Decision aids have been overall successful in improving the quality of health decision making. However, it is unclear whether the impact of the results of using decision aids also apply to older people (aged 65+). We sought to systematically review randomized controlled trials (RCTs) and clinical controlled trials (CCTs) evaluating the efficacy of decision aids as compared to usual care or alternative intervention(s) for older adults facing treatment, screening or care decisions.

A systematic search of (1) a Cochrane review of decision aids and (2) MEDLINE, Embase, PsycINFO, Cochrane library central registry of studies and Cinahl. We included published RCTs/CCTs of interventions designed to improve shared decision making (SDM) by older adults (aged 65+) and RCTs/CCTs that analysed the effect of the intervention in a subgroup with a mean age of 65+. Based on the International Patient Decision aid Standards (IPDAS), the primary outcomes were attributes of the decision and the decision process. Other behavioral, health, and health system effects were considered as secondary outcomes. If data could be pooled, a meta-analysis was conducted. Data for which meta-analysis was not possible were synthesized qualitatively.

The search strategy yielded 11,034 references. After abstract and full text screening, 22 papers were included. Decision aids performed better than control resp. usual care interventions by increasing knowledge and accurate risk perception in older people (decision attributes). With regard to decision process attributes, decision aids resulted in lower decisional conflict and more patient participation.

This review shows promising results on the effectiveness of decision aids for older adults. Decision aids improve older adults’ knowledge, increase their risk perception, decrease decisional conflict and seem to enhance participation in SDM. It must however be noted that the body of literature on the effectiveness of decision aids for older adults is still in its infancy. Only one decision aid was specifically developed for older adults, and the mean age in most studies was between 65 and 70, indicating that the oldest-old were not included. Future research should expand on the design, application and evaluation of decision aids for older, more vulnerable adults.