Forty-one studies, involving 2,539 participants in total, were included. Nineteen studies were described as double-blind (13 of which were placebo-controlled) and the remaining 22 were described as open label, most of which had no control group.
Disruptive behavioural disorders: 12 studies (1,389 participants) were identified. Six double-blind controlled studies found consistently that risperidone was better than placebo at reducing symptoms. Where calculable, the effect size ranged from -0.6 to -1.5 (3 studies). Further studies showed that significant behavioural improvement was seen within 1 to 2 weeks of commencing treatment, and that efficacy was maintained long term. One very small open-label uncontrolled study showed a positive response to olanzapine within 2 weeks.
Pervasive developmental disorders: 10 studies (548 participants) were identified. Five double-blind studies demonstrated greater efficacy in reducing symptoms, or preventing relapse, for risperidone compared with placebo. Where calculable, the effect size ranged from -0.8 to -1.2 (3 studies). Further studies reported longer term efficacy of risperidone, and one study reported benefits of treatment with olanzapine.
Tic disorders: 4 studies (84 participants) were identified. Two double-blind studies found significantly greater improvements in tic severity with risperidone compared with pimozide or placebo. Risperidone was also better at improving obsessive compulsive traits than clonidine (1 double-blind study). One double-blind study demonstrated ziprasidone to be more effective at reducing tic severity than placebo.
Schizophrenia and related disorders: 6 studies (190 participants) were identified. One double-blind study found clozapine to be better at improving symptoms than haloperidol, while another reported similar efficacy of haloperidol, risperidone and olanzapine. Open-label studies reported treatment responses for clozapine, olanzapine and risperidone.
Mania in bipolar disorder: 6 studies (179 participants) were identified. One double-blind study found quetiapine plus divalproex to give superior control of mania in adolescents than divalproex plus placebo, without an increase in major side-effects, although there was a higher drop-out rate amongst those on quetiapine. Further open-label studies have also reported efficacy of olanzapine and risperidone.
Safety and tolerability: 3 studies (149 participants), plus safety data from treatment efficacy studies, were identified. In general, risperidone, clozapine and olanzapine were well-tolerated; the most common side-effects were sedation and weight gain. One open-label study found that most weight gain with risperidone occurred in the first 2 months of treatment. One study suggested that this weight gain is reversible. Extrapyramidal symptoms were not often noted with atypical antipsychotics, nor were cognitive adverse events. Some serious adverse events were reported, particularly haematological adverse events with clozapine.