The validity of drug use responses in a household survey in Puerto Rico:: comparison of survey responses of cocaine and heroin use with hair tests

Background Despite the widespread use of household surveys to assess the epidemiology of illicit drug use and abuse, there is very little information about the willingness of respondents to disclose their use of drugs in household studies outside the US. Methods As part of a household study of substance use disorders in Puerto Rico, we collected hair specimens from a sub-sample of 114 respondents. Hair specimens were screened using a radio immunoassay. Screened-positive specimens were confirmed using gas chromatography/mass spectrometry. Results Using hair-test results as the standard, specificity of self-reports was 98% or higher for both drugs. The sensitivity of all self-reports was low, although lifetime use reports had somewhat higher sensitivities. The sensitivity of self-reports of recent cocaine use was particularly low, 7.1%. The sensitivity of heroin use reports was somewhat higher, 33.3% for recent use and 66.7% for lifetime use. The estimate of recent cocaine use based on hair tests was 13.7 times the estimate generated from interview reports. For heroin use, the test-based estimate was 2.9 times the rate generated from the interview reports. A shift from the cut-off level of 0.2 ng/mg to 0.5 ng/mg had only a marginal improvement on validity, with sensitivity increasing from 7.1% to 11.1% for self-reported recent cocaine use. Conclusions The results suggest that drug users, for the most part, are not willing to disclose their use of drugs in household surveys in Puerto Rico. Methods to increase the willingness of respondents to disclose their use of drugs are needed.