Imprecision Studies Research Articles

Evaluation of diagnostic reliability of DCA 2000 for rapid and simple monitoring of HbA1c

The monitoring of diabetic patients by evaluating glycated protein levels is now widely accepted and performed. The microchromatographic version of the high performance liquid chromatography method is the technique most frequently used in clinical practice. The DCA 2000 instrument (Bayer Diagnostics, Milan, Italy), based on an immunochemical technique, has been proposed for the rapid and simple evaluation of HbAlc, using even capillary blood. We evaluated 171 subjects including 22 healthy volunteers, 78 type 2 diabetic patients with different degrees of metabolic control, 11 women affected by gestational diabetes mellitus (GDM), 6 patients with hyperlipemia, 38 patients with chronic renal failure, 13 diabetic patients with chronic renal failure, and 3 patients with hemoglobinopathies. The DCA 2000 model was compared with the Diamat HPLC system. Data from within-run imprecision studies showed excellent precision, for both DCA 2000 and the HPLC system. The correlation between the two different systems, as shown by other statistical evaluations, was good (y = 0.911x + 0.462, r = 0.923). Results from the control group and diabetic patients were used to compare the two methods. Values obtained using the DCA 2000 were significantly lower (p < 0.0001) than those obtained with the HPLC system, in both healthy subjects and diabetic patients. To detect possible interferences, selected samples were analyzed from patients with hyperlipemia, diabetes and chronic renal failure, and hemoglobinopathies. While in the case of hyperlipemia, an acceptable correlation coefficient between the two systems was confirmed (y = 1.047x - 1.236, r = 0.876), in the case of chronic renal failure the correlation turned out to be very low (y = 0.254x + 3.456, r = 0.203). Our results indicate that the DCA 2000 gives accurate and reliable results in the clinical field of interest.

Reference interval for serum C-reactive protein in healthy blood donors using the Dade Behring N Latex CRP mono assay

Measurement of the serum concentration of the acute-phase reactant C-reactive protein (CRP) provides an useful objective marker in clinical practice, both in screening for diseases and in monitoring disease activity and response to therapy. The aim of this study was to evaluate the analytical performance of the assay in the concentration range from 0.2 to 15 mg/L, and to determine a reference interval for serum CRP in healthy blood donors using the Dade Behring N Latex CRP mono assay on the Dade Behring Nephelometer II. The assay is a particle-enhanced nephelometric immunoassay with a measuring range of 0.2-1100 mg/L. The sample volume is 40 microL, and the assay time is 6 min. The coefficients of variation for within-run and between-run imprecision studies were between 2.2 and 5.4% in serum pools with CRP concentrations between 1.29 and 11.98 mg/L. Linearity studies showed excellent correlation between the theoretical and the measured values. No interferences were detected from hemoglobin <1.0 mmol/L, bilirubin <512 micromol/L and Intralipid <20 g/L. Blood samples were collected from 268 healthy blood donors (107 females and 161 males) between 20 and 65 years old. The non-parametric reference intervals for serum CRP was calculated to be <0.2-14.7 mg/L in females and <0.2-9.16 mg/L in males. The Mann-Whitney U-test revealed no gender-related difference for serum CRP (p-value=0.72). A common reference interval for serum CRP for both genders was calculated to be <0.2-10.5 mg/L (median 0.98 mg/L, range <0.2-17.3 mg/L).

Evaluation of an automated immunoassay method for cytokine measurement using the Immulite Immunoassay system.

Cytokines are key mediators in cell regulation and communication. The concentration of these proteins can rapidly and importantly increase during severe clinical situations. However, current techniques are not adapted to stat measurement, thus making their clinical use limited. In this context, the commercialization of five new kits for cytokine measurement interleukin ((IL)-1, IL-6, IL-8, tumor necrosis factor (TNF)-alpha and IL-2R) on an automated immunoanalyzer, the Immulite, seems to be a new approach for the determination of these markers. We report here the evaluation of the performance of these tests. The technique is based on a solid phase (bead) two site chemiluminescent enzyme immunometric assay. The analysis is performed within 60 to 90 minutes and the calibration is stable for 15 days. The values of the between-run imprecision study were similar to those from the within-run study with coefficients of variation (CV) ranging from 2% (low values of IL-8) to 11.5% for intermediate concentrations of IL-6 (500 pg/ml). CVs were usually around 5%. The accuracy was determined by a linearity study using standards (except for IL-2R) provided by the National Institute for Biological Standards and Control (NIBSC). Slopes obtained during this study were close to 1 (r2 = 0.99), except for IL-6, for which the slope was 1.55. TNF-alpha values were close to those expected. IL-1 results were about 20% higher. IL-6 values were over estimated above 100 pg/ml and under estimated below this value. IL-8 study seemed to be impaired by the poor stability of this molecule in the NIBSC preparation. Correlation study with standard laboratory techniques gave variable results: for IL-1 (n = 43) the slope was 0.77 (study carried out using cell culture media), for IL-6 (n = 54) the slope was 0.78, for IL-8 (n = 37) the slope was 1.64, for TNF-alpha (n = 40) the slope was 0.33 and the slope for IL-2R (n = 51) was 5.1. For the last cytokine, the unit in Immulite assay was different from the one used in our comparison technique. Cross-calibration results were consistent with these data and show that the bias is probably linked to a calibration problem. The study demonstrated excellent practicality of the system, and good stability of the calibration curve (15 days). However, the sample volume required (350 microl for the IL-6 and the TNF-alpha) could constitute a limitation for pediatric measurements.

