Typical and atypical antipsychotics--the misleading dichotomy. Results from the Working Group 'Drugs in Psychiatry'...

(1) To investigate the risk of extrapyramidal motor side effects (EPS) associated with the prescription of different antipsychotics under naturalistic treatment conditions; (2) to test the rationale of the terms 'typical' and 'atypical' based on EPS rates.

Cross-sectional study in the federal state of Bavaria.

20 psychiatric hospitals in Bavaria.

6,061 inpatients, aged 18-65 years, with psychotic disorders.

Co-medication with the anticholinergic biperiden was used as an index of EPS. Odds ratios for EPS and numbers needed to harm [number of patients who would need to be treated to obtain one more case with an adverse outcome (i.e. EPS) as compared with the control treatment (clozapine)] were calculated to obtain risk estimates for 15 different antipsychotics.

Groups of 'typical' and 'atypical' antipsychotics were not homogeneous in their EPS rates, and showed wide variation within each group. Nor did the frequency of EPS allow a clear distinction between the groups. There were 2 reasons for this: first, EPS rates rose continuously over the whole spectrum of drugs under study, and therefore precluded the definition of a cut-off score; second, there was considerable overlap between the 2 groups as EPS rates of various 'atypicals' (e.g. amisulpride, risperidone and zotepine) did not differ from some 'typical' substances (e.g. fluphenazine), while one 'typical' antipsychotic (perazine) even had a lower EPS risk than most 'atypicals'.

The odds of inducing EPS are not distinguishable between 'typical' and 'atypical' antipsychotics as EPS rates rise on a continuous scale throughout both classes. We propose dropping the categorization of antipsychotics as 'typical' and 'atypical' and instead using risk estimates like number needed to harm for EPS to help in benefit/risk considerations for antipsychotic treatment.