Models of “Real World” health research: A critical review

Abstract Background The increasing call for real world evidence has resulted in a diverse range of studies being conducted using the broad “real world” descriptor. We conducted a critical review of how researchers are incorporating the real world in their evaluation of health innovations. Our aim was to develop a framework to categorise existing research approaches and to establish a more precise lexicon for real world health research models. Methods Five electronic databases were searched in July 2021 (PubMed, MEDLINE, Web-of-Science databases, Cochrane library and SCOPUS) for primary research articles presenting validations, evaluations, evidence or data drawn from real world, pragmatic or realist studies for health related innovations. 1762 records were identified. Two authors independently screened papers by title and abstract (n = 291), followed by a full-text screening (n = 104). 33 included studies were critically reviewed to determine the nature of the “real world” elements used in each of the research studies. Results It was shown that whilst authors used a wide variety of terms to describe their studies, they could be grouped into three categories according to the aim of the study. Two of the categories, the pragmatic study and the proof of concept study were derived from current published definitions. We identify the need for a category that provides evidence on the population using an innovation, rather than as an evaluation of the innovation itself. Conclusions We propose a fourth category – real world validation – and explain why it should be reserved for evidence meeting very specific criteria.