Daria Pašalić
Department of Medical Chemistry, Biochemistry and Clinical Chemistry
Zagreb University School of Medicine
Šalata ul 2.
10 000 Zagreb, Croatia
Phone +385 (1) 4590 205; +385 (1) 4566 940
E-mail: dariapasalic [at] gmail [dot] com

Useful links

P12-1 (Oral presentation)

Ožvald I1, Šurina B1, Flegar-Meštrić Z1, Vidas Ž2. P12-1: Advantage of tacrolimus determination on Abbott Architect analyzer. Biochemia Medica 2009;19(Suppl 1):S154.
1Department of clinical chemistry, Merkur University hospital, Zagreb, Croatia
2Surgical Clinic, Department of Organ Transplantation, Merkur University hospital, Zagreb, Croatia
Corresponding author:iozvald [at] gmail [dot] com
Introduction: Long life immunosuppressive therapy in transplant patients requiring intra- and interindividual therapeutic titrate. Lower therapeutic dose of drug reduces the toxic side effects. European Consensus Recommendation from the Committee on Tacrolimus Optimization recommended for the tacrolimus determination to use method which requires detection of functional sensitivity of 1 μg/L.
Materials and methods: Analytical validation of tacrolimus determination by chemiluminescent immunoassay (CMIA) on Abbott Architect system included: within-run imprecision on the human samples and commercially controls (N = 5); between-day imprecision on commercially controls (N = 15); inaccuracy on commercially controls (N = 15) and comparation determination on the human samples (N = 20) on 2 systems: IMX (MEIA) and Architect (CMIA) methods.
Results: Within-run imprecision on the human sample is 1.69% and on commercially controls for L1 is 1.24%, L2 is 3.61% and L3 is 2.38%; between-day imprecision on commercially controls is 4.01%; 4.49%; 3.38% for L1, L2, L3 respectively; inaccuracy on commercially controls for L1 is -4.95%, L2 is 0.39 and L3% is -2.23%. The results are within the target range of the Architect Tacrolims assay data and they are for imprecision and inaccuracy ≤ 10%. Linearity was confirmed with calibration curve in 6 points in concentration range from 0 to 30 ug/L. Preliminary data of method comparison on human samples (N = 20) show in the range of method linearity correlation coefficient from 0.957.
Conclusion: CMIA method for tacrolimus determination on the Abbott Architect system is sensitive and accurate test. Method has low functional sensitivity and elimination of interferences: hematocrit, high values of cholesterol, triglycerides, bilirubin, total protein and uric acid, what means that this method is more specific in comparison to other routine methods.
Obuljen J1, Mujagić R2, Tešija Kuna A3, Frakin-Užović I4, Svržan J3, Vrkić N3. P12-2: Comparison of colorimetric and turbidimetric method for total protein determination in urine. Biochemia Medica 2009;19(Suppl 1):S155-S156.
1Department of Laboratory Diagnostics, Zagreb Children’s Hospital, Zagreb, Croatia
2Medical Biochemistry Laboratory, Pula General Hospital, Pula, Croatia
3University Department of Chemistry, Sestre Milosrdnice University Hospital, Zagreb, Croatia
4Department of Transfusiology, Zadar General Hospital, Zadar, Croatia
Corresponding author:jasna [dot] obuljen [at] kdb [dot] hr
Background: The most frequently used automated methods for quantitative determination of total urinary proteins are colorimetric, dye binding methods (pyrogallol red) and turbidimetric, precipitation methods (benzethonium-chloride) which exhibit different sensitivities to various proteins. The aim of this study was analytical comparison of two methods for urine protein determination and evaluation of their diagnostic accuracy for determination of albuminuria.
Materials and methods: We examined 55 participants and determined total protein and albumin concentration in 24-hour urine collection. Total urine protein assays were performed on biochemical analyzer Olympus 2700 (reagents Olympus Urinary/CSF Reagent and Roche Diagnostics GmbH Urinary/CSF Protein), we calibrated both tests with the same protein standard from Olympus kit, commercial controls Bio-Rad Liquichek Urine Controls at two levels were used for inaccuracy control. Albumin was determined by nephelometric method on Behring Nephelometer analyzer II.
