Reviews of Individual Patient Data (IPD) Are Useful for Geriatrics: An Overview of Available IPD Reviews

Objectives To determine how many individual patient data (IPD) reviews that included older people were available in MEDLINE and whether the effectiveness of treatments differed between older and younger individuals. Design Overview of IPD reviews. Setting A MEDLINE search was conducted for IPD reviews of randomized controlled trials published before July 2012. Participants IPD reviews that presented a regression model that included age as a factor or a subgroup analysis of individuals aged 70 and older or in which all participants were aged 70 and older. Measurements Whether the IPD reviews reported similar conclusions for the younger and older populations was evaluated. Results Twenty-six IPD reviews with a subgroup of older individuals and eight reviews with only older individuals were included (median N=3,351). The most important reason for choosing an IPD review was the ability to perform a subgroup analysis in the older population. Fourteen IPD reviews suggested that older people should receive different treatments from younger people because of differences in effectiveness, six of which indicated that the investigated treatment(s) should be avoided in older adults. Conclusion IPD review is a valuable approach for generating evidence in older adults. Treatment effects frequently differed between older and younger individuals. Still, IPD results should be applied to older adults cautiously, because they are often excluded from primary trials. The collaborative sharing of raw data should be promoted to improve evidence-based decisions for this group.