Cost-effectiveness of constraint-induced movement therapy implementation in neurorehabilitation : the ACTIveARM project

Background: Constraint-induced movement therapy (CIMT) is a recommended intervention for improving arm recovery following stroke and traumatic brain injury; however, delivery in practice remains rare. Purpose: The aim of this study was to investigate the costs and cost effectiveness of CIMT delivery, and the use of a CIMT implementation package designed to improve CIMT uptake and delivery by therapists in Sydney, Australia. Methods: This economic evaluation was conducted with a subset of CIMT programmes (n = 20) delivered by neurological rehabilitation teams at five varied hospitals within a mixed methods implementation study (ACTIveARM). The costs of delivering the CIMT implementation package and publicly funded CIMT were calculated using a bottom-up approach. A cost-effectiveness analysis was conducted, using decision analytic modelling. We compared the uptake and outcomes of people who received CIMT from health services that had received a CIMT implementation package, with those receiving standard upper limb therapy. An Australian health care system perspective was used in the model, over a 3-week time horizon (the average timeframe of a CIMT programme). All costs were calculated in Australian dollars (AUD). Inputs were derived from the ACTIveARM study and relevant literature. The Action Research Arm Test was used to measure arm outcomes. Sensitivity analyses assessed the impact of improving CIMT uptake, scale-up of the implementation package and resource adjustment, including a ‘best-case’ scenario analysis. Results: The total cost of delivering the implementation package to nine teams across five hospitals was $110,336.43 AUD over 18 months. The mean cost of delivering an individual CIMT programme was $1233.38 AUD per participant, and $936.03 AUD per participant for group-based programmes. The incremental cost effectiveness ratio (ICER) of individual CIMT programmes was $8052 AUD per additional person achieving meaningful improvement in arm function, and $6045 AUD for group based CIMT. The ICER was most sensitive to reductions in staffing costs. In the ‘best-case’ scenario, the ICER for both individual and group-based CIMT was $245 AUD per additional person gaining a meaningful change in function. Conclusion: Therapists improved CIMT uptake and delivery with the support of an implementation package, however cost effectiveness was unclear. Clinical Trial Registration: https://anzctr.org.au/Trial ID: ACTRN12617001147370.