Cost-effectiveness of a community-based integrated care model compared with usual care for older adults with complex needs: a stepped-wedge cluster-randomised trial

OBJECTIVE: To assess the cost of implementation, delivery and cost-effectiveness (CE) of a flagship community-based integrated care model (OPEN ARCH) against the usual primary care. DESIGN: A 9-month stepped-wedge cluster-randomised trial. SETTING AND PARTICIPANTS: Community-dwelling older adults with chronic conditions and complex care needs were recruited from primary care (14 general practices) in Far North Queensland, Australia. METHODS: Costs and outcomes were measured at 3-month windows from the healthcare system and patient’s out-of-pocket perspectives for the analysis. Outcomes included functional status (Functional Independence Measure (FIM)) and health-related quality of life (EQ-5D-3L and AQoL-8D). Bayesian CE analysis with 10 000 Monte Carlo simulations was performed using the BCEA package in R (V.3.6.1). RESULTS: The OPEN ARCH model of care had an average cost of $A1354 per participant. The average age of participants was 81, and 55% of the cohort were men. Within-trial multilevel regression models adjusted for time, general practitioner cluster and baseline confounders showed no significant differences in costs, resource use or effect measures regardless of the analytical perspective. Probabilistic sensitivity analysis with 10 000 simulations showed that OPEN ARCH could be recommended over usual care for improving functional independence at a willing to pay above $A600 (US$440) per improvement of one point on the FIM Scale and for avoiding or reducing inpatient stay for any willingness-to-pay threshold up to $A50 000 (US$36 500). CONCLUSIONS AND IMPLICATIONS: OPEN ARCH was associated with a favourable Bayesian CE profile in improving functional status and dependency levels, avoiding or reducing inpatient stay compared with usual primary care in the Australian context. TRIAL REGISTRATION NUMBER: ACTRN12617000198325.