Objective: To summarise frailty measures and outcomes reported in trials involving frail older inpatients.

Methods: Databases were searched for randomised controlled trials enrolling frail older inpatients.

Results: Twenty-four articles describing twelve trials were included. Seven trials applied six tools to measure frailty, whilst five trials employed ad hoc measures. Eighty outcomes were examined with survival and functional status reported most commonly. Nine studies trialled multidisciplinary, geriatrician-led interventions. Statistically significant between-group differences were detected for at least one outcome in ten trials. All studies represented high risk of bias within at least one domain.

Conclusions: Heterogeneity of interventions, measurement of frailty and outcomes reported limit generalisability of findings. Many articles purport to study frail patients, yet do not enrol patients using any frailty measurement tool. Utilising validated instruments to measure frailty and a standard set of health outcomes relevant to older people would assist consistent reporting and evaluation of future studies.