Seven RCTs (n=419) were included. The sample size ranged from 31 to 87.
One study was judged to be at low risk of allocation concealment bias and six at moderate risk. All studies were described as randomised, but only two reported the randomisation method. Six reported double-blinding and one reported blinding of the rater only. All studies accounted for early withdrawals but only five reported reasons for withdrawal. Two studies reported intention-to-treat analysis and three reported completer analysis.
Clozapine versus placebo (2 trials, n=120): there was no significant difference between clozapine and placebo in the proportion of early withdrawals. Clozapine was associated with a significant improvement in the CGI (WMD -1.1, 95% CI: -1.24, -0.97) and UPDRS total and motor score (WMD -2.39, 95% CI: -3.58, -1.20 and WMD -1.74, 95% CI: -2.57, -0.92, respectively) compared with placebo. Both studies also reported a significant improvement in psychotic symptoms with clozapine compared with placebo when using different scales.
Clozapine versus quetiapine (1 trial, n=40 completed): there were no significant differences between clozapine and quetiapine for clinical efficacy, motor functioning, or adverse events.
Quetiapine versus placebo (2 trials, n=89): there were no significant differences between quetiapine and placebo in efficacy, safety, or the proportion of early withdrawals.
Olanzapine versus placebo (2 trials, n=170): olanzapine was associated with a significantly increased risk of early withdrawal (RR 2.11, 95% CI: 1.13, 3.92), particularly due to adverse events (RR 7.18, 95% CI: 1.76, 29.24), and a significant worsening of Parkinson’s symptom (UPDRS, SMD 0.59, 95% CI: 0.40, 0.78). There was no significant difference between olanzapine and placebo in clinical efficacy or the MMSE.