Motor features in psychotic disorders. II. Development of diagnostic criteria for catatonia.

Existing diagnostic criteria for catatonia have been based exclusively on theoretical assumptions. The present study aimed to develop empirical criteria for diagnosing catatonia. The same patient population (n=187) described in Part I (Peralta, V., Cuesta, M.J., Motor features in psychotic disorders. I. Factor structure and clinical correlates. Schizophr. Res., 00, 000-000) was used in this study. Fourteen catatonic signs with potential diagnostic value were cluster-analyzed to derive groups with and without a catatonic syndrome. Taking the cluster solution as the reference standard, the diagnostic value for individual signs was examined by means of multiple discriminant analysis, ROC analysis, and the parameters of sensibility, specificity, positive predictive power and negative predictive power. Cluster analysis divided the sample into a catatonic group (n=32) and a noncatatonic group (n=155). Discriminant analysis showed that 11 of the 14 potential diagnostic signs discriminated among groups: immobility/stupor, mutism, negativism, oppositionism, posturing, catalepsy, automatic obedience, echophenomena, rigidity, verbigeration and withdrawal. ROC analysis showed that any combination of three or more of these symptoms best fitted the cluster-derived catatonic syndrome. In conclusion, patients displaying three or more classical motor signs may be diagnosed with confidence as suffering from a catatonic syndrome.