Burnout among international versus domestic medical graduates in the United States and United Kingdom: a systematic review and meta-analysis of trainees and early-career physicians

International medical graduates constitute a substantial fraction of the physician workforce in high-income destination countries. Whether their burnout rates differ from those of domestic medical graduates has not been quantitatively synthesized in any prior meta-analysis we identified through searches of PubMed/MEDLINE, Embase, PsycINFO, Scopus, and CINAHL conducted on 27 March 2026 and re-run with expanded population terms on 27 April 2026 (Supplement S1). To synthesize available comparative burnout estimates between international and domestic medical graduate trainees and early-career physicians. Studies were eligible if they reported burnout as a dichotomous outcome with extractable odds ratios. Effect sizes were pooled on the log-OR scale using random-effects meta-analysis with Paule-Mandel τ² estimation and Hartung-Knapp-Sidik-Jonkman (HKSJ) adjustment to confidence intervals, paired with a 95% prediction interval. DerSimonian-Laird is reported as a comparability analysis only. Risk of bias was assessed using a modified Newcastle-Ottawa Scale; certainty of evidence was rated using GRADE. We identified five effect sizes from four studies comprising 92,235 physician-responses. Because two UK survey waves may include overlapping respondents, the primary meta-analysis used four independent effect sizes from four studies comprising 73,049 physician-responses (Al-Haddad 2021); the five-effect all-wave analysis is reported as sensitivity. Under the primary HKSJ-Paule-Mandel random-effects model, international medical graduate trainees and early-career physicians had lower odds of reporting burnout than their domestic counterparts (pooled OR 0.65, 95% CI 0.49–0.87; 95% prediction interval 0.37–1.15; I²=90.3%). The prediction interval crosses 1.0; certainty of evidence was rated VERY LOW (GRADE). Across this small evidence base, all effect sizes favoured international graduate trainees, but the magnitude is uncertain (HKSJ 95% CI 0.49–0.87; 95% prediction interval 0.37–1.15) and the certainty of evidence is VERY LOW. Given the small evidence base, conceptually non-equivalent burnout instruments across studies, mixed adjustment status across studies, English-language eligibility restriction, and confinement to predominantly trainee populations in two destination countries (US, UK), this finding should be interpreted as a hypothesis-generating signal rather than a settled conclusion.