Results: Bland and Altman statistical method showed disagreement between two methods (mean amount of disagreement = -43.9%; 95% CI= -57.3 to -30.5) which is specifically expressed in the lower concentration range. Inter-assay imprecision, intra-assay imprecision and inaccuracy for pyrogallol and benzethonium-chloride were: 3.1%, 2.6%, 9.4% and 5.4%, 3.1% and 8.9%, respectively. Receiver operating characteristic (ROC) analysis was used to assess diagnostic accuracy. Decision criteria for albuminuria was albumin concentration ≥ 0.030 g/24h. The area under the ROC curve for colorimetric method was 0.867 (95% CI = 0.751-0.942) with the optimal calculated cut-off value 0.131 g/24 h, and for the turbidimetric AUC = 0.920 (95% CI = 0.817-0.975) with the optimal cut-off value 0.069 g/24h. Comparison of areas under the ROC curve showed no statistically significant difference.
Conclusion: Imprecision and inaccuracy are acceptable for both methods according to the Westgard specification, but quantification of total urine protein was not comparable, specifically at the lower concentration range. The two examined methods showed similar diagnostic accuracy, but with different cut-off values, so clinical decisions should be made according to the optimal cut-off value for the certain method.
Hrabrić Vlah S, Dvornik Š, Grdovic D. P12-3: Analytical performance of the Gem®Premier TM4000 – comparative study. Biochemia Medica 2009;19(Suppl 1):S156.
Department of laboratora diagnostics, Rijeka Clinical Hospital Centre, Croatia
Corresponding author:snjezana [dot] hrabric [dot] vlah [at] gmail [dot] com
Objective: The aim of this study was to compare quantitative data on a new point-of-care testing (POCT) Gem® PremierTM4000 analyzer with reference analyzers. Special attention was given to technology of measurement concentration of hemoglobin and hematocrit on the POCT analyzers.
Materials and methods: Analytical performance of the Gem® PremierTM4000 was compared with Ciba Corning 865, Gem Premier 3000 and Dimension RxL analyzers for pH/pCO2, pO2/electrolytes/metabolites and for hemoglobin/hematocrit. Within-run and between-run imprecisions have been done for the Gem® PremierTM4000 analyzer. In the period of 30 days, 200 whole blood samples were analyzed.
Results: Satisfactory results were obtained on within-run imprecision for all study analytes (CV ≤ 4.89%), except for pO2 (CV = 14.66%) and on between-run imprecision for all parameters (CV ≤ 4.41%), except for lactate (Level I: CV = 8%, Level II: CV = 6.59%). The correlation with the comparative methods was satisfactory, correlation coefficients were between 0.9458-0.9948, for most parameters, except for sodium r = 0.8635. For hemoglobin correlation coefficients were excellent, r = 0.9923 (Gem® PremierTM4000 vs. Ciba Corning 865) and r = 0.9759 (Gem® PremierTM4000 vs. Gem® PremierTM 3000). Passing Bablok regression analysis of analytes concentrations showed good compatibility, except for sodium (Gem® PremierTM4000 vs. Ciba Corning 865; y = 34.7500 + 0.7500x) and hemoglobin (Gem® PremierTM4000 vs. Gem Premier 3000; y = 23.6923 + 0,8308x).
Conclusions: Results of analyses obtained by the comparative measurement analytes between study and reference analyzers were compatible. Our results present that co-oximetry is the only accurate and reliable method of hemoglobin determination.
Matišić D1, Galez D2, Čvorišćec D1. P12-4: Short validation of the Sebia Capillarys 2 analyzer. Biochemia Medica 2009;19(Suppl 1):S157.
1Clinical Institute of Laboratory Diagnosis, Zagreb University School of Medicine and Clinical Hospital Center Zagreb, Zagreb, Croatia
2Department of Laboratory Diagnostics, Karlovac General Hospital, Karlovac, Croatia
Corresponding author:dmatisic [at] kbc-zagreb [dot] hr
Introduction: Short validation of the Sebia Capillarys 2 analyzer for capillary electrophoresis was performed according to the guidelines of the European Committee for Laboratory Standards (ECCLS).
