Introduction Previous reviews have investigated the relationship between empathy and burnout. However, these are now out of date, did not capture the effect of the pandemic, did not include healthcare professionals other than doctors and nurses or medical students, did not assess the impact of differences in profession and did not pool the data, which made estimating the strength of the association unclear. We therefore aim to address these shortcomings in an up-to-date, rigorous, systematic review and meta-analysis. Methods and analysis Findings will be reported in accordance with the Preferred Reporting Items for Systematic reviews and Meta-Analyses guidelines and flowchart. We will search American Psychological Association (APA) PsycINFO, APA PsycArticles, Cumulative Index to Nursing and Allied Health Literature (CINAHL), The Cochrane Library, PubMed and Scopus. We will also search ResearchSquare and Social Science Research Network (SSRN) for preprints; ProQuest Dissertations and Theses and Electronic Theses Online Service for relevant theses. Forward and backward citation searches will identify additional studies. Two independent reviewers will screen titles, abstracts and full texts and extract data. Two independent reviewers will assess risk of bias using Risk of Bias 2 (RoB 2) for randomised controlled trials, Risk of Bias in Non-randomised Studies of Interventions (ROBINS-I) for non-randomised interventional studies and Risk of Bias in Non-randomised Studies of Exposures (ROBINS-E) for observational studies. For all included studies, we will summarise the study characteristics, including number of participants; health profession, specialty and career stage; country and gender. If data are suitable, we will pool results and conduct subgroup analyses (including by health profession, career stage and clinical specialty). We will also explore the relationships between subscales of empathy and burnout. We will use metaregression to explore the impact of theoretically derived factors (such as study design and profession) on the strength of the association. Sensitivity analyses will assess the impact of low-quality research. In our discussion, we will summarise results, the limitations and provide a general interpretation of the results and implications. Ethics and dissemination Ethical approval is not required for this review, as primary data will not be collected. The review will be disseminated through peer-reviewed publication and presentation at conferences. PROSPERO registration number CRD420251075618.
Edwards et al. (2026) studied this question.