Antipsychotics for behavioural and psychological problems in elderly people with dementia: a systematic review of adverse events.

OBJECTIVE: Although antipsychotics are important in the treatment of behavioural and psychological symptoms of dementia (BPSD), they have moderate efficacy and often cause adverse events. Recent safety warnings about increased frequency of cerebrovascular adverse events in elderly patients who use atypical antipsychotics mean that physicians now face a dilemma when weighing the benefits and risks of use of antipsychotics in this patient group. This study systematically reviews the reporting of adverse events of antipsychotics used to treat BPSD in randomised, controlled trials (RCTs).
METHODS: We searched the MEDLINE, EMBASE, PsychInfo and CINAHL databases (search period 1980 or 1986-April 2005) and the Cochrane controlled trials register (2005) for RCTs that used intention-to-treat analysis to evaluate the efficacy and harms of antipsychotics used to treat BPSD. Two independent reviewers assessed the reporting of adverse events.
RESULTS: Screening of 930 abstracts identified 12 eligible RCTs (2809 patients). Most participants were elderly people (mean age 80 years) with Alzheimer's, vascular or mixed dementia. Studies lasted from 3 to 16 weeks. Adverse events, though common, were described heterogeneously and incompletely. No RCT fulfilled all Consolidated Standards of Reporting Trials requirements for the reporting of harms. Atypical antipsychotics caused fewer extrapyramidal symptoms and less somnolence than typical antipsychotics, but these differences disappeared when dosages were increased. Only one trial reported cerebrovascular adverse events, with a number needed to harm of 14 (95% CI 8, 41).
CONCLUSIONS: At lower doses atypical antipsychotics may cause fewer adverse events in the treatment of BPSD, but there is uncertainty about their cerebrovascular safety profile. The RCTs included in this systematic review described adverse events too incompletely and heterogeneously to allow generation of clear treatment recommendations, and they do not provide sufficient evidence to support recent safety warnings. Better reporting on harms in RCTs is needed to enable rational treatment decisions with respect to use of antipsychotics for BPSD.