Self-reports are the primary method for collecting data on alcohol use and risky sexual practices among men-who-have-sex-with-men (MSM). The TimeLine FollowBack method (TLFB) has been considered the gold standard for collecting data on alcohol use and risky sexual behavior. Interactive Voice Response (IVR) technology has become increasingly popular as a new data collection method and offers several advantages. Given this background, three aims were posited for the present study that was conducted among a sample of problem drinking MSM to: (1) examine the correspondence between the IVR and TLFB methods for the assessment of alcohol use and risky sexual behavior; (2) identify factors that influence correspondence between methods for select alcohol use and risky sex variables; and, (3) examine the conditional relationship between alcohol use and engagement in unprotected anal intercourse (UAI) using IVR daily data.

Participants (N = 84) were problem drinking MSM, who were participating in a combined pharmacotherapy and psychotherapy clinical trial to moderate their alcohol consumption. Participants reported on their alcohol use and sexual risk behavior daily for 90-days using the IVR system. At treatment completion, participants were administered the TLFB interview and reported on their behavior during the same 90 day time frame. Study findings revealed moderate correlations between the IVR and TLFB methods for each of the alcohol use and sexual risk variables. T-tests indicated greater aggregate reports of drinking on the IVR, whereas greater aggregate reports for the sexual risk variables were generated from the TLFB method. A visual inspection of the limits of agreement indicated substantial individual variation for self-reports between methods across each of the alcohol use and sexual risk variables. Exploratory analyses revealed that TLFB-IVR correspondence for variables specific to alcohol use was affected by participant’s alcohol dependence severity, daily negative affect, and number of standard drinks consumed prior to or during data collection on the IVR. Multilevel analyses indicated that the risk of engaging in UAI was greater on days in which any alcohol use was consumed by the participant, regardless of the quantity of use. Findings are discussed regarding the utility of IVR data collection technology among MSM.