USING HEALTH INSURANCE DATABASES FOR EPIDEMIOLOGICAL RESEARCH: A SCOPING REVIEW

OBJECTIVES: This scoping review aimed to appraise the existing literature on using the claims databases for epidemiological studies and to draw inferences for using data from Pakistan's health insurance databases. METHODS: We conducted a scoping review of literature focusing on health insurance databases, querying MEDLINE, EMBASE and Google Scholar. We used the frameworks proposed by the Joanna Briggs Institute and Arksy and O'Malley for mapping our results. RESULTS: There was a considerable chronological increase in studies published using data from health insurance databases. Most of the studies in our search were from economically developed countries. Most of the studies (n=84) focussed on chronic non-communicable diseases, while a limited number (n=09) focussed on communicable (infectious) diseases. Our findings suggest that insurance databases could be utilised to study rare diseases, prospects of prolonged follow-up, and minimal research costs. This is especially important for countries like Pakistan, having limited resources to conduct regular, population-level epidemiological studies. Several methodological approaches (for instance, disease, pharmacy or intervention classification codes) were presented in these studies to extract epidemiological data from the insurance database. CONCLUSION: Health insurance databases are utilised as sources for epidemiological studies, predominantly for chronic illnesses, in economically developed countries. Methodological approaches described in these papers could be used to extract data for epidemiological research from health insurance databases in Pakistan. This could be especially useful for following the patterns of infectious disease in the country.