Repeating tests: Different roles in research studies and clinical medicine

Department of Medicine, Section of Rheumatology, Boston University School of Medicine, 72 E Concord St, E-533, Boston, MA 02118, USA. .
Biomarkers in Medicine (Impact Factor: 2.65). 10/2012; 6(5):691-703. DOI: 10.2217/bmm.12.57
Source: PubMed


Researchers often decide whether to average multiple results in order to produce more precise data, and clinicians often decide whether to repeat a laboratory test in order to confirm its validity or to follow a trend. Some of the major sources of variation in laboratory tests (analytical imprecision, within-subject biological variation and between-subject variation) and the effects of averaging multiple results from the same sample or from the same person over time are discussed quantitatively in this article. This analysis leads to the surprising conclusion that the strategy of averaging multiple results is only necessary and effective in a limited range of research studies. In clinical practice, it may be important to repeat a test in order to eliminate the possibility of a rare type of error that has nothing to do analytical imprecision or within-subject variation, and for this reason, paradoxically, it may be most important to repeat tests with the highest sensitivity and/or specificity (i.e., ones that are critical for clinical decision-making).

1 Follower
6 Reads
  • [Show abstract] [Hide abstract]
    ABSTRACT: Objective: The use of reference change values (RCV) has been advocated as a most appropriate for monitoring individuals in clinical laboratories. The aim of this study was to calculate of RCV of 24 analytes of immun analysis parameters. Methods: Twenty four serum analytes (Cortisol, DHEA-S, estradiol, follicle stimulating hormone (FSH), luteinizing hormone (LH), prolactin, testosterone, troponin I, NT-proBNP, parathormone (PTH), insulin, ferritin, folate, thyroid stimulating hormone (TSH), free T3, free T4, thyroglobulin antibody (anti TG), thyroid peroxidase antibody (anti TPO), alpha fetoprotein (AFP), CA 15-3, CA 19-9, CA 125, carcinoembryonic antigen (CEA) and total PSA) were analyzed with Roche kits (Roche Diagnostics, Mannheim, Germany), which were manifactured to use immun analyzer Cobas "6000 (Roche Diagnostics, Mannheim, Germany). We used analytic coefficient of variation (CVa) from internal quality control programme. Intra-individual biological variation (CVw) were obtained from Ricos current updated published in Westgards website. This database was updated in 2014. We calculated RCV using this formula RCV=21/2xZx(CVa2+CVw2)1/2 (Z=1.65, 95% probability, unidirectional; Z=2.33, 99% probability, unidirectional; Z=1.96, 95% probability, bidirectional; Z=2.58, 99% probability, bidirectional). Results: RCVs (Z=1.96) of these analytes were calculated as 44.30% for Cortisol, 31.25% for DHEA-S, 64.34% for estradiol, 35.14% for FSH, 64.58% for LH, 65.24% for prolactin, 27.28% for total testosterone, 39.24% for troponin I, 29.00% for NT-proBNP, 71.93% for PTH, 59.22% for insulin, 40.51% for ferritin, 68.04% for folate, 53.74% for TSH, 23.07% for free T3, 17.38% for free T4, 29.27% for anti TG, 42.52% for anti TPO, 35.42% for AFP, 21.58% for CA 15-3, 44.99% for CA 19-9, 68.92% for CA 125, 36.83% for CEA and 51.00% for total PSA. Conclusion: We suggest to use RCV as well as to use population-based reference interval. RCV could be a valuable tool for clinical decision in especially monitoring individuals.
    Gazi Medical Journal 01/2015; 26(3):85-87. DOI:10.12996/gmj.2015.27
  • [Show abstract] [Hide abstract]
    ABSTRACT: The use of Reference Change Values (RCV) has been advocated as very useful for monitoring individuals. Most of these are performed for monitoring individuals in acute situations and for following up the improvement or deterioration of chronic diseases. In our study, we aimed at evaluating the RCV calculation for 24 clinical chemistry analytes widely used in clinical laboratories and the utilization of this data. Twenty-four serum samples were analyzed with Abbott kits (Abbott Laboratories, Abbott Park, IL, USA), manufactured for use with the Architect c8000 (Abbott Laboratories, Abbott Park, IL, USA) auto-analyzer. We calculated RCV using the following formula: RCV = Z x 2 1/2x (CVA2 + CVw2)1/2. Four reference change values (RCV) were calculated for each analyte using four statistical probabilities (0.95, and 0.99, unidirectional and bidirectional). Moreover, by providing an interval after identifying upper and lower limits with the Reference Change Factor (RCF), serially measured tests were calculated by using two formulas: exp (Z x 2 1/2 x (CV(A)2 + CVw2)½/100) for RCF(UP) and (1/RCF(UP)) for RCF(DOWN). RCVs of these analytes were calculated as 14.63% for glucose, 29.88% for urea, 17.75% for ALP, 53.39% for CK, 46.98% for CK-MB, 21.00% amylase, 8.00% for total protein, 8.70% for albumin, 51.08% for total bilirubin, 86.34% for direct bilirubin, 6.40% for calcium, 15.03% for creatinine, 21.47% for urate, 14.19% for total cholesterol, 46.62% for triglyceride, 20.51% for HDL-cholesterol, 29.59% for AST, 46.31% for ALT, 31.54% for GGT, 20.92% for LDH, 19.75% for inorganic phosphate, 3.05% for sodium, 11.75% for potassium, 4.44% for chloride (RCV, p < 0.05, unidirectionally). We suggest using RCV as well as using population-based reference intervals in clinical laboratories. RCV could be available as a tool for making clinical decision, especially when monitoring individuals.
    Clinical laboratory 05/2015; 61(3-4):251-7. DOI:10.7754/Clin.Lab.2014.140906 · 1.13 Impact Factor