[Assessment of needs and evaluation of interventions in mental health. Current approach].

For needs assessment research to contribute to mental health program planning and evaluation, the term "need" requires better conceptualization and operationalization. Recent advances in this area are reviewed. The two major American contributions focus on "true prevalence" of need, but rely heavily on simplistic definitions. The major British methodology, developed on treated samples, operationalizes "need" as multidimensional and dynamic. However, its feasibility as a methodology in general populations remains to be demonstrated. Findings from recent mental health consumer preference surveys suggest that a multi-partite approach might be most productive for planning and evaluation.