Contextual heterogeneity: unlocking local relevance from global evidence

Article type
Authors
Gruen R, Green S, Lavis J
Abstract
Abstract: Most people have a particular environment or perspective in mind when using research literature. This may be quite different to that in which research studies were conducted. Such differences in context affect the interpretation and application of research evidence, as well as being issues shaping primary study design. Relevant contextual factors include available physical, human, political, financial and social resources, the organization of health services, social factors such as political interests and cultural norms, and the physical environment. Contextual factors can affect several aspects of interpretation, such as the relative importance of a health problem, the practicality or appropriateness of an intervention, the relevance of outcome measures, and the balance of benefits and harms. Relevant contextual characteristics depend on the condition, intervention or diagnostic test in question. Some, such as per capita health expenditure relative to the cost of an intervention, are germane. Others may be quite specific. The traditional characteristics of studies reported in most Cochrane reviews do not facilitate the identification of important factors, nor their description. We present a structured approach to accommodating both generic and specific aspects of contextual heterogeneity in systematic reviews. These were developed for systematic reviews of health system interventions and have evolved in the work of the Global Evidence Mapping Initiative, a collaborative with the goal of improving methods for applicability and relevance of research evidence. We will demonstrate how such an approach can enhance the applicability and usefulness of a systematic review.