A refined typology of approaches for sustaining health interventions: a mixed methods study and Delphi technique with health and implementation science experts

Abstract Background A typology of approaches used by external agencies to evaluate the sustainment of evidence-based intervention/initiatives (EBIs) was recently proposed. As the typology was initially developed by a single research group, further refinement was warranted to ensure its validity. Drawing on expert insights and real-world examples of program sustainment, this study aimed to: (1) refine the typology and (2) identify key research efforts to improve its utility and application. Methods We used a multi-method design with a purposeful sample of international experts in public health and implementation science. Qualitative data from a one-day workshop were synthesized using rapid analysis. A proof-of-concept case study analysis was conducted with 22 health EBIs described in peer review papers, independently classified by typology approach. Final typology revisions combined insights from both analyses. A three-round modified Delphi technique was used to identify research priorities aligned with the typology to advance sustainment research. Results Seventeen experts across three countries participated and agreed that the approaches in the initial typology ‘Self-Sustainment’, ‘Static Sustainment’, and ‘Dynamic Sustainment’ reflected practice. Three refinements were recommended: (1) Characterise external sustainment support as a continuum (static to dynamic) to capture the complexity of sustaining EBIs in real-world settings; (2) reflect the temporal nature of support by grounding classifications to a ‘ defined period of time ’; and (3) clarify typology language to improve usability. The revised typology now includes: ‘ Independent Sustainment’ where no active external support is provided to sustain the EBI; ‘Static Sustainment’ where fixed (static) active external support is provided; and ‘Dynamic Sustainment’ where active external support adapts over time in response to changes. The highest priority research needs to advance typology were: trials that test the effectiveness of sustainment approaches (mean = 8.6 SD = 0.7), mapping and case study analysis to explore typology transferability (mean = 8.2, SD = 1.1), and practitioner feedback to assess typology utility (mean = 7.2 SD = 1.1). Conclusion This study provides a refined and more transferable typology of approaches external agencies use to support EBIs to sustain. Selecting an approach early can guide agencies to plan for the lifecycle of an EBI and may improve the likelihood it continues in practice.