Evaluation of the first automated thyroglobulin assay.

The aim of this study was to investigate technical and analytical performance of the first automated thyroglobulin (Tg) assay (DPC-Immulite; Diagnostic Products Corporation, Los Angeles, USA). In imprecision studies using several human serum pools ranging from 21 to 58 replicates, a coefficient of variation of 9.0% was obtained at a mean Tg concentration of 0.84 ng/ml and of 6.1% at a Tg concentration of 62.1 ng/ml. In a method comparison with a non-automated assay (BRAHMS LUMItest Tg, BRAHMS, Berlin, Germany) using 383 sera of 303 patients with thyroid carcinoma, regression analysis according to Passing and Bablock yielded in the following equation: Immulite Tg = 1.6 x BRAHMS Tg-0.1 ng/ml (Pearson's r = 0.979). Sera obtained from 59 patients with thyroid carcinoma enabled comparative follow-up studies; in all cases qualitative agreement was found with regard to increase or decrease of serum Tg; in eight cases, however, Tg was detected with the Immulite assay but not with the BRAHMS assay. Further follow-up proved the presence of thyroid tissue in these patients. From these and further methodological data (dilution linearity, interference studies, carry-over study, high-dose hook properties, and short report time) it is concluded that the DPC-Immulite Tg assay meets the requirements of routine diagnostic use.

Evaluation of the Dade Behring N Latex Cystatin C assay on the Dade Behring Nephelometer II System

The Dade Behring N Latex Cystatin C assay, a particle-enhanced nephelometric immunoassay for measuring serum cystatin C, was evaluated on the Dade Behring Nephelometer II. The assay time was 6 min and the throughput was 75 samples per hour. The sample volume was 40 microL and the measuring range was 0.25-7.90 mg/L. Imprecision studies revealed within-run CVs < 1.8% and between-run CVs < 1.8% in the concentration range 0.87-4.63 mg/L. Recovery was 92.4-101.3%. Linearity studies showed excellent correlation between the theoretical and obtained values. No interferences were detected from haemoglobin < 1.0 mmol/ L, bilirubin <512 micromol/L and Intralipid <20 g/L. Stability of cystatin C in serum was 7 days at temperatures from 20 degrees C to 20 degrees C and 6 months at -80 degrees C. Measurements of cystatin C in heparin-plasma and EDTA-plasma did not differ significantly from cystatin C measured in serum. Fifty patient samples run on the Dade Behring Nephelometer II (y) were compared to the Dako Cystatin C assay (x). The Passing-Bablok regression analysis revealed y = 1.105x - 0.340. In conclusion, the Dade Behring N Latex Cystatin C assay was precise and correlated with the Dako Cystatin C assay.

Analytical and clinical evaluation of an electrochemiluminescence immunoassay for the determination of CA 125

The CA 125 II assay on the Elecsys(R) 2010 analyzer was evaluated in an international multicenter trial. Imprecision studies yielded within-run CVs of 0.8-3.3% and between-day CVs of 2.4-10.9%; CVs for total imprecision in the manufacturer's laboratory were 2.4-7.8%. The linear range of the assay extended to at least 4500 kilounits/L (three decades). Interference from triglycerides (10.3 mmol/L), bilirubin (850 micromol/L), hemoglobin (1.1 mmol/L), anticoagulants (plasma), and several widely used drugs was undetectable. Method comparisons with five other CA 125 II assays showed good correlation but differences in standardization. A 95th percentile cutoff value of 35 kilounits/L was calculated from values measured in 593 apparently healthy (pre- and postmenopausal) women. In 95% of patients with benign gynecological diseases CA 125 was </=190 kilounits/L; 63% of patients with newly diagnosed ovarian carcinoma had values >190 kilounits/L. A comparison of CA 125 values obtained with the Elecsys test and with other common CA 125 tests in monitored patients being treated for ovarian cancer showed identical patterns. In conclusion, the Elecsys CA 125 II assay is linear over a broad range, yields precise and accurate results, is free from interferences, and compares well with other assays.

