Predicting the Trajectory of Absence-Drinking Dyads

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): The application proposes a secondary analysis of outcome from treated alcoholics. It will use data from Project MATCH, a large-scale multi-site clinical trial of treatments for alcoholism completed in 1996, to test several models for analyzing patterns of abstinence and drinking behaviors over time. The MATCH data contains for each participant retrospective self-reports of daily consumption over a period of 455 days from treatment entry. This data (N=1726) will address research aims concerned with the identification of behaviors defined as abstinence/drinking dyads, where a dyad is the sequence of 1 or more days of not drinking (abstinence) and 1 or more days of drinking. Concurrently, it will be possible to identify a participant's pattern of abstinence/drinking dyads over the follow-up period as post-treatment behavior trajectories. Because it is feasible to suppose that there will be similarities among the post-treatment drinking behavior trajectories (continual abstinence is 1 common possibility), it is reasonable to attempt to identify homogeneous groupings of these trajectories. Subsequently, client characteristics and environmental factors (e.g. treatment conditions) can be employed in the attempt to discriminate between the classes of grouped trajectories. The general hypothesis is that this method will yield a better understanding of a person's change in established patterns of drinking over time and result in a better understanding of the relative impact of alternative treatment strategies. The application presents an innovative way of conceptualizing alcoholism treatment outcome. Methods commonly employed currently treat outcome from alcoholism treatment as either a binary event (abstinent-drinking) or use indicators of frequency (e.g. percent (%) days abstinent) or intensity (e.g. drinks on a drinking day) averaged over the follow-up period. Both of these approaches to outcome fail to recognize and represent the trial and error involved in the transition from an old behavior (e.g. drinking) to a new behavior (e.g. abstinence). The MATCH data provides an opportunity to develop models that better represent what is observable in clinical trials with long-term outcome follow-up phases, multiple attempts at abstinence. Most importantly these objectives can be achieved at minimum cost.