The role of natural experiments in hepatology research: filling the gap between clinical trials and service evaluations.

Research developing and testing interventions that address the social determinants of liver disease are urgently needed; however, this cannot be achieved using conventional clinical research designs. A different approach is needed to conduct widely applicable, inclusive, and community-based research that addresses upstream factors driving liver morbidity. Natural experimental studies encompass a well-established field of research methodology that is less familiar to clinical hepatologists than conventional research methods such as the randomized control trial. The key strength of natural experiments is that, when robustly designed, they can be used to imply causality from routinely collected data. As such, they are well placed to test the impact of community interventions that aim to address social determinants of liver disease that cannot feasibly be assessed in a randomized control trial. In this review, we define natural experiments and their potential utility. We then work through examples of where they have already been used in clinical hepatology to highlight a range of research designs, analytical approaches, and best practices regarding their conduct and reporting. In doing so, we hope to equip clinical hepatologists with another tool to ensure the hepatology community can meet the global liver disease epidemic with evidence-based interventions.