Drug Report Validity of Formerly Homeless vs. Never Homeless Adults

The trauma and stress of homelessness is believed to influence individuals long after they attain permanent housing. We investigate the degree to which prior homeless experiences may influence the accuracy of self-reports obtained in a substance use survey. Data to do so come from two community surveys in Chicago, Illinois that employed probability sampling methods and audio computer-assisted self-interviews (ACASI). One survey was of the general population aged 18-40 and the second survey was of men-who-have-sex-with-men (MSM) aged 18-55. Using logistic regression models that adjusted for the complex design of each survey sample, the effects of self-reported homeless experiences on drug use concordance, measured by agreement between drug use reports and biological assays, were investigated. In the general population sample, 14.8% reported ever having been homeless, and 19.9% of the MSM sample reported a lifetime homeless experience. Logistic regression models that adjusted for demographic variables revealed that, in each sample, having ever been homeless was not associated with past 30-day cocaine reporting accuracy. Ever having slept in public spaces, however, was predictive of lower cocaine use reporting accuracy in both the general population (OR = 0.13, 95% CI = 0.04-0.39) and MSM (OR = 0.11, 95% CI = 0.02-0.59) samples. We conclude that traumatic homeless experiences may be another source of measurement error when assessing substance use behavior in epidemiologic surveys.