Impact of testosterone assay standardization efforts assessed via accuracy-based proficiency testing.

BACKGROUND:We reported observations on analytical performance in testosterone measurements of various methods/assays from the study carried out using accuracy-based proficiency testing (PT) during 2012-2013. In 2016, we re-evaluated analytical performance of testosterone assays using accuracy-based PT to assess effectiveness of CDC efforts toward standardization. METHODS:Five single-donor human serum samples from female and male adult donors were analyzed for testosterone by New York State Department of Health-certified clinical laboratories using 16 immunoassays and LC-MS/MS methods. Target values were determined using the CDC reference measurement procedure. RESULTS:Testosterone targets for the 5 samples were 43.5, 160, 294, 457, and 534 ng/dL. The biases of individual result of the 65 participant laboratories against the target for each sample were calculated. Of participants, 87.7% had ≥4 of the 5 results within the minimum allowable total error limits (± 25.1%), a 14.7% increase from the previous study. The improved PT scores were attributed to better analytical accuracy and precision, and laboratories' selection of more accurate assays/methods. CONCLUSIONS:Improved analytical accuracy and precision for testosterone assays were demonstrated over a 3.5-year period after the first CDC-directed accuracy-based proficiency testing. Additional effort is needed to improve accuracy/precision of measurements, especially at low concentrations.