Katharine Dudley, Xiaomeng Liu, Saskia De Haan

The World Health Organization (WHO) Model Lists of Essential Medicines lists chlorpromazine as one of its five medicines used in psychotic disorders. To determine chlorpromazine dose response and dose side-effect relationships for schizophrenia and schizophrenia-like psychoses. We searched the Cochrane Schizophrenia Group's Study-Based Register of Trials (December 2008; 2 October 2014; 19 December 2016). All relevant randomised controlled trials (RCTs) comparing low doses of chlorpromazine (≤ 400 mg/day), medium dose (401 mg/day to 800 mg/day) or higher doses (> 800 mg/day) for people with schizophrenia, and which reported clinical outcomes. We included studies meeting review criteria and providing useable data. Review authors extracted data independently. For dichotomous data, we calculated fixed-effect risk ratios (RR) and their 95% confidence intervals (CIs). For continuous data, we calculated mean differences (MD) and their 95% CIs based on a fixed-effect model. We assessed risk of bias for included studies and graded trial quality using GRADE (Grading of Recommendations Assessment, Development and Evaluation). As a result of searches undertaken in 2014, we found one new study and in 2016 more data for already included studies. Five relevant studies with 1132 participants (585 are relevant to this review) are now included. All are hospital-based trials and, despite over 60 years of chlorpromazine use, have durations of less than six months and all are at least at moderate risk of bias. We found only data on low-dose (≤ 400 mg/day) versus medium-dose chlorpromazine (401 mg/day to 800 mg/day) and low-dose versus high-dose chlorpromazine (> 800 mg/day).When low-dose chlorpromazine (≤ 400 mg/day) was compared to medium-dose chlorpromazine (401 mg/day to 800 mg/day), there was no clear benefit of one dose over the other for both global and mental state outcomes (low-quality and very low-quality evidence). There was also no clear evidence for people in one dosage group being more likely to leave the study early, over the other dosage group (moderate-quality evidence). Similar numbers of participants from each group experienced agitation and restlessness (very low-quality evidence). However, significantly more people in the medium-dose group (401 mg/day to 800 mg/day) experienced extrapyramidal symptoms in the short term (2 RCTS, n = 108, RR 0.47, 95% CI 0.30 to 0.74, moderate-quality evidence). No data for death were available.When low-dose chlorpromazine (≤ 400 mg/day) was compared to high-dose chlorpromazine (> 800 mg/day), data from one study with 416 patients were available. Clear evidence of a benefit of the high dose was found with regards to global state. The low-dose group had significantly fewer people improving (RR 1.13, 95% CI 1.01 to 1.25, moderate-quality evidence). There was also a marked difference between the number of people leaving the study from each group for any reason, with significantly more people leaving from the high-dose group (RR 0.60, 95% CI 0.40 to 0.89, moderate-quality evidence). More people in the low-dose group had to leave the study due to deterioration in behaviour (RR 2.70, 95% CI 1.34 to 5.44, low-quality evidence). There was clear evidence of a greater risk of people experiencing extrapyramidal symptoms in general in the high-dose group (RR 0.43, 95% CI 0.32 to 0.59, moderate-quality evidence). One death was reported in the high-dose group yet no effect was shown between the two dosage groups (RR 0.33, 95% CI 0.01 to 8.14, moderate-quality evidence). No data for mental state were available. The dosage of chlorpromazine has changed drastically over the past 50 years with lower doses now being the preferred of choice. However, this change was gradual and arose not due to trial-based evidence, but due to clinical experience and consensus. Chlorpromazine is one of the most widely used antipsychotic drugs yet appropriate use of lower levels has come about after many years of trial and error with much higher doses. In the absence of high-grade evaluative studies, clinicians have had no alternative but to learn from experience. However, such an approach can lack scientific rigor and does not allow for proper dissemination of information that would assist clinicians find the optimum treatment dosage for their patients. In the future, data for recently released medication should be available from high-quality trials and studies to provide optimum treatment to patients in the shortest amount of time.