Effectiveness of a community-based rehabilitation programme following hip fracture: results from the Fracture in the Elderly Multidisciplinary Rehabilitation phase III (FEMuR III) randomised controlled trial

OBJECTIVE: To determine whether an enhanced community rehabilitation intervention (the Fracture in the Elderly Multidisciplinary Rehabilitation (FEMuR) intervention) was more effective than usual National Health Service care, following surgical repair of hip fracture, in terms of the recovery of activities of daily living (ADLs).

DESIGN: Definitive, pragmatic, multisite, parallel-group, two-armed, superiority randomised controlled trial with 1:1 allocation ratio.

SETTING: Participant recruitment in 13 hospitals across England and Wales, with the FEMuR intervention delivered in the community.

PARTICIPANTS: Patients aged over 60 years, with mental capacity, recovering from surgical treatment for hip fracture and living in their own home prior to fracture.

INTERVENTIONS: Usual rehabilitation care (control) was compared with usual rehabilitation care plus the FEMuR intervention, which comprised a patient-held workbook and goal-setting diary to improve self-efficacy, and six additional therapy sessions delivered in-person in the community, or remotely during COVID-19 restrictions (intervention), to increase the practice of exercise and ADL.

PRIMARY AND SECONDARY OUTCOME MEASURES: Primary outcome was the Nottingham Extended Activities of Daily Living (NEADL) scale at 12 months. Secondary outcomes included: Hospital Anxiety and Depression Scale, Falls Self-Efficacy-International scale, hip pain intensity, fear of falling, grip strength and Short Physical Performance Battery. Outcomes were collected by research assistants in participants' homes, whenever possible, but had to be collected remotely during COVID-19 restrictions.

RESULTS: In total, 205 participants were randomised (n=104 experimental; n=101 control). Trial processes were adversely affected by the COVID-19 pandemic. There were 20 deaths, 34 withdrawals and three lost to follow-up. At 52 weeks, there was no significant difference in NEADL score between the FEMuR intervention and control groups. Joint modelling analysis testing for difference in longitudinal outcome adjusted for missing values also found no significant difference with a mean difference of 0.1 (95% CI -1.1, 1.3). There were no significant between-group differences in secondary outcomes. Sensitivity analyses, examining the impact of COVID-19 restrictions, produced similar results. A median of 4.5 extra rehabilitation sessions were delivered to the FEMuR intervention group, with a median of two sessions delivered in-person. Instrumental variable regression did not find any effect of the amount of rehabilitation on the main outcome. There were 53 unrelated serious adverse events (SAEs) including 11 deaths in the control group: 41 SAEs including nine deaths in the FEMuR intervention group.

CONCLUSIONS: The FEMuR intervention was not more effective than usual rehabilitation care. The trial was severely impacted by COVID-19. Possible reasons for lack of effect included limited intervention fidelity (fewer sessions than planned and remote delivery), lack of usual levels of support from health professionals and families, and change in recovery beliefs and behaviours during the pandemic.

TRIAL REGISTRATION NUMBER: ISRCTN28376407.