DSM-III mental disorders in general medical sector: predictors of outcome

In 1987, a multicentric research was carried out in Emilia-Romagna, an administrative area in northern Italy. Nine hundred and forty one people in the general medical sector were contacted to enter a two-stage procedure to identify DSM-III cases. Cases were recontacted for 3-month, 1-year and 2-year follow-ups. Eighty cases completing at least the one-year follow-up were analysed. The sample was split into "still-cases" and "no longer-cases" depending on the persistence of a DSM-III diagnosis at the last follow-up. Both clinical and social features proved to be predictors of the outcome. A disorder lasting one year or longer, past recurrences, Axis I comorbidity, older age, being married and having children were the features which at the first assessment predicted an unfavourable outcome. Having children and past episodes proved to be the best predictors of remaining a case in a stratified logistic regression model. The persistence of major difficulties during the follow-up period was also correlated to a "still-case" status.