Materials and method: Analytical properties of the analyzer were evaluated by determination of within-run imprecision, between-run imprecision, inaccuracy, and comparative determination serum samples on Sebia Capillarys 2 and Olympus HITE 320 analyzers. Within-run imprecision was examined in 10 consecutive measurements of two control sera during 3 days. Between-run imprecision was examined in duplicate in two control sera during 10 days. Inaccuracy was calculated as a percentage of deviation of mean determined value from the mean value declared for control sera. Comparative determinations was made using 40 serum samples.
Results: Within-run imprecision showed coefficients of variation < 4.64% for all fractions. Between-run imprecision showed coefficients of variation < 5.92% for all fractions. Inaccuracy testing showed bias < 8.32% for all fractions except alfa 2-globulin in Hipergamma Control where bias was 12.60%. Results of the comparison showed correlation coefficients to range from 0.8873 to 0.9665 for all fractions. Passing Bablok regression analysis showed that there was no significant deviation from linearity.
Conclusion: Results of the short validation showed the Sebia Capillarys 2 analyzer to be a precise, accurate and reliable instrument with high degree of correlation to Olympus HITE 320 analyzer. Full automatization, high resolution, high speed, high sensitivity and reproducibility and possible connection of the analyzer to laboratory information system have significantly broadened the possibilities for its application, making the routine work faster and easier than before.
Badža N1, Jurčić-Karlović E2, Doljanin Z3. P12-5: Analytical evaluation of biochemistry analyzer Pentra400 (Horiba ABX). Biochemia Medica 2009;19(Suppl 1):S158.
1Sunce Policlinic, Zadar, Croatia
2Sunce Policlinic, Rijeka, Croatia
3Sunce Policlinic, Split, Croatia
Corresponding author:erika [dot] jurcic-karlovic [at] sunce [dot] hr
Aim: To estimate analytical performance of biochemistry analyzer Pentra400, Horiba ABX, France, according to the ECCLS protocol by determination of creatinine, cholesterol, glucose, ALP, AST, CRP, sodium and potassium concentrations using reagents of same manufacturer.
Material and methods: Analytical assessment comprised within-run as well as between-run imprecision and inaccuracy. Within-run imprecision was determined by multiple (10x) measurements of commercial control samples (N/P) and human sera, and expressed as a coefficient of variation (CV). Between-run imprecision was determined on N/P control samples and pool sera during 15 days on two replicates. Inaccuracy was expressed as a percentage of bias (R%) of the determinated mean value from the declared mean value. Inaccuracy was evaluated using data from between-run imprecision measurement on control samples.
Results: Satisfactory results were achieved for within-run imprecision for creatinine, glucose, cholesterol, AST, CRP, sodium and potassium. Poorer result for the control sample P was obtained for ALP (CV 4.1%). Satisfactory results were achieved for between-run imprecision for glucose, cholesterol, AST, CRP and potassium. Poorer but acceptable results of control sample P were obtained for creatinine (CV 3.1%) and ALP (CV 4.5%). The results of between-run imprecision for sodium were not satisfactory (CV 1.7% for control N; CV 1.9% for control P; CV 1.0% for human sera, respectively). Deviation from declared control sample values showed a satisfactory level of accuracy for all tested parameters.
Conclusion: Results of the study indicate that biochemistry analyzer Pentra400 provides accurate and precise results and is convenient for use in routine laboratory.
Juričić G, Ravnić R, Trogrlić M, Honović L. P12-6: Harmonization of different analytical systems in everyday laboratory practice. Biochemia Medica 2009;19(Suppl 1):S159.
Laboratory of Medical biochemistry, Pula General Hospital, Pula, Croatia
Corresponding author:gjuricic [at] gmail [dot] com
Background: In accordance with the rules of good laboratory practice and with the introduction of harmonization of laboratory test results, in everyday practice we face the problem of harmonizing various analytical systems. The aim was to compare the measured concentrations of total and conjugated bilirubin obtained by various modifications of methods with diazo-reagent on biochemical analyzers Architect c8000 and Olympus AU400. Also, we consider the clinical justification of introduction of unique reference interval (Harmonization of laboratory test results in medical biochemistry; CCMB, 2004.).
