Physical exercises to prevent and rehabilitate hospital-associated disability in hospitalised older people: A protocol for a living systematic review with network meta-analysis.

Background: During hospitalisation, older adults are at risk of developing functional declineunrelated to the condition for which they were admitted. The loss of independence in at least one activity of daily living is referred to as hospital-associated disability (HAD). A loss of functional independence in hospitalised older people is associated with a greater risk of nursing home placement and increased caregiver burden after hospital dismissal. It is essential to raise awareness of the HAD problem among older patients and to implement adequate preventive and treatment measures. Objectives: To evaluate systematically the effectiveness of physical exercise to prevent and rehabilitate HAD in people aged 65 years and over who are hospitalised in an acute care setting. To assess factors (training intensity, volume, and frequency) that potentially influence the effectiveness of physical exercise on HAD in such patients through meta-regression. Methods: We will conduct a living systematic review of randomised controlled trials including a network meta-analysis. For the outcome HAD, various physical exercise modalities such as resistance training, aerobic exercises or cycling could be proposed. Unlike pair-wise meta-analyses, which allow only head-to-head comparisons, network meta-analysis enables us to compare all modalities simultaneously. Guidelines from the Cochrane Handbook for a systematic review of interventions will be followed. Discussion: A network meta-analysis offers several advantages which are relevant in the context of the present review question: i) it allows the integration of multiple comparisons within one analytical framework; ii) it enables to the evaluation of the comparative effectiveness of each exercise modality, thus allowing for a hierarchy of interventions; iii) it increased precision of the effect estimates compared to traditional meta-analyses. The latter advantage is crucial for outcomes like functional decline after a hospital stay.