Levels and patterns of alcohol consumption using timeline follow-back, daily diaries and real-time "electronic interviews".

Abstract

This study was designed to compare the Timeline Follow-Back (TLFB) to daily and real-time assessments of drinking. Our purpose was to evaluate overall correspondence and day-to-day agreement between these two methods among both problem and moderate drinkers. In Study 1, problem drinkers (n = 20) reported their alcohol consumption daily during 28 days of brief treatment. In Study 2, moderate drinkers (n = 48), recruited from the community, used a palm-top computer to record their drinking for 30 days. In both studies participants completed the TLFB covering the recording period. Participants in Study 1 reported fewer drinking days, fewer drinks per drinking day and fewer total drinks per day on the TLFB, and those in Study 2 reported fewer drinks per drinking day, fewer ounces per drinking day, fewer total drinks per day and fewer total ounces per day. The magnitude of the difference, however, was modest. There was considerable between-person variation in day-to-day correspondence of TLFB and the daily and real-time reports. Neither person characteristics (gender, education and income) nor the distributional characteristics of drinking (including average consumption, variation) predicted concordance between TLFB and real-time reports. The Timeline Follow-Back method captured overall levels of drinking quite well compared to a 28-day daily diary and a 30-day electronic interview. Vast individual differences in day-to-day correspondence suggest that the TLFB may be less useful for detecting patterns of consumption.