Effects of interactive voice response self-monitoring on natural resolution of drinking problems: utilization and...

Most problem drinkers do not seek help, and many recover on their own. A randomized controlled trial evaluated whether supportive interactive voice response (IVR) self-monitoring facilitated such "natural" resolutions. Based on behavioral economics, effects on drinking outcomes were hypothesized to vary with drinkers' baseline "time horizons," reflecting preferences among commodities of different value available over different delays and with their IVR utilization.

Recently resolved untreated problem drinkers were randomized to a 24-week IVR self-monitoring program (n = 87) or an assessment-only control condition (n = 98). Baseline interviews assessed outcome predictors including behavioral economic measures of reward preferences (delay discounting, pre-resolution monetary allocation to alcohol vs. savings). Six-month outcomes were categorized as resolved abstinent, resolved nonabstinent, unresolved, or missing. Complier average causal effect (CACE) models examined IVR self-monitoring effects.

IVR self-monitoring compliers (≥70% scheduled calls completed) were older and had greater pre-resolution drinking control and lower discounting than noncompliers (<70%). A CACE model interaction showed that observed compliers in the IVR group with shorter time horizons (expressed by greater pre-resolution spending on alcohol than savings) were more likely to attain moderation than abstinent resolutions compared with predicted compliers in the control group with shorter time horizons and with all noncompliers. Intention-to-treat analytical models revealed no IVR-related effects. More balanced spending on savings versus alcohol predicted moderation in both approaches.

IVR interventions should consider factors affecting IVR utilization and drinking outcomes, including person-specific behavioral economic variables. CACE models provide tools to evaluate interventions involving extended participation.