Premium
Long Term Outcome for Patients with Drinking Problems: the search for predictors
Author(s) -
EDWARDS GRIFFITH,
BROWN DAVID,
OPPENHEIMER EDNA,
SHEEHAN MARGARET,
TAYLOR COLIN,
DUCKITT ANNIE
Publication year - 1988
Publication title -
british journal of addiction
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.424
H-Index - 193
eISSN - 1360-0443
pISSN - 0952-0481
DOI - 10.1111/j.1360-0443.1988.tb01584.x
Subject(s) - outcome (game theory) , matching (statistics) , test (biology) , psychology , term (time) , clinical psychology , medicine , mathematics , pathology , ecology , physics , mathematical economics , quantum mechanics , biology
Summary Previous work on prediction of outcome among patients with drinking problems is briefly reviewed. The literature suggests that few predictors of consistent strength have been identified. Two reasons frequently put forward for conducting such investigations are the belief that findings may be useful to clinical policies either by indicating which patients should be excluded from treatment or by assisting in matching. At present such amibitions seem premature. The separation between theory‐rich research on community samples and often very barely descriptive studies on clinical populations should be overcome. With the issues raised by this review taken as the background, results are then presented on a 10‐year follow‐up of a group of 99 married male alcoholics seen at the Maudsley Hospital. The usual predictive approach of running individual Intake variables against a range of Outcome variables reveals little of significance and actual drinking behaviour cannot be predicted at all. When, however, outcome is measured in terms of dimensions deriving from a principal components analysis, a number of very significant predictors are found which bear differentially on the two components. Prediction research may provide a test‐bed for theory but advances will depend on the development of ideas on the nature of outcome processes.