Biostatistician Says Athletes' Doping Tests Flawed

A US biostatistician, who testified for runner Mary Decker in a 1997 doping hearing, believes that doping tests among athletes are inherently flawed. In the latest issue of Nature, Donald Berry cites a lack of validated sensitivity and specificity data for individual doping tests—data which are necessary to know the accuracy of positive tests.

Berry argues that, even when a test has a very high specificity (eg, 95%), the false-positive rate in a non-doper who provides 8 different urine samples remains remarkably high (eg, 34%).* A comparable false-positive rate would also apply to the large number of samples assayed by a particular doping laboratory. Invariably some percentage of these tests in non-dopers would turn up positive and, consequently, ruin careers. Berry concludes that doping labs "must prospectively define and publicize a standard testing procedure, including unambiguous criteria for concluding positivity...and validate that procedure in blinded experiments."

The flip side of Berry's argument, the reliability of negative doping tests, was explored recently by Danish investigators, who showed that the detection of recombinant human erythropoietin (rHuEPO) in intentionally doped volunteers was astonishingly inconsistent within and between official doping labs.

* 1 – 0.95 to the 8th power = 1 – 0.66 = 0.34.

Photo: iStockPhoto