Consistency of Recall of Sexual and Drug-Using Behaviors for Homeless Persons with Dual Diagnosis

In this paper we present data from a study of the consistency of recall of several sexual and drug-using behaviors gathered from face-to-face interviews with homeless persons who have dual diagnosis (severe mental illness plus a substance use disorder). Factors associated with inconsistent reporting are also examined. One hundred and thirty-three individuals participating in a longitudinal study of HIV risk behaviors were recruited for a retrospective recall study. Participants were completing monthly interviews that included assessments of sexual behaviors and drug use. Participants completed an additional interview using the same instruments and were randomly assigned to a 3- or 6-month recall condition; the data from the additional interview were correlated with the regular, monthly interviews. Results indicated that reports of sexual and substance-using behaviors were generally reported consistently for both 3- and 6-month recall periods; however, coefficients for the 3-month interval were generally better than those for the 6-month interval. One exception was that protected intercourse was generally not reported consistently. Inconsistent reporting of sexual and drug-using behaviors was associated with the severity of substance abuse; inconsistent reporting of sexual behaviors was also associated with African-American race. Our data suggest that a 3-month recall interval is generally superior to a 6-month interval for persons with severe mental illness.