POSITIVE TEST RESULTS INDICATING DRUG USE IS A MISTAKE-RIDDEN INFERENTIAL PROCESS
Berry, D. A., & Chastain, L. (2004). Inferences about testosterone abuse among athletes. Chance, 17, 5-8.
This article compares the method of diagnosis used by the medical profession and that of the drugs-in-sport (DIS) movement to arrive at conclusions about a condition. It highlights what is required for correct identification. The basic thesis is that drug-test results alone are insufficient for correct inference. Testosterone [the steroid revealing the highest proportion of positive DIS test results] is the substance used to illustrate the problem as it applies to endogenous substances.
The basic problem with DIS testing is the "Failure to recognize relevant statistical issues when making inferences from laboratory tests leads to flawed conclusions in the case of an accusation of testosterone use" (p. 5-6).
Urinary testosterone concentrations fluctuate significantly in both genders. The comparison of variable appearances with a static ratio of testosterone to epitestosterone (until 2005 it was 6:1 and now is 4:1) will lead to false conclusions. The variations in endogenous testosterone are associated with "ethnicity, age, sex, circadian rhythm, training and competition, diet, nutritional supplementation, environmental factors, alcohol ingestion, enzyme deficiencies, decreased epitestosterone excretion, menstruation, pregnancy, birth control pills, other hormonal therapy, consumption of meat from animals supplemented with anabolic steroids, polycystic ovary syndrome (a common endocrine disorder), and other pathologic medical conditions" (p. 6). [Normally occurring levels in excess of 4:1 in a large sample of football players is ~10%. However, in sports where high natural levels of testosterone favor performance, the percentage is likely to be greater (Catlin, D. H., Hatton, C. K., & Starcevic, S. H. (1997). Issues in detecting abuse of xenobiotic anabolic steroids and testosterone by analysis of athletes' urine. Clinical Chemistry, 43, 1280-1288).]
"Any laboratory can give false positive results, whether because of inherent variability in the testing procedure, inherent biologic variability, or inappropriate handling of source material" (p.7). Thus, to declare laboratory test results infallible will result in false positives. "It seems unreasonable to impugn 1% or even a tenth of a percent of nonusers who happen to be tested" (p. 7).
"The diagnostic accuracy of any laboratory test is defined as the ability to discriminate between two types of individuals – in this case users and nonusers. Specificity and sensitivity characterize diagnostic tests. In the anti-doping analogy regarding the T/E ratio, specificity is the proportion of nonusers with T/E <6 . . . and sensitivity is the proportion of users with T/E <6".(p. 7). Those proportions can be estimated.
Sensitivity must be addressed in the context of the effect of a particular banned substance upon the T/E ratio. DIS laboratories fail to address sensitivity or recognize its relevance. "Specificity, generally addressed by keeping records of tests performed on athletes, introduces a clear bias of unknown magnitude because some athletes in their databases may be users but have 'normal' T/E. For example, Asian males have baseline urine concentrations of testosterone and T/E that can be so low that testosterone doping many not elevate their T/E to that of some nonusers among Caucasian athletes". (p/ 7).
"Conclusions about the likelihood of testosterone doping require consideration of three components: specificity and sensitivity of the testing procedure, and the prior probability of use. DIS officials consider only specificity. The result is a flawed process of inference" (p. 8). Thus, the assumption of a positive test result in the DIS scheme indicating drug use is a mistake-ridden inferential process.
Implication. The assumption of a positive test result in the DIS scheme indicating drug use is a mistake-ridden inferential process. DIS testing does not adhere to the standards of testing and interpretation that are required of the medical and pharmaceutical professions. Yet, it is allowed to "destroy" innocent athletes.
Return to Table of Contents for this issue.