Materials and methods: Concentration of total and conjugated bilirubin was determined in 41 sample. Comparison of methods was performed with MedCalc statistical software using Passing-Bablok regression and Bland-Altman plot.
Results: Measured concentrations of total bilirubin were on average 14.6% (95% CI = -19.0 to -10.1) lower on Architect c8000. Methods for determination of total bilirubin on Architect c8000 and Olympus AU400 can be harmonized with the direction of Passing-Bablok regression equation: y (tot.bil.Arch) = 0.934 x (tot.bil.Oly) – 1.906. Measured concentrations of conjugated bilirubin were on average 51.6% (95% CI = 42.4 to 60.9) higher on Architect c8000. Methods for determination of conjugated bilirubin on Architect c8000 and Olympus AU400 can not be properly harmonized because of significant deviation from linearity. Analytical deflection of total and conjugated bilirubin concentrations was more visible in the area of lower concentrations.
Conclusion: Considering significant disagreement between different methods with diazo-reagent for determination of conjugated bilirubin on two analytical systems, the legitimacy of the introduction of unique reference interval by Harmonization is questionable.
Salamunić I, Bilopavlović N, Galetović A, Pauković Sekulić B, Tandara L. P12-7: Comparison of different immunoassays for tumor marker. Biochemia Medica 2009;19(Suppl 1):S160.
Department of Medical Laboratory Diagnosis, Split University Hospital Center, Split, Croatia
Corresponding author:ilza [dot] salamunic [at] gmail [dot] com
Background: The comparability of immunoassay tests on different analyzer is extremly important during the assay validation process. In this study, we investigated the interchangeability of tumor markers (TM) on two different analyzer.
Methods: We tested patient samples for CEA, CA 19-9, CA 15-3, PSA and fPSA. We used Elecsys 2010 (Roche) with electrochemiluminescence detection and AU 3000i (Olympus) with chemiluminescent detection. We performed statistical comparative analysis using commercially available program MedCalc Software.
Results: The correlation coefficient (r) between Elecsys 2010 and AU 3000i were: CEA (N = 180, r = 0.993), CA 19-9 (N = 131, r = 0.979), CA 15-3 (N = 105, r = 0.826), PSA (N = 148, r = 0.986) and fPSA (N = 61, r = 0.976). The parameters of Passing-Bablok regression show significant systematic differences among analytical systems for: CEA (95% CI, a = -0.701 to -0.470; b = 0.869 to 0.948), CA 19-9 (95% CI, a = 1.533 to 2.476; b = 0.890 to 0.970), PSA (95% CI, b = 0.916 to 0.966), fPSA (95% CI, b = 0.735 to 0.829) not for CA 15-3 (95% CI, a = -1.233 to 2.327, b = 0.867 to 1.074). Systematic differences between tested analytical systems evaluated by Bland-Altman method were: CEA Elecsys-AU3000i (95% CI = -50.4 to 43.2), CA 19-9 Elecsys-AU3000i (95% CI = -111.1 to 87.6), CA 15-3 Elecsys-AU3000i (95% CI = -727 to 817.8), PSA Elecsys-AU3000i (95% CI= -11.1 to 8.9) and fPSA Elecsys-AU3000i (95% CI = -1.13 to 0.49).
Conclusion: The results of determining TM cannot be substitute for one analytical system to another. If TM method is changed, parallel measurements of TM with both methods for an appropiate time span strongly recommended.
Mujagić R1, Vrkić N2, Getaldić B2, Vukelić N2, Futač D2, KvaternikM2. P12-8: Biochemical indicators of iron status and modalities for calculation of transferrin saturation. Biochemia Medica 2009;19(Suppl 1):S161.
