Estimating total human deaths from rabies: A statistical framework leveraging reports of rabies transmission from deceased organ and tissue donors
- 1Department of Statistics, University of Oxford, Oxford, United Kingdom
- 2Pandemic Sciences Institute, University of Oxford, Oxford, United Kingdom
Abstract
Estimating the global burden of human rabies is notoriously challenging because the infection is often misdiagnosed or not reported. Although rare, instances of donor-derived human-to-human rabies transmission have been identified after organ or tissue transplantation. Here, we present a statistical framework leveraging these instances to estimate both the probability that a death from rabies is diagnosed for an infected individual and the number of total deaths from rabies. Of the 33 canine-rabies-endemic and 13 canine-rabies-free countries assessed, rabies was identified among deceased donors after transplantation in China (n = 4) between 2015 and 2024 and the United States of America (US) (n = 3) between 2000 and 2024. Assuming the same proportion of deaths from rabies among deceased donors and decedents in the general population, we estimated that China and the US diagnosed rabies in only 28.7% (95%CI: 14.8–56.5%) and 6.1% (95%CI: 2.4–21.1%) of rabies decedents, respectively, during the study years. In 2024, these corresponded to 581 (95%CI: 308–1,056) and 75 (95%CI: 19–242) estimated total deaths from rabies compared to 167 and 4 reported deaths from rabies, respectively. Although parameters were imprecisely estimated for the remaining countries, for countries with a relatively large number of deceased donors, our framework provided informative upper confidence bounds on total deaths from rabies, which can inform the design of surveillance and control measures. For example, in 2023, this upper bound was estimated to be 2,227 deaths from rabies for India. Parameter estimates were sensitive to the assumed relative risk of rabies between deceased donors and decedents in the general population, highlighting the need for future research to estimate this quantity and improve inference within our framework. Our findings show substantial underreporting of human deaths from rabies in both canine-rabies-endemic and canine-rabies-free countries, underscoring the need to strengthen surveillance and diagnostic capacity and raise risk awareness. Author summary: Rabies is a fatal zoonotic viral disease responsible for a substantial mortality burden in humans globally. However, estimating the true global burden of rabies infection in humans is notoriously challenging. In this study, we estimated the total number of deaths from rabies and the probability of diagnosing a death from rabies for an infected individual in canine-rabies-endemic and canine-rabies-free countries. To do so, we used a statistical framework that leverages instances of donor-derived human-to-human rabies transmission events documented in China and the US, alongside data on reported deaths from rabies, deceased donors, and all-cause deaths. We also applied this framework to other countries with no documentation of such events to assess its broader utility and to provide upper bounds on their total number of deaths from rabies. Our results suggest substantial underreporting occurs not only in canine-rabies-endemic countries but also in the US, where most human rabies infections are caused by wildlife exposure, including bats. While the exact estimates of underreporting are sensitive to underlying model assumptions, our study emphasises the importance of enhancing rabies risk awareness and strengthening rabies surveillance and diagnostic capacity.
Related
- Leveraging genomes and metagenome-assembled genomes unveils novel diversity and transmission between public transport and clinical settings in Acinetobacter
- TxCyto: A machine learning framework for estimating cytokine activity from whole transcriptome
- A generative framework for estimating brain energy landscapes from continuous neuroimaging observations.
- A statistical framework for system-level synthesis
- Long-Term Kidney and Patient Survival after Deceased Donor Transplantation in Two Brothers with Alport Syndrome from the Same Donor
BibTeX
@article{kim2026estimating,
title = {{Estimating total human deaths from rabies: A statistical framework leveraging reports of rabies transmission from deceased organ and tissue donors}},
author = {Younjung Kim and Junjie Chen and Alex Yan and Tarek Alrefae and Sumali Bajaj and Christl A. Donnelly},
journal = {PLOS Neglected Tropical Diseases},
year = {2026},
month = sep,
doi = {10.1371/journal.pntd.0014731},
url = {https://journals.plos.org/plosntds/article?id=10.1371/journal.pntd.0014731},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Estimating total human deaths from rabies: A statistical framework leveraging reports of rabies transmission from deceased organ and tissue donors
AU - Younjung Kim
AU - Junjie Chen
AU - Alex Yan
AU - Tarek Alrefae
AU - Sumali Bajaj
AU - Christl A. Donnelly
PY - 2026
DA - 2026/09/21
JO - PLOS Neglected Tropical Diseases
DO - 10.1371/journal.pntd.0014731
UR - https://journals.plos.org/plosntds/article?id=10.1371/journal.pntd.0014731
AB - Estimating the global burden of human rabies is notoriously challenging because the infection is often misdiagnosed or not reported. Although rare, instances of donor-derived human-to-human rabies transmission have been identified after organ or tissue transplantation. Here, we present a statistical framework leveraging these instances to estimate both the probability that a death from rabies is diagnosed for an infected individual and the number of total deaths from rabies. Of the 33 canine-rabies-endemic and 13 canine-rabies-free countries assessed, rabies was identified among deceased donors after transplantation in China (n = 4) between 2015 and 2024 and the United States of America (US) (n = 3) between 2000 and 2024. Assuming the same proportion of deaths from rabies among deceased donors and decedents in the general population, we estimated that China and the US diagnosed rabies in only 28.7% (95%CI: 14.8–56.5%) and 6.1% (95%CI: 2.4–21.1%) of rabies decedents, respectively, during the study years. In 2024, these corresponded to 581 (95%CI: 308–1,056) and 75 (95%CI: 19–242) estimated total deaths from rabies compared to 167 and 4 reported deaths from rabies, respectively. Although parameters were imprecisely estimated for the remaining countries, for countries with a relatively large number of deceased donors, our framework provided informative upper confidence bounds on total deaths from rabies, which can inform the design of surveillance and control measures. For example, in 2023, this upper bound was estimated to be 2,227 deaths from rabies for India. Parameter estimates were sensitive to the assumed relative risk of rabies between deceased donors and decedents in the general population, highlighting the need for future research to estimate this quantity and improve inference within our framework. Our findings show substantial underreporting of human deaths from rabies in both canine-rabies-endemic and canine-rabies-free countries, underscoring the need to strengthen surveillance and diagnostic capacity and raise risk awareness. Author summary: Rabies is a fatal zoonotic viral disease responsible for a substantial mortality burden in humans globally. However, estimating the true global burden of rabies infection in humans is notoriously challenging. In this study, we estimated the total number of deaths from rabies and the probability of diagnosing a death from rabies for an infected individual in canine-rabies-endemic and canine-rabies-free countries. To do so, we used a statistical framework that leverages instances of donor-derived human-to-human rabies transmission events documented in China and the US, alongside data on reported deaths from rabies, deceased donors, and all-cause deaths. We also applied this framework to other countries with no documentation of such events to assess its broader utility and to provide upper bounds on their total number of deaths from rabies. Our results suggest substantial underreporting occurs not only in canine-rabies-endemic countries but also in the US, where most human rabies infections are caused by wildlife exposure, including bats. While the exact estimates of underreporting are sensitive to underlying model assumptions, our study emphasises the importance of enhancing rabies risk awareness and strengthening rabies surveillance and diagnostic capacity.
SN - 1935-2735
ER -