Acessibilidade / Reportar erro

Cluster Analysis and Heuristic Analysis in building a typology of alcoholics

Objectives

To develop and to describe a binary typology of alcoholics.

Methods

Data derived from 329 patients followed in an observational study. Correspondence Analysis was applied to identify the least number of variables capable of representing the maximum variability of the subjects; Cluster Analysis was performed to identify two subgroups; and Heuristic Analysis performed with the objective of organizing the variables into groups of meaning. Bivariate Analyses were used to test possible associations of Clusters to 51 variables – demographic or of clinical relevance.

Results

Correspondence Analysis derived 20 variables, afterwards categorized into three groups of meaning through Heuristic Analysis: “behaviors”, “beliefs” and “feelings”. The results indicate that subgroup 1 patients present a more severe clinical, psychopathological and social profile. Identified feelings and beliefs were associated to drinking and relapsing variables, and behaviors to social repercussions.

Conclusions

The results of this study suggest that the typology here presented may contribute to the treatment of these alcoholics subtypes by means of therapeutic strategies adapted to the identified characteristics of each subgroup of patients.

Alcoholism; typology; diagnosis; cluster analysis; heuristic analysis


Instituto de Psiquiatria da Universidade Federal do Rio de Janeiro Av. Venceslau Brás, 71 Fundos, 22295-140 Rio de Janeiro - RJ Brasil, Tel./Fax: (55 21) 3873-5510 - Rio de Janeiro - RJ - Brazil
E-mail: editora@ipub.ufrj.br