Popular PARADE Columnist Refutes Accuracy of Urinalysis
Marilyn Vos Savant, a popular columnist for the widely read weekly Parade provides convincing statistical evidence that drug testing, even at a 99 percent accuracy level, can ultimately do no better than a coin flip in determining whether or not someone is really a drug user (Marilyn Vos Savant, "Ask Marilyn: Tests for AIDS and Drugs: How Accurate Are They? Parade, 3/28/93, p. 24).
Vos Savant goes into great depth as to why, once an individual has tested positive, the odds that he or she is actually a drug user are only 50-50. Even with the most sophisticated methods, a great margin for error exists. This is a result mainly of analytic sensitivity (the rate of positive test results in people who are actually users) and analytic specificity (the rate of negative test results in people who are actually not users). These factors are biochemical, not statistical, hence laboratory error, although a possibility, is not the primary concern.