Point: The Quest for Clean Competition in Sports: Are We the Dopes?

A 2011 Q&A article published in Clinical Chemistry (1) covered the recent state of affairs and the outlook for future efforts to combat the use of illicit or performance-enhancing drugs in sports (i.e., doping). The interviewees in that article made several poignant admissions: that the rate of detection for doping is <100% (and maybe far less), that deterrent measures need to be stronger yet fair, and that future additions to the pharmacopoeia of doping agents will seriously challenge our ability to provide reliable and comprehensive testing.Given these challenges, it may be time for our society to reconsider the purpose of antidoping testing, and for those of us in the clinical chemistry community to question whether we should be doing this testing at all.One of the most important principles of laboratory testing is the statistical concept implied by the Bayes theorem: the likelihood of a finding being true is related to the pretest probability of truth times the likelihood ratio associated with the test. Thus, in the case of a drug test, if it is tremendously unlikely that the subject is using the drug, a positive drug test result is most likely a false positive. Conversely, if a subject is highly likely to be using a drug, a negative drug test result is most likely a false negative. Consider, for example, a hypothetical screening population in which 90% of athletes use a banned substance. Bayesian analysis tells us that, even if the test used for the banned substance is 95% sensitive and 95% specific for detection, more than 32% of the negative tests are false negatives. Because antidoping testing is often designed to afford higher specificity at the cost of lower sensitivity to lower the probability of false-positive results, and because athletes are known to actively evade detection, the …