IPD meta-analyses are important to improve evidence-based decisions in the geriatric population

Article type
Authors
van de Glind E1, Rhodius-Meester H2, Reitsma JB3, Hooft L4, van Munster B5
1Department of Internal Medicine, Section of Geriatrics/Dutch Cochrane Centre, Academic Medical Center, Amsterdam, The Netherlands
2Department of Geriatric Medicine, Slotervaart Hospital, Amsterdam, The Netherlands
3Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, The Netherlands
4Dutch Cochrane Centre, Academic Medical Center, Amsterdam, The Netherlands
5Gelre Hospitals, Department of Geriatric Medicine, Apeldoorn, The Netherlands
Abstract
Background: In a meta-analysis of individual patient data (IPD), original trial data may be used to analyse the effects of an intervention in a particular subgroup. IPD meta-analyses may, therefore, be a helpful method to generate evidence that is more applicable to the geriatric population, which may have treatment effects that differ from a younger population.

Objectives: To provide an overview of treatment differences between the older and younger patients based on meta-analyses on an IPD-level.

Methods: A MEDLINE search was conducted for IPD meta-analyses of randomized controlled trials (RCTs) published before July 2012. IPDmeta-analyses involving patients with a mean age of≥70 years or describing a subgroup in this age range were included. We evaluated whether the IPDmeta-analyses reported similar conclusions for both the younger and older populations.

Results: Twenty-six IPD meta-analyses with a subgroup of older individuals were included (median N = 3.581). The most important reason for applying the IPD methodology was the ability to perform a subgroup analysis in the older population, as well as multivariable analysis with additional patient characteristics (e.g., different tumor parameters or type of drug). Fourteen IPD meta-analyses suggested that older people should receive distinct treatments compared to younger people due to differences in effectiveness, whereof eight reviews indicated that the investigated treatment(s) should be avoided or adjusted in older patients. However, in six reviews the investigated treatment was more effective in older than younger patients.

Conclusions: IPD meta-analysis is a valuable approach for generating evidence for older patients. In 54% of the included IPD meta-analyses, treatment effects differed between older and younger patients. The collaborative sharing of raw data should be promoted and facilitated to improve evidence-based decisions in the growing population of older and more vulnerable patients in our society.