Practice-Based Evidence and Public Health

Introduction

Practice-based evidence (PBE) is a term often used to describe information collected from evaluations of interventions in real world conditions. Including PBE in systematic reviews may increase their external validity and provide useful information about implementation. Yet challenges arise from the lack of a standard definition for PBE and the complexity of the concept. Our goal was to assess the relative contributions of PBE and explicitly research-based evidence (RBE) in systematic reviews conducted for the Guide to Community Preventive Services (Community Guide). The Community Guide is an evidence-based resource for recommendations about effective community-based public health programs, services, and policies. These recommendations are made by the independent, non-federal Community Preventive Services Task Force, based on systematic reviews of the available evidence.

Methods

To differentiate practice-based and research-based evidence, we developed the following operational definitions: a) PBE: “evaluations of interventions that have been implemented in practice with the primary intent of improving health;” and b) RBE: “studies conducted with the intent of testing hypotheses on causal relationships between an intervention and outcomes of interest.” We then applied these definitions to determine the prevalence of PBE within a purposive sample of Community Guide reviews - all reviews in five topics chosen to ensure a full range of intervention types and settings. The key feature used to distinguish the operational definitions was whether allocation of subjects to intervention and comparison conditions was dictated by researchers (RBE; e.g. through randomization) or under the control of others (PBE; e.g., practitioners, legislators, administrators). 

Results

Within the selected vaccine, tobacco control, excessive alcohol consumption, cancer, and diabetes prevention topics, there were 99 reviews containing 1774 studies conducted at worksites and in the healthcare system and community. Fifty two percent of studies were identified as PBE and 48% were RBE. Community-based and policy interventions had a higher proportion of PBE, while RBE was more frequent for healthcare system interventions. PBE and RBE studies did not differ substantially in quality based on the Community Guide’s quality scoring methods.

Conclusion

Studies considered to be PBE are prevalent in the public health literature. Many are of sufficient rigor to be incorporated into systematic reviews that support evidence-based recommendations for public health practice. This finding suggests it is possible to attain the balance required in public health between studies emphasizing internal validity (RBE) and those that permit greater generalizability and documentation of feasibility (PBE).