The cost of non-drug interventions that improve function and reduce dementia-related behaviors

Background: To determine the net cost of non-drug interventions that maintain or improve a person with dementia’s physical function and/or reduce challenging behaviors. Cost data are needed to inform the adoption of non-drug interventions in health systems and the development of policies to incentivize their use. Methods: We modified a person-level microsimulation to model the cost of four non-drug interventions relative to usual care: Collaborative Care, Care of Persons with Dementia in their Environments (COPE), Tailored Activity Program (TAP), and Skills2Care. We also conducted a value of information analysis to quantify the optimal sample size of conducting a new randomized trial that would reduce uncertainty on the cost savings of each intervention from a societal perspective. Finally, we conducted sensitivity analyses. Results: Collaborative Care, TAP and COPE were cost savings compared to usual care (-$572, -$1,816, and -$5,262, respectively). Skills2Care results in a $89 net increase in cost compared to usual care. The value of information analysis identified the optimal sample size of a potential future study: Skills2Care (optimal n = 8,560), TAP (optimal n = 5,650), COPE (optimal n = 3,910) and Collaborative Care (optimal n = 3,630). In one-way sensitivity analyses, when we applied a pessimistic assumption for the treatment effect, COPE and TAP were still cost saving, while Collaborative Care cost more than usual care. Conclusions did not materially change in sensitivity analyses that varied treatment cost. Conclusions: Non-drug dementia care interventions that maintain or improve a person with dementia’s function and/or reduce challenging behaviors present a viable clinical / economic model of care for health systems.