatients from Sestre Milosrdnice University Hospital Center with collected 24-hour urine sample were invited to participate in the survey. All patients who have consented to this survey were asked 10 questions about basic demographic data, urine collection, fluid volume intake on the day of urine collection, type of container for collection and collected urine volume. Patients were also asked to state the resource they have used to gather information on how to prepare for urine collection.
Results: We have enrolled 59 outpatients (21 males). Majority of participants were older than 65 years (29/59). Even 20 patients out of 59, did not follow correct urine collection procedure. Almost a quarter of patients either deliberately increases or decreases the fluid volume intake on the collecting day. Although the most (35/59) of the patients answered that the water bottle is the most suitable for urine collection, even 29 participants have chosen used bottle of soft drink, and 3/59 have used the bottle of milk for urine collection. A substantial proportion of patients (4/59) have even discarded a portion of urine that exceeded the bottle volume. When asked to state the resource they have used to gather information, most of the patients (37/59) stated that they were informed by the laboratory staff about the proper urine collection procedure.
Conclusions: A substantial proportion of patients are not following correct procedure for 24-hour urine collection. This clearly indicated the room for improvement. Better education of outpatients is needed. A concerted action of physicians and laboratory staff is the best approach to address this important issue.
Key words: 24-hour urine sample; urine collection; preanalytical factors
P67
How can we automate the handling of interfering factors in health laboratories?
Farkas K*, Siska A, Valczer E, Telkes M, Seres E
University of Szeged, Dept. of Laboratory Medicine, Szeged, Hungary
Background: There is a wide variety of modern automated systems to handle the effects of interfering factors (e.g. lipaemia, haemolysis etc.) on patient’s results. Many of them require expansive reorganization of laboratory workflows or new measuring instruments. The aim of this study was to apply these automatisms in our laboratory without changing workflows and instrumentation.
Materials and methods: Automatic index measurements of each serum samples have been introduced. The manual evaluation of serum indexes have been automated by using our Laboratory Information System. Shortly thereafter, a quick, uniform and objective interpretation of interfering factors in our routine laboratory workflow has been realized. In addition, the difference between our previous evaluation strategy and the actually used evaluating protocol was examined between 2007 and 2012.
Results: With this automated method we were able to reduce the errors below 0.005%.
Conclusions: On the basis of our results we could present an almost completely automated solution of handling interfering factors.
Key words: serum index; automation; interpretation; interfering factors
P68
Improvement of preanalytical phase quality in laboratories of University Hospital
Bunešová M*, Moučková Š, Moravcová L
Department of Clinical Biochemistry and Pathobiochemistry, 2nd Medical Faculty, Charles University, University Hospital Motol, Prague, Czech Republic
Objective: The observation of the state and the improvement of preanalytical phase quality by the influence of Society of Clinical Biochemistry Recommendation, accreditation of laboratories and the whole hospital and in the consequences cooperation of laboratories with departments with ambulances and with University Hospital clinics.
Materials and methods: The observation and evaluation of samples that were rejected by laboratories when taking-over due to some preanalytical mistakes. Te observation results were evaluated in relation with the total number of accepted samples as values quality indicators values. The study specially focused on inaccurately identified specimens. The observation was done from 2007 to 2012 at the Institute of Clinical Biochemistry and from 2012 at other Hospital laboratories too).
Results: Since the year 2007 we have been monitoring gradual but permanent growth of preanalytical phase quality. It is manifested by decrease of mistakes, by decrease of rejected samples number, inaccurately identified tubes and by the improvement of numerical values of preanalytical quality indicators identification errors and in the same day repeated requirements. The usage of the hospital information system allows us to differentiate quality of preanalytical phase even according to the level of demands of patients. The number of rejected samples at the emergency (0.14 %) was compared with the number of rejected samples at the intensive care unit (3.8 %) and with the department of long-term care (2.9 %).
Conclusion: The care of quality improvement of preanalytical phase is secured completely. We use Recommendation of Society of Clinical Biochemistry about the rejection of inappropriate samples and its concordance with the accreditation process according to ISO 15189. The cooperation of laboratories with department managers has been developed, trainings of nurses are being held, the NIS is used, the system of mutual connection between laboratories and departments and in case of need the support of top management is provided even with a possible deployment of deep analysis of serious errors and defects causes. The target is to reduce the risk of patient care in the hospital and for the future to create a pattern that could inspire the whole Czech Republic to improve significantly the preanalytical phase Quality
Key words: defective sample; identification; preanalytical mistakes
P69
Pre-analytical factors of routine coagulation assays in Norwegian hospital laboratories
Kristoffersen AH*1, Stavelin A2, Vannes S1, Kristensen GB3
1Laboratory of Clinical Biochemistry, Haukeland University Hospital, Bergen, Norway
2Norwegian Quality Improvement of Primary Care Laboratories (NOKLUS), Department of Public Health and Primary Health Care, University of Bergen, Bergen, Norway
3Norwegian Clinical Chemistry EQA Programme (NKK), Bergen, Norway
Background: External quality assessment schemes mainly consider the quality in the analytical phase, but studies have shown that up to 70% of the errors occur in the pre-analytical phase. Therefore, the Norwegian Clinical Chemistry EQA Programme (NKK) initiated a program on the pre-analytical phase to get an overview on how important aspects in the pre-analytical phase were handled in the hospital laboratories.
Materials and methods: Hospital laboratories in Norway were invited to fill in a web-based questionnaire regarding pre-analytical routines for the most common coagulation assays (APTT, INR, fibrinogen and D-dimer). The results focusing on the routines of how to detect and handle high/low hematocrit, hemolysis, bilirubinemia and/or lipemia, are presented.
Results: 57 of 69 (83%) laboratories responded, of which ~80% used instruments with mechanical clot detection. Most laboratories (89%) had no routines for detecting abnormal hematocrit (high/low). One of the laboratories corrected for low hematocrit, but none corrected for high. After centrifugation, a visual check for hemolysis was performed by 79% of the laboratories. When hemolysis was detected, 1/3 would reject the sample, 1/3 would report the result with a comment and 1/3 would report the result without a comment. 40% gave additional comments about sample rejection if strongly hemolysed. Lipemia and bilirubinemia was detected by visual inspection in 57% and 47% of the laboratories, respectively. D-dimer results would be reported without a comment despite lipemia or bilirubinemia in 38-63% of the laboratories answering this question, while this number was 75-91% for the other analyses. Only one laboratory had an instrument automatically checking for hemolysis, bilirubinemia and lipemia.
Conclusion: There is a large variation in how Norwegian laboratories detect and handle possibly interfering factors when coagulation assays are requested. The guidelines are quite vague on these issues, and it is difficult to assess the clinical consequences of this variation.
Key words: pre-analytical phase; coagulation assays; hemolysis; bilirubinemia; lipemia