Multicenter evaluation of the Bayer Immuno 1™ CA 15-3™ assay

We conducted a multicenter evaluation of the analytical and clinical features of the automated Bayer Immuno 1 CA 15-3 assay and compared assay performance to two manual tests. Results of the 10-day imprecision study of the Bayer Immuno 1 assay pooled across four evaluation sites and three lots of reagent produced total CV < or = 4%. Lot-to-lot reproducibility for 26 different lots of reagents and calibrators manufactured over a 2-year period was demonstrated (CV, 1.1%). Results for the Bayer Immuno 1 assay correlated well with the Biomira TRUQUANT BR 27.29 and Centocor CA 15-3 RIAs (r > or = 0.94). The upper limit of the reference interval for the Bayer Immuno 1 assay was 35.9 kilounits/L (35.9 units/mL); values were similar for all methods. Longitudinal monitoring of healthy women yielded assay values with an average CV of 11% and 21% for the Bayer Immuno 1 and Biomira assays, respectively. The Bayer Immuno 1 assay demonstrated the analytical features, intermethod correlation, and long-term performance characteristics that are essential for longitudinal monitoring of breast cancer patients.

Automated measurement of mouse apolipoprotein B: convenient screening tool for mouse models of atherosclerosis

Although mice are commonly used for studies of atherosclerosis, investigators have had no convenient way to quantify apolipoprotein (apo) B, the major protein of atherogenic lipoproteins, in this model. We now report an automated immunoturbidimetric assay for mouse apo B with an NCCLS imprecision study CV < 5%. Added hemoglobin up to 50 g/L did not interfere with the assay, nor did one freeze-thaw cycle of serum samples. Assay linearity extends to apo B concentrations of 325 mg/L. We have used the assay to determine serum apo B concentrations under several atherogenic conditions, including the apo E "knock-out" genotype and treatment with a high-cholesterol diet. Our assay can be used to survey inbred mouse strains for variants in apo B concentrations or regulation. Moreover, the mouse can now be used as a convenient small-animal model to screen compounds that may lower apo B concentrations.

Use of precision profiles to evaluate precision of the automated leukocyte differential

The commonly used methods of assessing the precision of the automated leukocyte differential have certain drawbacks that affect the validity and comparability of results. In the present report, we introduce a procedure based on building precision profiles from a large number of within-run imprecision experiments. The profiles are fitted to the function for the CV of proportions, which yields the number of theoretically differentiated leukocytes. Differences between fitted curves are evaluated for statistical significance by the F-test. As an example, we compared the precision of two hematology analyzers, a flow-cytometric technique involving fluorescence-labeled monoclonal antibodies, and the manual differential. We were able to establish definite differences in precision between different analyzers and different leukocyte classes. Our data also indicated that conventional within-run imprecision studies may completely misjudge analyzer precision. Furthermore, we could demonstrate that the precision of analyzers that analyze a fixed amount of blood rather than a fixed number of leukocytes is strongly influenced by the leukocyte count of the sample, leading to high imprecision for leukopenic samples. We believe the proposed procedure is a useful addition to currently used protocols; it yields clear results and creates a statistical basis of comparison between various instruments and techniques of differentiation.

Multicenter evaluation of a new capillary blood prothrombin time monitoring system

The analytical performance of the new capillary blood prothrombin time monitoring system CoaguChek was examined in a multicenter evaluation at six hospitals. The coefficients of variation of the INR obtained in the CoaguChek imprecision study were approximately 7% in the control plasma provided (within-run and day-to-day) and 4% in blood (within-run). The prothrombin times were ascertained in capillary blood (CoaguChek PT Test) and citrated venous plasma (Hepato Quick, Thromborel S) from 359 patients under oral anticoagulation therapy with phenprocoumon, acenocoumarin or warfarin. The agreement with the test results obtained with the comparison methods was acceptable versus Hepato Quick assay (n = 359; y = 1.23 x -0.49, r = 0.888) and good versus Thromborel S method (n = 359; y = 1.09 x -0.28, r = 0.895). A simplified assessment of all test results (n = 795) in a nine-field comparison table showed a concordance with the comparison methods of more than 80 (Hepato Quick: 81%, Thromborel S: 83%). The concordance between Hepato-Quick and Thromborel S was slightly higher (88%). Its good analytical performance and convenient handling recommend the CoaguChek system as a suitable system for decentralized prothrombin time testing.