1Medical Biochemistry Laboratory, Pula General Hospital, Pula, Croatia
2University Department of Chemistry, Sestre Milosrdnice University Hospital, Zagreb, Croatia
Corresponding author:renat [dot] mujagic [at] pu [dot] t-com [dot] hr
Background: Estimation and differentiation of functional versus absolute iron deficiency as well as iron toxicity in clinical praxis is based on biochemical indicators of iron status such as transferrin saturation or feritin concentration. Selection of diagnostic tests for iron status estimation is very important because therapy strategy is based on these results.
Materials and methods: Blood samples (N = 34) had been collected from the patients with liver disease. At the request of the physician, blood samples were analyzed for the biochemical indicators of iron status: iron concentration (Fe), unsaturated iron binding capacity (UIBC) and total iron binding capacity (TIBC). Transferrin concentration (Tf) was additionally determined in the residual samples. Furthermore, transferrin saturation (TSAT) was calculated by using the widely accepted two different empirical equations: Equation I (Fe&UIBC): TSAT% = 100 x Fe/(UIBC + Fe) and Equation II (Fe&Tf): Tsat% = 3.98 x Fe/Tf.
Results: Total error of the calculated TSAT value summarizes analytical errors of the measured parameters that participate in equations. Derived parameters TSAT% and Tsat% were compared by Passing&Bablok and Bland&Altman statistical methods (MedCalc, v10.1.8). There was no significant deviations from linearity (P > 0.10) and the computed linear regression equation was: TSAT% = 1.084 x Tsat% + 0.774. Transferrin saturation that has been computed by the first equation (Tsat%) was 11.4% (95% CI = 9.4-13.3) higher in comparison with the TSAT that has been computed by the second equation (TSAT%).
Conclusion: Therefore, method evaluation is recommended before applying a new or improved assay for direct spectrophotometric determination of UIBC. Transferrin saturation calculated by second equation is more accurate method in comparison with the other alternative methods to assess TSAT.
Unić A, Đerek L, Serdar T, Stančin N, Šprajc N, Romić Ž. P12-9: Comparison of CA19-9, CEA and AFP on Vitros ECI and Cobas e411 analyzers. Biochemia Medica 2009;19(Suppl 1):S162.
Dubrava University hospital, Zagreb, Croatia
Corresponding author:adrianaunic [at] gmail [dot] com
Background: We compared the values of three tumor markers, CA19-9, CEA and AFP, obtained on Vitros ECi (Ortho Clinical Diagnostics, Johnson and Johnson, UK) and Cobas e411 analyzer (Roche Diagnostics, Japan), as a part of method validation for the new analyzer. The aim of the study was to examine the relationship between two methods for determinating tumor markers concetrations.
Materials and methods: The study included 26 sera for CA19-9, 20 for CEA and 18 for AFP. Vitros ECi is immunochemical analyzer that works on chemiluminiscence and Cobas e411 on electrochemiluminiscence principle. Methods were compared using Passing-Bablok regression.
Results: The values of the intercept and slope with 95% confidence interval obtained with Passing-Bablok regression were as follows: for CA19-9 0.8418 (-0.1021-2.9266) and 0.9977 (0.7793-1.0639); for CEA 1.0289 (0.8478-1.1875) and 0.9016 (0.8388-0.9979); for AFP 0.7081 (0.5436-0.8375) and 0.7932 (0.7357-0.8515). The Cusum linearity test performed on Passing-Bablok regression showed that there was no significant deviation from linearity for any of the above tumor markers combinations(P > 0.10).
Conclusion: Data obtained by Passing-Bablok regression showed that the results are not completely comparable and they do not follow the same linearity (slope and intercept on the y axis does not include 1 or 0). Only CA19-9 fulfilled settings of Passing-Bablok regression. CEA showed good comparability, but does not follow the same linearity. AFP did not meet the requirement of comparability, nor linearity. These results confirm the importance of recommendations on monitoring the values of tumor markers on the same analyzer with the same method.
Sikirica M, Flegar-Meštrić Z. P12-10: Comparison betweeen manual count erithrocytes and leukocytes with two test strips urine analyzers: LabUMat and Clinitek 500. Biochemia Medica 2009;19(Suppl 1):S163.
