Firearm deaths supply the best lungs in the transplant system.
Donors who died of a gunshot wound give up a lung 2.29 times as often as everyone else, are transplanted at higher rates inside every donation pathway, and their recipients survive the first year slightly better. This is the exact inverse of the drowning-donor deficit, measured the same way.
Source: OPTN/UNOS STAR, Dec 2025 release · 24,890 gunshot-wound donors (median age 26) — 14,043 suicides, 9,500 homicides
9,611 gave a lung against 5686 expected at other-mechanism age-specific rates — a surplus of 3925 donors.
01 · The finding
Above par on every organ, and furthest above on the lung
Share of donors from whom each organ was recovered Gunshot All other
Kidneys 96.8%, livers 88.9%, hearts 61.1%, pancreases 39.4% — every one above the rate for other mechanisms. Lungs are recovered from 38.6% against 16.9%. Conditioning on the donation having proceeded changes nothing: 39.4% against 17.6%.
The clinical logic is unremarkable. A penetrating head injury in a young, previously healthy person produces brain death with an otherwise intact body — short ICU course, little time for ventilator-associated injury, no systemic disease. Median donor age is 26.
02 · Not an age artefact
Roughly double in every age band
Lung recovery by donor age band Gunshot All other
These donors are young, and youth predicts lung recovery for everyone — so age is the obvious explanation to rule out. It does not account for the finding: the advantage holds in every band, and after direct standardisation over the gunshot age distribution the observed/expected ratio is 1.69.
03 · The organs perform
Used more, and marginally better one-year survival
Measure
Gunshot
All other
p
Recovered lungs transplanted
94.8%
91.6%
1.3e-48
— within brain death
95.7%
93.2%
1.1e-33
— within circulatory death
74.5%
70.0%
0.010
Severe PGD at 72 h
32.3% (n=2,122)
32.9%
0.58
1-year graft survival
86.3% (n=6,184)
85.0%
0.009
Unlike the utilisation gap in our asphyxiation brief, this one survives stratification. Gunshot lungs are transplanted more often within brain-death and circulatory-death donation, so it is not confounding by pathway. Early graft function is identical (32.3% vs 32.9%, p = 0.58) — these lungs are not better at resisting reperfusion injury. One-year graft survival is 86.3% against 85.0%; statistically significant on 6,184 grafts, but a 1.3-point difference that no one would act on clinically.
Machine perfusion is used less in this group than average (5.6% against 6.7%) — consistent with lungs that do not need rescuing.
04 · The supply picture
More of them every year, a smaller share of the whole
Era
Gunshot donors
Per year
Share of all donors
Share of lung donors
Lung recovery
2000-2010
7,639
694
9.6%
19.5%
33.9%
2011-2019
7,202
800
8.5%
17.1%
45.9%
2020-2025
5,726
954
6.3%
13.8%
45.7%
Both directions are true and they are easy to confuse. Gunshot donors rose from 694 to 954 per year in absolute terms. Their share of all deceased donors fell from 9.6% to 6.3%, because the donor pool grew faster still — largely through drug-intoxication deaths. This is not a growing share of donation; it is a growing number within a faster-growing whole.
They remain heavily over-represented where it counts: 13.8% of lung donors from 6.3% of donors.
05 · Reading it honestly
A descriptive fact about supply, not a recommendation
This brief reports where transplantable lungs currently come from. It is not an argument about firearms policy in either direction, and nothing here should be read as one. The clinical content is narrow and worth stating plainly: mechanism of death is a poor guide to lung quality in general — our asphyxiation analysis found nothing, and the drowning deficit may reflect caution rather than organ injury — but gunshot-wound donation is the one mechanism where the association with usable lungs is large, consistent across age and donation pathway, and matched by outcome.
The likely explanation is not the bullet. It is what surrounds it: youth, absence of chronic disease, an isolated injury, rapid progression to brain death, and a short ventilated course. Those are the donor characteristics a lung programme actually wants, and this mechanism concentrates them. The same logic predicts that any cause of death sharing that profile would look similar — which is testable, and which we have not yet tested.
Where this sits in the series. Three mechanisms, one method. Asphyxiation: no signal anywhere, and the one significant result was confounding. Drowning: lungs recovered at half the expected rate while every other organ is taken, a gap we can size but cannot adjudicate. Gunshot: recovered at 1.69 times expectation with outcomes to match. Mechanism of death carries real information about lung availability — it just is not the information the injury mechanism itself would suggest.