Characteristics of Randomized Controlled Trials Designed for Elderly: A Systematic Review

Objectives To quantify the proportion of randomized controlled trials (RCTs) specifically designed for elderly, and to assess their characteristics, as compared to RCTs not specifically designed for elderly. Design Review and synthesis of published literature. Measurements We searched PubMed for articles published in the year 2012. We included RCTs. Articles were excluded if not conducted with human subjects or if findings of secondary analyses were reported. A random sample of 10% was drawn and of this selection the following trial characteristics were extracted: sample size, disease category, age of sample, and age-related inclusion criteria. Clinical trials were defined to be specifically designed for elderly if a lower age cut-off of ≥ 55 years was used, or when participants had an average age of ≥ 70 years. Results The search strategy yielded 26,740 articles, from which a random sample was drawn, resulting in 2375 articles. After exclusion, data was extracted from 1369 publications. Of these 1369 RCTs, 96 (7%) were specifically designed for elderly. In comparison with trials not designed for older adults, trials designed for elderly contained a significantly larger median number of participants (125 vs. 80, p = 0.008) significantly more trials designed for elderly fell into the disease categories eye (6% vs. 2%, p = 0.005), musculoskeletal (13% vs. 7%, p = 0.023) and circulatory system (16% vs. 9%, p = 0.039). No significant difference was observed with regard to the other disease categories. Conclusion There is a low proportion of RCTs specifically designed for elderly. As older patients will increasingly form the majority in medical practice, there is an urgent need for stronger evidence for the formulation of treatment guidelines specifically for older adults.

Overview publication

TitleCharacteristics of Randomized Controlled Trials Designed for Elderly: A Systematic Review
Date15 May 2015
Issue namePLOS ONE
Issue number10(5): e0126709.
AuthorsBroekhuizen K, Pothof A, de Craen AJM & Mooijaart SP
Read Read publication