Institute of Clinical Chemistry, Merkur University Hospital, Zagreb, Croatia
Corresponding author:mirjana [dot] sikirica [at] gmail [dot] com
Introduction: Test strip analisys plays an important role in urinalysis. Automated test strips urine analyzers are now available to report semiquantitative data. In this study, we wanted to compare the reflectance readings with manual microscopy.
Materials and methods: We evalated concordance level between two automated test strips urine analyzers: LabUMat 77Elektronika Kft., Hungary and Clinitek 500, Bayer Corporation, Diagnostics Division, Tarrytown, NY with manual microscopy of supravitally stained specimens in detecting and counting erithrocytes (RBC) and leukocytes (WBC). We analyzed 165 urine specimens whithout centrifgation for RBC and WBC counting by two urine analyzers, Clinitek 500 and LabUMat based on reflectance masurement with manual microscopy performed according the European Urinalysis Guidelines (ELM - European Urinalysis Group)) recommendations.
Results: The results have been categorized into four ranges and calculated concordance level expressed as percent (%). The agreement for RBC and WBC made by two urine analyzers Clinitek 500 versus LabUMat showed a concordance level of 90% and 84% respectivly; between Clinitek 500 with manual microscopy of 83% and 81% and between urine analyzer LabUMat with manual microscopy showed a concordance level of 80% and 78% respectivly.
Conclusion: Both test strip analyzers showed similar results provides clinically acceptable results. The differences observed between the results obtained for RBC and WBC with test strip urine analyzers with those obtained with manual microscopy of supravitally stained specimens are based on the different methodology. The strip test analyzers measure enzymatic activity and do not count cell units.Therefore each positive results mast be confirmation by microscopy.
Ćorić J1, Jadrić R2, PanjetaM1. P12-11: Evaluation of fructosamine assay on the Vitalab analyzer. Biochemia Medica 2009;19(Suppl 1):S164.
1Institute for Clinical Chemistry and Biochemistry, Clinical Centre University of Sarajevo, Sarajevo, Bosnia and Herzegovina
2Department for Biochemistry, Medical Faculty, University of Sarajevo, Sarajevo, Bosnia and Herzegovina
Corresponding author:rjadric [at] hotmail [dot] com
Introduction: Fructosamine is a glycated protein and a time-averaged indicator of blood glucose level, that is used to assess the glycaemic status of diabetes.
Material and methods: We evaluated a fructosamine method on Vitalab analyzer. Fructosamine, in its ketoaminic form, reduces in an alkaline medium, with nitroblue tetrazolium, to formazan. The reaction rate, measured photometrically at 550 nm, is directly proportional to the concentration of fructosamine present in the sample.
Results: The sensitivity of the assay was 10 umol/L. The within-run precision (N = 20) was frim 1-2.3% (CV) for samples with fructosamine concentrations from 197-587 umol/L. The between-day precision (N = 20) was 2.2 % for a sample with a mean concentrations of 552 umol/L and 4.7% for a sample with a mean concentrations of 217 umol/L. The extended measurement range is up to 2000 umol/L with automatic dilution. Regression analysis comparing the Vitalab assay (y) and Abbott method (x) yielded the following equation: y = 1.10x – 21.1 and correlation coefficient (r) of 0.98.
Conclusion: The fructosamine method on Vitalab analyzer is accurate and precise.
Kožić I, Rumenjak V. P12-12: Comparison of the values of metabolites glucose, urea and creatinine using whole blood (EDTA) and serum as a sample on different analyzers. Biochemia Medica 2009;19(Suppl 1):S164-S165.
Department for clinical laboratory diagnostics, Sveti Duh General Hospital, Zagreb, Croatia
Corresponding author:irenakozic [at] yahoo [dot] com
Background: One way to make TAT (turn around time) shorter in emergency laboratories is performing the tests using whole blood as a sample. The purpose of this work was to compare the obtained values of glucose, BUN and creatinine using whole blood (EDTA) as a sample on a blood gas analyzer Stat Profile Critical Care Xpress CCX (Nova Biomedical) and the values of metabolites using serum as a sample on the Olympus AU 400 biochemistry analyzer.