Development of an enzyme-linked immunosorbent assay for 2,3-dinor-6-keto-prostagland in F 1α in urine using a monoclonal antibody

Objectives: To develop and validate an enzyme-linked immunosorbent assay (ELISA) for measurement of urinary 2,3-dinor-6-keto-prostaglandin F 1α (2,3D6KPGF 1α) using a monoclonal antibody and a horseradish peroxidase-linked antigen. Design and Methods: Assay validation included optimization of the standard curve, antibody cross-reactivity, accuracy and imprecision studies together with preliminary measurement of clinical samples. Results: Optimal conditions of the standard curve (0.078–10.0 μg/L) used 2 mg/L of antibody, and 3 μg/L of peroxidase conjugate in each well, at pH 7.2. The coefficient of variation of various concentrations of the standard curve averaged 6.8%. Antibody cross-reactivity was <0.01% for related prostanoids. Recovery of known amounts (0.1–5.0 μg/L) of 2,3D6KPGF 1α added to an urinary sample was 101.2 ± 6.3%. Imprecision studies with non-pregnant (0.24 μg/L) and pregnant (2.5 μg/L) samples displayed an intraassay variability of 8.9 and 9.9%, and an interassay variability of 9.6 and 10.0%, respectively. Urinary measurements in the non-pregnant and pregnant states were similar to those previously reported. An apparent decreased concentration was observed early in pregnancy in future preeclampsia. Conclusion: With similar precision and validity, our assay method is time- and cost-saving. Preliminary urinary measurements show that this analyte may be of interest as an early marker for preeclampsia.

Epidemiological studies and radiation protection: The potential value of a new study at U.S. nuclear power plants

Risk estimates of fatal cancer constitute the principal component of the health detriment after low doses of ionizing radiation protection recommendations by ICRP and NCRP are based upon them. These risks are presently derived from the Life Span Study of the A bomb survivors in Japan and have recently been supported by a number of relatively imprecise studies of occupational exposure of workers in the U.K., in the U.S., and in Russia. Further verification by more precise studies would be most useful. The most promising future study is that of workers in the U.S. nuclear industry, a study which was proposed some years ago (by New York University) and endorsed by a special committee of the National Academy of Sciences, the National Research Council. The study has not yet been undertaken despite its intrinsic capability to yield more precise results than all other worker studies. This paper describes the present status of background risk estimation for radiation protection and urges that the study of workers in the U.S. nuclear power industry be undertaken without further delay.

Evaluation and correlation studies of common analytes on the beckman synchron CX5 and CX7

The Synchron CX7 was evaluated in order to transfer the analytes comprising the renal and liver function panels, cardiac enzymes, calcium and urate tests from the CX5 to the CX7. Imprecision studies were performed for intra-batch (21 analyses in one run at two levels) and inter-batch (greater than 150 analyses over a two week period at two levels). The mean, standard deviation and coefficient of variation were calculated. Patient correlation was performed using samples previously run on the CX5. Linearity studies were undertaken using a set of AQAP material (Cycle 21) purchased for this study. Linear regression analysis was carried out according to Demlng. The imprecision studies showed acceptable performance for all chemistries evaluated (17 in top 50% of laboratories in AQAP Cycle 24). The urea, creatinine calcium and glucose are analysed on the CX3 module of the CX7 and therefore have different methods to the CX5. The patient correlation studies showed that these analytes could be transferred to the CX7 with retention of the current reference ranges as could those tests using the same method.

Comparison of seven serum assays on fourautomatic analysers

Serum enzyme assays used on four different analysers (Hitachi 737, Hitachi 705, Cobas-bio and RA-2X) were compared by determining the activity of seven different enzymes (AST, ALT, LD, ALP, GGT, CK and AMS). Performance checks (quality control procedure) and replications (the study of the total analytic imprecision and of its components) were conducted and the methods were compared by linear regression analysis with statistical inference on the curves following the protocols of the National Committee for Clinical Laboratory Standards (NCCLS: PSEP- 2, PSEP-3, PSEP-4). The correlation coefficients between the methods (r = 0.991-0.999), together with the other statistical parameters, indicated that the methods are well correlated on all the instruments. The total imprecision was good for all analytes, except ALT. Among the instruments tested, the RA-2X gave more variable results, although the total imprecision was acceptable. There was no relevant carry-over effect. The evaluation of performance claims indicated that the expected error did not substantially affect the results at the level of clinical decisions.

