Organ-donor cases suggest most human rabies deaths go undiagnosed and uncounted

Most human deaths from rabies are never diagnosed as rabies, according to a new statistical framework that uses one of the disease's rarest events — transmission from an infected organ or tissue donor — to estimate how many deaths official surveillance misses [s1]. The method suggests that even China and the United States, with comparatively strong health systems, detected only a minority of their rabies deaths: an estimated 28·7% in China and 6·1% in the United States over the study years [s1].

The result matters because rabies is almost always fatal once symptoms appear, and almost always preventable before them — so an undercounted death is also, usually, a preventable one that the system never registered as a failure.

The idea behind the method

Estimating rabies mortality is notoriously hard because the infection is frequently misdiagnosed or never reported [s1]. The authors' insight is to exploit a grim natural experiment. On rare occasions, a person who died of undiagnosed rabies has become an organ or tissue donor, and the disease was identified only after it was transmitted to recipients [s1]. Because deceased donors are screened and followed far more intensively than the general population, the rate at which rabies turns up among them offers a way to back-calculate how often it goes undetected among everyone else.

Of 33 canine-rabies-endemic and 13 canine-rabies-free countries assessed, rabies was identified among deceased donors in just two: China, with four instances between 2015 and 2024, and the United States, with three between 2000 and 2024 [s1]. Assuming the share of rabies deaths among deceased donors mirrors that among decedents in the general population, the model estimates what proportion of true rabies deaths each country actually diagnosed [s1].

What it found

For 2024, the framework estimated 581 total rabies deaths in China (95% CI 308–1,056) against 167 reported, and 75 in the United States (19–242) against just 4 reported [s1]. The implied detection rates — 28·7% (14·8–56·5%) in China and 6·1% (2·4–21·1%) in the United States — mean the great majority of deaths in both countries went uncounted [s1].

For most of the other countries the parameters were too imprecise to yield a point estimate, but where a country had a relatively large number of deceased donors the method still produced an informative upper bound [s1]. For India in 2023, that upper bound on total rabies deaths was 2,227 [s1].

How much weight it can carry

The authors are candid about the framework's fragility. The estimates are sensitive to an assumed quantity — the relative risk of having died of rabies among deceased donors versus the general population — that is not well measured, and they flag the need for future research to pin it down and sharpen the inference [s1]. The wide confidence intervals are the honest expression of that uncertainty; the lower bounds still sit above the reported counts, which is the durable part of the finding.

The direction, in other words, is robust even where the magnitude is not: reported deaths understate true deaths, in both endemic and rabies-free countries [s1]. That is consistent with the long-standing global picture, in which canine rabies is estimated to cause roughly 59,000 human deaths a year (95% CI 25,000–159,000), more than 3·7 million disability-adjusted life years and about 8·6 billion US dollars in economic losses, concentrated in the poorest regions of the world [s2].

Why undercounting is not a bookkeeping problem

Rabies is one of the few diseases where the tools to eliminate human deaths already exist — dog vaccination is the single most effective intervention, and post-exposure prophylaxis prevents the disease in people who are bitten [s2]. Surveillance is what tells a health system whether those tools are reaching the places that need them. If deaths are being missed at the rates this model implies, then the case for investment looks weaker on paper than it is in reality, and the gaps in dog vaccination and prophylaxis access stay hidden [s1][s2].

The paper's own conclusion is a call to strengthen surveillance and diagnostic capacity and to raise awareness of the risk [s1]. Its more provocative contribution is the demonstration that a country can believe it has rabies nearly under control while diagnosing only a fraction of the deaths it causes.

This article is informational and is not medical advice.

Sources

• [s1] "Estimating total human deaths from rabies: A statistical framework leveraging reports of rabies transmission from deceased organ and tissue donors," PLOS Neglected Tropical Diseases, 21 September 2026. https://doi.org/10.1371/journal.pntd.0014731
• [s2] "Estimating the global burden of endemic canine rabies," PLOS Neglected Tropical Diseases, 16 April 2015. https://doi.org/10.1371/journal.pntd.0003709