Materials and methods: As samples we used whole blood (EDTA) and serums of hospital patients. Repeatability of serial sampling and day by day repeatability of the given metabolites was determined using quality control samples.
Results: Repeatability of serial sampling for glucose described by coefficient of variation was CV = 0.6%, for BUN CV = 2.7% and for creatinine 1%. Day by day repeatability is for glucose CV = 0.5%, for BUN CV = 4.6% and for creatinine CV = 4%. The obtained results were statistically processed the correlation coefficient between the two analyzers and two different samples for glucose (N = 88) was R2 = 0.8919, for BUN (N = 60) was R2 = 0.964 nad for creatinine (N = 84) R2 = 0.9819.
Conclusions: From the obtained results it can be seen that the correlation of the results for the metabolites glucose, BUN and creatinine using serum as a sample on the biochemistry analyzer Oympus AU 400 and the results for the same metabolites using whole blood as a sample on the blood gas analyzer CCX is significant. Also using whole blood as a sample we can make TAT considerably shorter.
Rumenjak V, Kožić I. P12-13: Comparison of the values of metabolites glucose, urea and creatinine using whole blood (EDTA) and serum as a sample on different analyzers. Biochemia Medica 2009;19(Suppl 1):S165-S166.
Department for clinical laboratory diagnostics, Sveti Duh General Hospital, Zagreb, Croatia
Corresponding author:irenakozic [at] yahoo [dot] com
Aim: It is now recognized that the accuracy of commonly used hospital glucose meters is adversely affected by the biochemical composition and matrix of blood in critical ill patients. The aim of the study is to evaluate a new generation meter (StatStrip Glucose Nova Biomedical Waltham, MA) that has been designed to compensate for these interference effects.
Material and methods: Whole blood venous samples were collected from patients admitted to ICU and from patients attending for dialysis treatment. Correlation studies were performed by analyzing the venous whole blood samples on two commonly used glucose meters and StatStrip Glucose. Serum samples prepared from the whole blood sample were tested using a reference plasma hexokinase method (Olympus AU 400). Hematocrit values were measured using a Sysmex KX 2100. Method correlation was assessed by linear regression analysis and the clinical accuracy was assessed by comparing the pattern of results to the ISO15197 criteria for blood glucose monitoring.
Results: Linear regression analysis showed good correlations for all tested meters. Median bias from the reference method was lower for the StatStrip than for either Roche Accu-chek Performa or Roche Accu-chek Active. For the Dialysis patient population StatStrip Glucose met the accuracy acceptance criteria of ISO 15197 but Roche Accu-chek Performa did not. For the ICU patients there is a negative bias associated with increasing hematocrit for the Roche Accu chek Active meter. StatStrip Glucose was unaffected by hematocrit.
Conclusions: StatStrip Glucose offers the potential for improved accuracy of glucose monitoring in hospitalized patients.
Ruljančić N1, Mihanović M2, BakližaA1, Rumenjak V3. P12-14: Analytical evaluation of TSH, tT3 and tT4 assays. Biochemia Medica 2009;19(Suppl 1):S166-S167.
1Clinical laboratory, Sveti Ivan Psychiatric hospital, Zagreb, Croatia
2Sveti Ivan Psychiatric hospital, Zagreb, Croatia
3Department for clinical laboratory diagnostics, Sveti Duh General Hospital, Zagreb, Croatia
Corresponding author:n_ruljancic [at] yahoo [dot] com
Introduction: The short evaluation for analytical system was performed according to CLSI protocols. Statistical analysis was performed using EP-Evaluator®: EP5-protocol – imprecision (SD and CV within- and between run, total CV); EP10-protocol – preliminary evaluation for linearity, recovery, precision, bias, carryover and drift; EP9-protocol – method comparison. Interpretations of results were performed according to three criteria: vendor’s claimed for SDP within run and CVP within run; national periodic proficiency testing surveys for bias (B%DK) and european biologic goals and calculated biological allowable total errors TE%B, CV%B and B%B.
