[Diagnostic variability in a cohort of patients with multiple admissions in the last two decades]

To investigate possible changes in the admission pattern correlated with improvement in psychiatric attention and if there is variability in the diagnosis of patients previously identified as having multiple admissions during the study period.

Our study is based on fifty-nine patients who were admitted on at least ten occasions between 1983-2000. The medical records of the frequent users were reviewed and data were extracted on several variables: principal psychiatric diagnosis (DSM-IV TR criteria) and diagnostic changes as main variables and period, admissions density, admission interval and period between last admission and the end of study, as related variables.

Friedman's variance non-parametric analysis for related samples, Pearson's chi-squared test and sequence graphing.

Improved psychiatric care in our city would seem to correlate with the decreasing number of patients with multiple admissions to the point where there have been no admissions in the past year. Diagnosis (DSM-IV axis I and II) variables showed important variations from one admissions to the next. The psychosocial and adaptation problems that became manifest during the successive admissions tended to improve over time as did the overall functioning of the patients. Our findings suggest that this population is composed of three distinct subgroups. Identification of the subgroups may have important implications for the management and treatment modality.