1598 a typology of community-based complex interventions to sustain independence in older people

Abstract Introduction Provision of community-based health services to support independence of older people, and further research in this area, would be improved by a typology of these complex interventions - thereby enabling evidence synthesis and the identification of effective intervention components. We aimed to produce such a typology in preparation for a systematic review and network meta-analysis. Methods The typology was developed based upon the descriptions of these interventions in published reports. This involved four stages: (1) systematic identification of relevant RCTs and related publications; (2) the extraction of descriptions of the interventions (including control/comparison) using the Template for Intervention Description and Replication (TIDieR); (3) a qualitative synthesis generating categories of key intervention features and (4) grouping the interventions based on the categories. Results Our search identified 496 reports of 129 studies, involving 266 intervention arms. 19 intervention components were identified: Formal homecare; Physical exercise; Health education; ADL training; Providing aids and adaptations; Nutritional support; Psychological therapy; Technology for communication and engagement; Cognitive training; Engagement in meaningful activities; Care voucher provision; Alternative medicine; Social skills training; Welfare rights advice; Medication review; Monitoring; Routine risk screening; Multifactorial-action from care planning; and Routine review following multifactorial-action from care planning. Multifactorial-action from care planning refers to a process of individualised, multidomain assessment and management resulting in a tailored selection of action components, as in comprehensive geriatric assessment. 63 different intervention types (combinations of these components) were identified. Conclusions The typology provides an empirical basis for service planning and evidence synthesis. Target populations are not explicitly integrated and should be considered separately. The components, being broad actions, are likely to endure; further components may be identified. However, the huge potential number of intervention types constitutes a challenge to typical approaches to effectiveness research. We recommend better reporting about organisational aspects of interventions and usual care.