Drug Testing and Unintended Consequences

Via Hit & Run, I find this article from the American Academy of Pediatrics arguing against random drug testing of students and employees. I was surprised to find relatively little on the problem of false positives. Here’s what the article says on that score:

School staff members and/or parents need to be able to assess possible false-positive results, especially when screening test results are positive for amphetamines or opioids. Over-the-counter cold medications containing pseudoephedrine can cause false-positive screening results for amphetamine, although follow-up testing with gas chromatography and mass spectrometry is highly specific and can reliably confirm the presence of amphetamine. Ingestion of foods that contain poppy seeds makes interpretation of drug testing more difficult, because it can cause screening and gas chromatography and mass spectrometry results to be falsely positive for morphine and/or codeine.

Unfortunately, the article doesn’t report the percentage of positive test results that don't actually involve drug use. For some (possibly outdated) calculations like that, see here. Still, the article makes a number of strong arguments against random drug testing, including this one showing the AAP understands the importance of incentives:

With the exception of marijuana, the window of detection for most drugs of abuse is 72 hours or less. ... Standard drug-testing panels also do not detect many of the drugs most frequently abused by adolescents, such as alcohol, ecstasy ..., and inhalants, and information on the limitations of screening tests and ways to defeat them is widely available to adolescents via the Internet. Widespread implementation of drug testing may, therefore, inadvertently encourage more students to abuse alcohol, which is associated with more adolescent deaths than any illicit drug but is not included in many standard testing panels. Mandatory drug testing may also motivate some drug-involved adolescents to change from using drugs with relatively less associated morbidity and mortality, such as marijuana, to those that pose greater danger (eg, inhalants) but are not detected by screening tests. [emphasis added]

In short, drug tests are most effective at detecting the least harmful drugs, so implementation of random drug testing could induce some drugs users to substitute into more harmful drugs to avoid detection. Nice.