Materials and methods: 1. method – Analyzer Elisys Uno ELISA assay for TSH (2nd gen., analytic sensitivity lower than 0.10 mlU/L), T3 and T4; 2. method – Analyzer Victors ECI CLIA method for TSH (3rd gen., analytic sensitivity lower than 0.02 mlU/L), T3 and T4. For samples we have used: a) Biorad control serum L1, L2, L3 for EP5 and EP10 protocol and b) 92 patient’s serum samples for method comparison according EP9-protocol.
Results: EP5-protocol: a) users SDE and CV%E lower than SDP within run and CVP within run for TSH, T3 and T4; b) users CV%totE higher than CV%B for TSH, T3 and T4. EP10-protocol: a) users CV%totE higher than CV%B for TSH, T3 and T4; b) users bias B%E lower than B%DK for TSH, T3 and T4 and c) t-value lower than 4.6 for intercept, slope, %carryover, recovery, linearity, except for TSH slope (t = 16.6). EP9-protocol: a) correlation coefficient higher than 0.975 for TSH (r = 0.9775, B% = -1.4), r lower than 0.975 for T3 (r = 0.8360, B% = 0.16) and T4 (r = 0.8070, B% = 1.8%); b) TE% between method higher than allowable TE%B.
Conclusion: Methods for determination of TSH, T3 and T4 met allowable vendor’s claimed for imprecision but did not met european biologic goals for CV%B. Methods met allowable bias (B%DK) according to national periodic proficiency testing surveys. Methods comparisons show that these two methods should not be used like equivalence.
Šupak Smolčić V1, Bilić-Zulle L1,2, Fišić E1. P12-15: Analytic validation of biochemistry analyzer Cobas 6000 analyzer series modul C501. Biochemia Medica 2009;19(Suppl 1):S167-S168.
1Clinical laboratory, Sveti Ivan Psychiatric hospital, Zagreb, Croatia
2Sveti Ivan Psychiatric hospital, Zagreb, Croatia
3Department for clinical laboratory diagnostics, Sveti Duh General Hospital, Zagreb, Croatia
Corresponding author:vesnasupak [at] gmail [dot] com
Introduction: Cobas 6000 (Hitachi, Japan) is biochemistry analyzer for spectrophotometric, immunoturbidimetric and ion-selective determination of biochemical analytes. Here we present part of comprehensive analyzer validation.
Materials and methods: Validation was made for 12 analytes which represent following groups: metabolites, enzymes, trace elements, specific proteins and electrolytes. Research included determination of within-run imprecision (N = 20), between-run imprecision (N = 30), and method comparison with routine biochemistry analyzer (Olympus AU640) (N = 50).
Results: Within-run imprecision coefficients of variations are: glucose = 2.8%, urea = 0.8%, creatinine = 4.0%, GGT = 0.8%, AST = 0.6%, ALT = 0.8%, IgG = 3.2%, IgA = 1.1%, IgM = 1.5%, Fe = 0.9%, Na = 0.4% i K = 0.6%, and between-run imprecision coefficients of variations are: glucose = 1.6%, urea = 2.0%, creatinine = 2.6%, GGT = 1.9%, AST = 2.7%, ALT = 1.7%, IgG = 2.5%, IgA = 3.2%, IgM = 7.9%, Fe = 3.2%, Na = 1.3%, K = 2.0%. Method comparison study showed correlation coefficient r > 0.99 regarding all tested analytes accept for Na r = 0.97. Passing-Bablok regression analysis provided linear equation for tested analytes as well as 95% confidence interval for intercept and slope. Intercept confidence interval for urea, Na and K includes zero as value, and slope confidence interval includes one as value which makes these analytes in accordance with routine method. Method comparison regression analysis for AST, ALT, IgM and Fe shows small constant difference (intercept CI does not include zero), and for IgG small proportional difference (slope CI does not include one). Both, small constant and proportional differences, are found for glucose, creatinine, GGT and IgA.
Conclusion: Regarding low CV values tested analyzer has satisfactory accuracy and precision. Most analytes are coherent on both analyzers whereas a part of them require adjustments of slope and intercept for complete accordance.