Supporting rehabilitation stakeholders in making service delivery decisions: a rapid review of multi-criteria decision analysis methods

Abstract

Purpose This review aimed to synthesize knowledge about multi-criteria decision analysis methods for supporting rehabilitation service design and delivery decisions, including: (1) describing the use of these methods within rehabilitation, (2) identifying decision types that can be supported by these methods, (3) describing client and family involvement, and (4) identifying implementation considerations. Methods We conducted a rapid review in collaboration with a knowledge partner, searching four databases for peer-reviewed articles reporting primary research. We extracted relevant data from included studies and synthesized it descriptively and with conventional content analysis. Results We identified 717 records, of which 54 met inclusion criteria. Multi-criteria decision analysis methods were primarily used to understand the strength of clients’ and clinicians’ preferences (n = 44), and five focused on supporting decision making. Shared decision making with stakeholders was evident in only two studies. Clients and families were mostly engaged in data collection and sometimes in selecting the relevant criteria. Good practices for supporting external validity were inconsistently reported. Implementation considerations included managing cognitive complexity and offering authentic choices. Conclusions Multi-criteria decision analysis methods are promising for better understanding client and family preferences and priorities across rehabilitation professions, contexts, and caseloads. Further work is required to use these methods in shared decision making, for which increased use of qualitative methods and stakeholder engagement is recommended. IMPLICATIONS FOR REHABILITATION Multi-criteria decision analysis methods are promising for evidence-based, shared decision making for rehabilitation. However, most studies to date have focused on estimating stakeholder preferences, not supporting shared decision making. Cognitive complexity and modelling authentic and realistic decision choices are major barriers to implementation. Stakeholder-engagement and qualitative methods are recommended to address these barriers.