Objectives: The aim of this study is to determine the predictive value on rehospitalization of sociodemographic variables, positive/negative symptoms and thought disorders. The results are part of research project founded by the Basque Health Department.
Methods: A 18 month follow-up study of a cohort of 60 patients with acute exacerbation of schizophrenia was carried out. The assessment was performed with DSM III-R diagnostic criteria, PANSS and CGI rating scales, and SCID-P semistructured interview. All patients received antipsychotic treatment. The sociodemographic and disease data, the dimensional score of the PANSS subscales, the score of CGI scale, the items 2, 12, 13 and 14 of the PANSS as indicators of formal thought disorders; and the items 1, 5, 6, 17 and 23 of the PANSS as content thought disorders were established as predictors. The predictive value was determined by the Cox regression test (Lee 1992).
Results: We did not find predictive value either in the PANSS scores or in the 9 thought disorders evaluated (Wald and RR tests were not significative). Nevertheless, considering the values of standard error obtained in the Cox regression we were not in a position to assure that they did not have an incidence in the hospitalizations. The CGI was the only scale that showed prognostic value (Wald test = 1.9945; RR = 1.7499). Our results indicated that the lower number of previous hospitalizations (Wald test = 1.1437; RR = 1.1437) and the high level of studies (Wald test = 2.4258; RR = 1.8052) diminished the risk of rehospitalization.
Conclusions: 1 o The predictive value on rehospitalization for the positive/negative symptoms and thought disorders was not confirmed. 2 o CGI is the only scale with predictive value. That fact makes us consider the importance of what German psychiatrists called "smelling the schizophrenia" or "The smell of schizophrenia". 3 o Our results indicate that the lower number of previous hospitalizations, and the high level of studies diminish the risk of rehospitalization.