Search

Abstract: Although we are in the era of evidence-based medicine, there is still a substantial gap between theory and current practice with the application of reference intervals as decision making tools. Different laboratories may have different reference intervals for the same tests using the same analytical methods and platforms. These differences have the potential to confuse physicians making the assessment and monitoring of patients more difficult by providing discordant information. This paper attempts to demonstrate how to use evidence-based approach for harmonising reference intervals. In order to consider harmonisation we must first have an appreciation of the various factors that influence the determination of that reference interval such as the choice of individuals within the population studied, biological variability of the analyte studied, partitioning, sample collection, analytical aspects such as bias and statistical models. An a priori approach for determining reference intervals, whilst recommended, may be beyond the scope of most laboratories and consideration should be given to the use of a validated indirect a posteriori approach. Regardless of method used, the continuing application of an evidence-based approach in harmonised reference intervals to meet the quality expectations of physicians should be pursued. [Copyright &y& Elsevier]

Objective: To evaluate the clinical usefulness for the measurement of testosterone (T) values in the “low” female range with a direct radioimmunoassay (RIA) for total T by comparing total T values measured by this assay with values determined by conventional RIA after organic solvent extraction/column chromatography. Design: Cross-sectional study. Setting: Victoria, Australia. Patient(s): Two hundred fifty-nine healthy women, aged 18–75 years, recruited from the community. Intervention(s): Fasting serum samples were obtained and stored at −80°C. Main Outcome Measure(s): Total T measurement. Total T was measured by the direct RIA method using antibody-coated tubes and iodine-labeled T tracer. For comparison, total T levels were also measured using the conventional RIA method after organic solvent (ethyl acetate:hexane [3:2]) extraction and celite column partition chromatography before RIA. Result(s): The mean T level by direct RIA was 0.76 nmol/L (median, 0.70; SD, 0.54; minimum, 0.10; maximum, 3.2). The mean difference between the two measurements (direct RIA − conventional RIA) was −0.28 (SD, 0.3). The direct RIA value was 63% (95% confidence interval, 26%–155%) of the conventional RIA estimate. For the classification of values at the lower end of the range, there was very good agreement beyond chance (kappa = 0.68–0.74) for values in approximately the lowest 10th–40th percentiles. Conclusion(s): The direct RIA is a clinically useful assay for the study of the issue of “low” T within the female population. [Copyright &y& Elsevier]