21,853 war casualties in 149 Lebanese hospitals. Four in five deaths never reached care.

Casualty counts in armed conflict usually reach the public as running totals from a ministry or a UN agency. What they rarely carry is the clinical shape of the injuries underneath — where people were hurt, by what, and whether they lived long enough to be treated.

A study published in Conflict and Health on 17 December supplies that shape for Lebanon. Its authors describe it as the first nationally representative analysis of war-related injuries and deaths in the country [s1].

What the records contain

The study is a multicentre retrospective observational analysis of data from 149 hospitals across Lebanon, covering October 2023 to March 2025 [s1]. Mortality data came from the National Hospital Mortality System; morbidity data came from hospitals subsidised by the Ministry of Public Health [s1]. Injury cases were identified using ICD-10 codes [s1].

A total of 21,853 individuals were included. Of these, 80.45% sustained injuries and 19.55% were recorded as fatalities [s1].

Males accounted for 81.09% of cases, and Lebanese nationals for 90.08% [s1]. The South and Nabatiyeh governorates reported the highest injury and fatality rates [s1]. The leading mechanisms of injury were blast injuries, at 29.79%, and gunshots, at 14.04% [s1].

One reported figure does not sit easily with the rest. The paper states that the most affected age group was 40–44 years, at 38.65% [s1]. A single five-year band accounting for nearly two fifths of all cases would be an extraordinary concentration in a population-wide casualty series, and the abstract does not describe how age categories were bounded. The figure is reported here as published; readers should treat the age distribution as the least interpretable part of the dataset until the full breakdown is examined.

The number that describes the health system

Among those recorded as fatalities, 81.46% were dead on arrival [s1].

Of those who arrived injured, 51.9% were discharged from the emergency department, 40.86% were admitted, and 7.24% required intensive care [s1].

Those two statistics belong together. A trauma system's leverage is concentrated in the first hour after injury; if four in five deaths occur before a patient reaches a hospital door, then almost all of the mortality in this dataset sits outside the reach of hospital care entirely. The authors attribute the pattern among blast victims to limited access to timely care and to the targeting of healthcare infrastructure [s1].

That framing is the paper's, and it is a causal claim that a retrospective hospital dataset cannot by itself establish. Pre-hospital death is also driven by injury severity — high-energy blast trauma kills quickly regardless of ambulance availability — and by geography, since the worst-affected governorates are the ones furthest from tertiary centres. The dataset cannot separate those.

The context on attacks against health care

A separate study, published in Population Health Metrics on 14 December, examined temporal trends in attacks on health care across five conflicts — Lebanon, Myanmar, the occupied Palestinian territory, Sudan and Ukraine — using two databases: WHO's Surveillance System for Attacks on Health Care and the Safeguarding Health in Conflict Coalition dataset maintained by Insecurity Insight [s2]. Data collection ended on 31 December 2024 [s2].

Across all five countries combined, the WHO system recorded 4,289 attacks and the coalition dataset 5,454 [s2]. For every country except Lebanon, the two databases differed significantly in the number of attacks reported [s2] — which is itself a finding about how poorly this category of harm is counted.

Lebanon was also the outlier on timing. In Myanmar, the occupied Palestinian territory, Sudan and Ukraine, the highest number of attacks fell in the first two months of conflict; in Lebanon, the peak came in the 9-to-11-month period [s2]. Across all five, peaks were associated either with the immediate or early phase of a conflict or with major escalations [s2].

What the study can support

These are administrative and surveillance records, not a prospective registry. Cases are those that reached a participating hospital or were captured by the national mortality system; deaths that never entered either — in areas cut off during the conflict, or where bodies were not brought to facilities — are missing by construction. The bivariate analyses reported show associations between incident status and age, sex, nationality, hospital type, region and injury mechanism, all at p<0.001 [s1], which in a sample of this size establishes that the categories differ, not why.

The comparison period is also unusual: the window opens in October 2023 and closes in March 2025, spanning both lower-intensity exchanges and the escalation of 2024 [s1]. The abstract does not report the split between those phases.

Why it matters

Lebanon entered 2024 with an economy in collapse and a health system that had already lost staff, funding and capacity. The authors situate the findings against that baseline [s1]. The practical implication of an 81% dead-on-arrival share is that post-conflict investment in hospital beds and intensive care would not have changed most of these outcomes; pre-hospital capability — ambulance coverage, bleeding control, safe evacuation corridors — is where the mortality sits.

The two studies together also make a measurement point. A country can produce a nationally representative casualty dataset from routine hospital records while the count of attacks on its health facilities differs by more than a thousand between the two international databases that track it [s2]. The clinical record is, in this instance, the better-instrumented half.

What to watch

Whether the full age-distribution data are published and reconcile the 40–44 figure, and whether national trauma-system planning in Lebanon shifts toward pre-hospital care in response.

This article is informational and is not medical advice.

Sources

• [s1] Al-Hajj S, Harb H, Barakat S, et al. "Injury morbidity and mortality during conflict: evidence from the 2024 war on Lebanon." Conflict and Health, 19(1), published online 17 December 2025. https://doi.org/10.1186/s13031-025-00722-9
• [s2] Usmany J, Barten DG, Goniewicz K, et al. "Attacks on healthcare in conflict-affected countries: a comparison of temporal trends in ongoing conflicts in Lebanon, Myanmar, occupied Palestinian territory, Sudan and Ukraine using WHO SSA and SHCC data, 2018–2024." Population Health Metrics, 24(1), published online 14 December 2025. https://doi.org/10.1186/s12963-025-00442-5