Thirteen studies (n=1,000) evaluating pharmacological interventions were included in the review: 5 double-blind RCTs (n=387), 2 open-label RCTs (n=86), 2 quasi-experimental studies (n=186) and 2 non-controlled prospective cohort studies (n=283) that evaluated antipsychotic agents and 2 RCTs (n=58) of other psychoactive agents.
Five studies (n=156) evaluated non-pharmacological interventions: 2 RCTs (n=91), 2 quasi-experimental studies (n=56) and one non-controlled prospective cohort study (n=9).
Pharmacological interventions.
Antipsychotic agents - double-blind RCTs (5 studies).
Conventional versus atypical antipsychotic agents (2 RCTs): one study reported that olanzapine was more effective and had less severe extrapyramidal symptoms than haloperidol; the other reported that clozapine and chlorpromazine were of similar efficacy.
Atypical antipsychotic agents versus each other (1 RCT): the study reported that olanzapine and risperidone had similar effects on cognitive and psychiatric measures, with the only difference in side-effects being greater weight gain with olanzapine.
Conventional antipsychotics versus each other (2 RCTs): these RCTs had small sample size and limited generalisability. The RCT that compared haloperidol with placebo did not report either side-effects or effectiveness.
Antipsychotic agents - open-label RCTs, quasi-experimental studies and large prospective single-agent studies. Methodological flaws included small sample sizes, lack of blinding and randomisation, absence of a control group and use of a retrospective design.
The open-label RCTs reported that olanzapine and risperidone were associated with less parkinsonism and improved negative and depressive symptoms in comparison with conventional antipsychotics (1 crossover RCT), and that olanzapine was associated with greater improvements on symptoms measures compared with haloperidol (1 RCT).
The results of the other studies were also reported.
Antidepressant augmentation of antipsychotic agents (2 RCTs): the RCTs reported improvements in some symptom measures with mianserin and trazodone compared with placebo (1 small double-blind RCT) and with citalopram compared with no citalopram (1 small single-blind RCT).
Non-pharmacological studies.
Three studies (2 RCTs and 1 non-controlled prospective cohort study) evaluated CBSST. The smaller pilot RCT reported greater improvements in positive and negative symptoms with CBSST compared with usual care. The other RCT reported significant improvements in some outcomes ('social functioning', cognitive insight and performance) but not others (symptoms, hospitalisations and living skills) in patients allocated to CBSST compared with usual care. The non-controlled cohort study evaluated only 9 patients.
One quasi-experimental study evaluated FAST and reported significant improvement in community functioning skills with FAST compared with usual care, but no difference in psychiatric symptoms between treatments.
One quasi-experimental study evaluated ST+HM and reported greater improvements in independent living skills and social functioning in patients allocated to ST+HM compared with HM alone.