Functional outcomes of drug treatment in Alzheimer's disease: a systematic review and meta-analysis.

BACKGROUND: Patient functioning is an important outcome in Alzheimer's disease, but treatment-related improvements in patient function are difficult to quantify because a number of different scales are used in its measurement. OBJECTIVE: To evaluate systematically the evidence relating to patient functioning as an outcome measure in the drug treatment of Alzheimer's disease. Data were obtained by searching MEDLINE((R)), EMBASE, The Cochrane Library and the International Pharmaceutical Abstracts from 1980 through to December 2005 for studies assessing functional outcomes with donepezil, galantamine, rivastigmine and memantine in Alzheimer's disease. Reference lists were searched manually and pharmaceutical manufacturers were invited to submit dossiers. Trained reviewers abstracted data and assessed the internal validity (quality) of trials using predefined criteria. Standardised effect sizes (i.e. Cohen's standardised mean difference [d]) for various functional outcome scales and pooled mean incidence and 95% CIs for adverse events were calculated and summarised qualitatively and quantitatively. Meta regression was used to explore potential heterogeneity. RESULTS: Overall, the standardised effect size for functional outcome measures was small (d = 0.1-0.4) among included studies. However, effect sizes consistently favoured drug treatment over placebo. For all drugs, pooled standardised effect sizes were consistent in both short (<24 weeks; d = 0.25; 95% CI 0.13, 0.37) and long trials (>/=24 weeks; d = 0.29; 95% CI 0.22, 0.36). The pooled effect size was not significantly affected by parameters such as disease severity, age, gender and drug dose. Adverse events were generally limited to gastrointestinal problems, weight loss and dizziness, all of which were reported in <20% of patients on average. CONCLUSIONS: Standardised estimates of effect size across diverse functional outcome measures for drug treatment in patients with Alzheimer's disease were small and the data reflect only a modest trend favouring active treatment over placebo. However, given the current lack of other effective treatments for Alzheimer's disease, this trend supports the clinical benefits of these treatments with regard to this important health outcome. [ABSTRACT FROM AUTHOR]