Background: A recent meta-analysis rejected the "delayed onset of antipsychotic action hypothesis" that had been described in textbooks for decades. Since meta-analyses are prone to a number of methodological problems, we attempted a replication by a) using a large database of individual patient data rather than meta-analysis, b) including another antipsychotic and c) extending the analysis from four weeks to one year.
Methods: We pooled the data of seven randomized trials involving amisulpride. The data included 1708 patients with schizophrenia and positive symptoms and we examined the incremental percentage Brief Psychiatric Rating Scale (BPRS) reduction over time.
Results: The "early onset of antipsychotic action hypothesis" was confirmed, as the reduction of overall and positive symptoms until week two was larger than the additional reduction until week four (p<.0001). Furthermore, in a subset with long-term data (n=748) approximately 68% of the mean BPRS change at one year was already achieved at four weeks in the observed cases.
Conclusions: A substantial amount of the antipsychotic drug effect seems to occur during the first weeks of treatment. Subsequent analyses are needed to establish how long an antipsychotic should be tried before it is considered ineffective and alternative strategies implemented.