2021/01/18 by Ellicott C. Matthay, Matthay, Ellicott C., Erin Hagan +11 · 2 citations
Health Professions · #Health Policy Implementation Science #Public Health Policies and Education #Primary Care and Health Outcomes
paper · pdf · doi:10.48550/arxiv.2101.07392
Evidence for Action (E4A), a signature program of the Robert Wood Johnson\nFoundation, funds investigator-initiated research on the impacts of social\nprograms and policies on population health and health inequities. Across\nthousands of letters of intent and full proposals E4A has received since 2015,\none of the most common methodological challenges faced by applicants is\nselecting realistic effect sizes to inform power and sample size calculations.\nE4A prioritizes health studies that are both (1) adequately powered to detect\neffect sizes that may reasonably be expected for the given intervention and (2)\nlikely to achieve intervention effects sizes that, if demonstrated, correspond\nto actionable evidence for population health stakeholders. However, little\nguidance exists to inform the selection of effect sizes for population health\nresearch proposals. We draw on examples of five rigorously evaluated population\nhealth interventions. These examples illustrate considerations for selecting\nrealistic and actionable effect sizes as inputs to power and sample size\ncalculations for research proposals to study population health interventions.\nWe show that plausible effects sizes for population health inteventions may be\nsmaller than commonly cited guidelines suggest. Effect sizes achieved with\npopulation health interventions depend on the characteristics of the\nintervention, the target population, and the outcomes studied. Population\nhealth impact depends on the proportion of the population receiving the\nintervention. When adequately powered, even studies of interventions with small\neffect sizes can offer valuable evidence to inform population health if such\ninterventions can be implemented broadly. Demonstrating the effectiveness of\nsuch interventions, however, requires large sample sizes.\n