Structured analyses of interventions to prevent delirium.

BACKGROUND Delirium is one of the most serious complications in hospitalized elderly, with incidences ranging from 3-56%. The objective of this meta-analysis was two-fold, first to investigate if interventions to prevent delirium are effective and second to explore which factors increase the effectiveness of these interventions. METHODS An electronic search was carried out on articles published between January 1979 and July 2009. Abstracts were reviewed, data were extracted and methodologic quality was assessed by two independent reviewers. Effect sizes of the interventions were expressed as ORs (odds ratios) and 95%CIs (confidence intervals). A random effect model was used to provide pooled ORs. To explore which factors increase the effectiveness of the interventions, ORs were stratified for several factors. RESULTS Sixteen relevant studies were found. Overall the included studies showed a positive result of any intervention to prevent delirium (pooled OR: 0.64; 95%CI: 0.46-0.88). The largest effect was seen in studies on populations with an incidence of delirium above 30% in the control group (pooled OR: 0.34; 95%CI: 0.16-0.71 versus 0.76; 95%CI: 0.60-0.97). CONCLUSIONS Interventions to prevent delirium are effective. Interventions seem to be more effective when the incidence of delirium in the population under study is above 30%. To maximize the options for a cost-effective strategy of delirium prevention it might be useful to offer an intervention to a selected population.