Direct concurrent measurement of urinary vanillylmandelic acid, 5‐hydroxyindoleacetic acid and homovanillic acid by HPLC. Three methodologies compared

Three different direct HPLC methods for the determination of 3-methoxy-4-hydroxymandelic acid (VMA, vanillylmandelic acid), 5-hydroxyindoleacetic acid (5-HIAA) and 3-methoxy-4-hydroxyphenylacetic acid (HVA, homovanillic acid) in urine were compared: two spectrofluorometric methods, applying discontinuous gradients, and one serial coulometric linear gradient method. The imprecision study (n = 6) revealed comparable coefficients of variation (CV), intra-assay ranging 1.4-11.1%, and inter-assay ranging 5.9-11.8% for physiological and moderately elevated levels of VMA, 5-HIAA and HVA. All methods showed good linearities up to 100 mumol/L for each of the three compounds studied. Analytical recoveries were 97-114% for VMA, 87-103% for 5-HIAA, and 80-95% for HVA. Recoveries were not dependent on urinary relative densities in the range 1.010-1.030 kg/L or on protein content (prior to acidification) in the range 0.1-3 g/L, or on the pH of conservation in the range 2-5 or on storage temperature in the range -20 - +22 degrees C for three weeks. The distributed-sample comparison revealed acceptable correlations and clinically unimportant accuracy differences between the methods. It is concluded that direct fluorometric and electrochemical HPLC methods can be used in the determination of major catecholamine and serotonin metabolites in human urine for clinical diagnosis and follow-up of neural crest and carcinoid tumours.

How to improve estimates of imprecision.

Authors sometimes report total imprecision as being less than within-run imprecision. This leads one to question how well the underlying statistical concepts are understood. This article explains statistical concepts relevant to imprecision studies and reviews proper use of the relevant statistical terms. Calculation of total imprecision and its components is illustrated by a numerical example. Two of the most common calculation mistakes are pointed out. Finally, we recommend improvements in experimental design that will result in more precise estimates.

A comparative study of five different methods for the determination of 3-methoxy-4-hydroxymandelic acid in urine.

Five different methods for the determination of 3-methoxy-4-hydroxymandelic acid (vanilmandelic acid, VMA) in urine were compared: a GLC-FID catecholamine metabolite profiling method, an HPLC method with electrochemical detection, the method of Pisano et al. [1962) Clin. Chim. Acta 7, 285-291), a one dimensional paper chromatographic method with diazotized p-nitroaniline staining and the commercially available Bio-Rad VMA by Column Test. The comparison consisted of an imprecision study, a linearity check, a recovery study, a split sample comparison and an interference study. The best results of the imprecision study (n = 8) were found with the Bio-Rad and the HPLC method (within-run imprecision had a coefficient of variation (CV) of 5.1% and 1.4%; between-days CV of 5.9% and 6.0% respectively for values of 32.4 mumol/l and 24.5 mumol/l). The Pisano method had the poorest within-run CV (14.6%) and between-days CV (16.8%) for a value of 23.2 mumol/l. All methods showed good linearity. The mean recovery of the HPLC method was 101.3%; the mean recovery of the other four methods ranged from 93.9%-96.0%. The split sample comparison showed that the accuracy of the HPLC, the GLC and the Pisano method is comparable. The accuracy of the paper chromatographic method and the Bio-Rad method had a positive bias compared with the HPLC method. Especially the positive bias of the Bio-Rad method can be very large. The HPLC method was not influenced by the compounds tested in the interference study, whereas the GLC method in some cases only suffered from overloading problems. The Pisano and the Bio-Rad method were most influenced by the interfering compounds tested. We conclude that the HPLC and the GLC methods are superior to the other three VMA methods. From an analytical point of view HPLC is the method of choice for determining 3-methoxy-4-hydroxymandelic acid in urine.

A study of geographical imprecision: the tropics

Australian Geographical StudiesVolume 17, Issue 1 p. 3-17 A study of geographical imprecision: the tropics JOHN OLIVER, JOHN OLIVER *John Oliver is Professor of Geography at James Cook University of North Queensland. This paper is based on a Presidential Address to the Institute of Australian Geographers at Townsville in August, 1978.Search for more papers by this author JOHN OLIVER, JOHN OLIVER *John Oliver is Professor of Geography at James Cook University of North Queensland. This paper is based on a Presidential Address to the Institute of Australian Geographers at Townsville in August, 1978.Search for more papers by this author First published: April 1979 https://doi.org/10.1111/j.1467-8470.1979.tb00330.xCitations: 2 AboutPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Citing Literature Volume17, Issue1April 1979Pages 3-17 RelatedInformation