Optimisation: defining and exploring a concept to enhance the impact of public health initiatives

Abstract

BACKGROUND: Repeated, data-driven optimisation processes have been applied in many fields to rapidly transform the performance of products, processes and interventions. While such processes may similarly be employed to enhance the impact of public health initiatives, optimisation has not been defined in the context of public health and there has been little exploration of its key concepts. METHODS: We used a modified, three-round Delphi study with an international group of researchers, public health policy-makers and practitioners to (1) generate a consensus-based definition of optimisation in the context of public health and (2i) describe key considerations for optimisation in that context. A pre-workshop literature review and elicitation of participant views regarding optimisation in public health (round 1) were followed by a daylong workshop and facilitated face-to-face group discussions to refine the definition and generate key considerations (round 2); finally, post-workshop discussions were undertaken to refine and finalise the findings (round 3). A thematic analysis was performed at each round. Study findings reflect an iterative consultation process with study participants. RESULTS: Thirty of 33 invited individuals (91%) participated in the study. Participants reached consensus on the following definition of optimisation in public health: "A deliberate, iterative and data-driven process to improve a health intervention and/or its implementation to meet stakeholder-defined public health impacts within resource constraints". A range of optimisation considerations were explored. Optimisation was considered most suitable when existing public health initiatives are not sufficiently effective, meaningful improvements from an optimisation process are anticipated, quality data to assess impacts are routinely available, and there are stable and ongoing resources to support it. Participants believed optimisation could be applied to improve the impacts of an intervention, an implementation strategy or both, on outcomes valued by stakeholders or end users. While optimisation processes were thought to be facilitated by an understanding of the mechanisms of an intervention or implementation strategy, no agreement was reached regarding the best approach to inform decisions about modifications to improve impact. CONCLUSIONS: The study findings provide a strong basis for future research to explore the potential impact of optimisation in the field of public health.