Myanmar's war has pushed millions further from hospital care, mapping study finds

Armed conflict in Myanmar has not merely damaged the health system; it has concentrated its harm on the people who could least reach care to begin with. A population-weighted analysis of all 330 townships finds that fighting after the February 2021 coup fell disproportionately on communities that were already farthest from a hospital, and that the number of people living under both burdens at once — poor access and intense conflict — grew from 0.4 million in early 2021 to 3.1 million by early 2024 [s1].

What the analysis measured

The study is a longitudinal ecological analysis covering 330 townships across 16 quarters from 2021 to 2024 [s1]. For every populated square kilometre, the authors estimated travel time to the nearest of 68 general hospitals — the tier that provides surgical care — using the pre-coup facility registry, OpenStreetMap road data and WorldPop population estimates [s1]. Conflict intensity came from 35,647 battle and explosion or remote-violence events recorded by the Armed Conflict Location and Event Data Project, and 655 facility attacks logged by the Safeguarding Health in Conflict Coalition were analysed separately [s1].

The baseline was already unequal. Before the coup, 5.8 million people — 12% of the population — lived more than two hours from a general hospital [s1]. Because the analysis holds access fixed at that pre-coup level rather than modelling roads and facilities destroyed since, the authors describe their estimates as optimistic bounds: the real picture is likely worse [s1].

Conflict tracked the gaps

To test whether the war concentrated on underserved populations, the study used a concentration index, where zero means an even spread and a positive value means the burden falls on those with worse access. Conflict events returned an index of 0.126 (95% CI 0.053 to 0.199) and conflict fatalities 0.169 (95% CI 0.078 to 0.263) — both significantly skewed toward the worst-access populations [s1]. Attacks on health facilities, by contrast, showed no significant concentration (0.049, 95% CI −0.041 to 0.145), which the authors attribute to a blunt geography: the most remote townships have few fixed facilities left to target [s1].

Adding the two exposures together, the study accumulated 8.4 million person-years in which people faced poor or very poor access and high or extreme conflict simultaneously over 2021–2024 [s1]. That double-burden exposure rose roughly eightfold quarter on quarter, from about 100,000 person-years in early 2021 to 780,000 by early 2024 [s1].

What that looks like on the ground

A separate qualitative study from Sagaing Region, one of the conflict's epicentres, records what the travel-time maps cannot. Interviewing 26 healthcare workers and six community members, it found that junta-controlled rural health centres had stopped functioning in areas held by resistance forces, and that anti-junta health workers and communities had improvised alternative primary-care systems in their place [s2]. Those systems could deliver basic primary care but little secondary care, and were constrained by shortages of staff, supplies, funding and infrastructure, by restricted travel and communication, and by outright safety concerns [s2]. The prevailing needs its participants described — conflict injuries, infectious diseases, chronic non-communicable diseases and mental health — are precisely the conditions that a two-hour journey to a shuttered hospital leaves unmet [s2].

What the study cannot say

This is an ecological analysis, not a study of individual patients, and it cannot report who was denied care or what happened to them [s1]. Holding access at the pre-coup baseline means the travel times are conservative and the trends should be read as lower bounds, not exact counts [s1]. The Sagaing interviews, for their part, capture the experiences of anti-junta providers and the communities they serve; they are not a neutral audit of every actor in the conflict, and the authors present them as situated testimony [s2]. Neither study measures deaths averted or incurred.

Why it matters

The finding that matters is not that war is bad for hospitals — it is that war in Myanmar has widened a gap that already ran along geography, so that surgical care is receding fastest from the people who had the least of it. That is the same dynamic driving health-system collapse in eastern DRC, and the concentration-index framework the authors built is explicitly designed to be reused in other conflict settings [s1]. The measurable question from here is whether humanitarian access can be negotiated into exactly the townships the index flags — the ones where poor access and intense fighting now coincide — rather than the ones easiest to reach.

Sources

• Geographical accessibility to hospital care amid armed conflict in Myanmar: a population-weighted analysis of compounding health inequality, 2021–2024, BMJ Global Health, 1 September 2026
• Newly emerged resistance healthcare systems in northwest Myanmar's Sagaing Region post-military coup, BMJ Global Health, 8 January 2026