African swine fever in Nepal: early outbreaks clustered in west and peaked in summer 2022

African swine fever reached Nepal on 19 March 2022, when an outbreak was confirmed in Kirtipur Municipality in Kathmandu district, Bagamati Province. By July 2024, a total of 43 outbreaks had been reported across 21 of the country’s 77 districts, 42 of them in domestic pigs and one in a wild boar from Shuklaphanta National Park in Kanchanpur district of Sudurpaschim Province. A retrospective study now published in the journal Animal Diseases has assembled the first spatial and temporal characterization of these early-phase outbreaks, mapping where the disease appeared, when it peaked, and which environmental variables tracked its occurrence.

The scale of loss on affected farms was severe. Across the 42 domestic pig outbreaks, the total pig population involved was 24,508 animals, of which 17,465 were infected and 16,578 died. That corresponds to an overall mortality rate of 67.6% and a case-fatality rate of 94.9%. African swine fever, caused by the African swine fever virus, is a highly contagious and often fatal disease of domestic and wild pigs, with mortality that can reach 100%. The disease had already spread widely across Asia after emerging in China in 2018, with 18 countries reporting outbreaks, and its arrival in Nepal threatened a sector in which pig farming is an important livelihood for many vulnerable ethnic communities.

The geography of the epidemic was uneven. Outbreaks were reported in all seven provinces of Nepal, but Bagamati Province, home to the Kathmandu Valley, recorded the greatest number, accounting for 42% of the total, or 18 of 43 outbreaks. Sudurpaschim Province in the west followed with 16%, or 7 outbreaks, while Gandaki in central Nepal and Koshi in the east each recorded 14%, or 6 outbreaks. The authors stress that the relatively small number of reported outbreaks over the 28-month study period should be read in context: in China, African swine fever was estimated to have reduced the national pig population by roughly 40%, and in Vietnam more than 20% of the pig population died or was culled within five months of the first outbreak. Nepal’s national pig population is approximately 1.35 million, and differences in surveillance capacity, reporting completeness and case detection may also have shaped the counts, so direct comparisons between countries should be interpreted cautiously.

Farm-level data showed which production systems bore the brunt. Herd sizes on the 42 outbreak farms ranged from five to 8,120 pigs, with a median farm population of 120 pigs and an interquartile range of 41 to 395. Only six farms had more than 1,000 pigs, the threshold considered large-scale in Nepal; two of those were in Koshi Province and four in Bagamati Province. Most outbreaks, 36 of the 42 domestic pig outbreaks, occurred on small- and medium-scale holdings, although commercial farms were also affected. Population sizes were relatively greater in Bagamati, with an interquartile range of 110 to 815, and in Lumbini, with an interquartile range of 240 to 480. The authors conclude that African swine fever poses a significant threat to small and medium-sized farms while large-scale farms remain susceptible, indicating that biosecurity challenges exist across farming systems of every scale.

Mortality and case-fatality figures varied markedly by province. Mortality rates on outbreak farms ranged from a minimum of 1.90%, with a 95% confidence interval of 1.48 to 2.31, to a maximum of 100%, with a confidence interval of 99.95 to 100.00. Case-fatality rates ranged from a minimum of 20%, with a confidence interval of 9.05 to 35.65, to the same maximum of 100%. Karnali Province had the highest median farm mortality rate at 81.82%, followed by Bagamati at 71.01%, while Sudurpaschim had the lowest median mortality rate at 20.50%. For case fatality, Koshi and Madhesh provinces recorded the highest median at 100%, Gandaki followed at 98.28%, and Sudurpaschim was notably lower at 61.43%. The authors suggest that the divergence between mortality and case-fatality trends among provinces might reflect differences in detection timing, reporting practices or farm management rather than viral virulence alone.

The temporal pattern was strongly seasonal. After the first reported outbreak in March 2022, case numbers peaked in June and July 2022, with seven outbreaks in each month. A sharp decline followed, with sporadic cases from August 2022 to May 2023, and no outbreak at all from June to December 2023. Three outbreaks occurred in January 2024, all confined to Tanahun district in Gandaki Province, and no further outbreaks were reported as of July 2024. Grouped by season, the greatest number of outbreaks occurred in summer, spanning June, July and August, which recorded a peak of 18 outbreaks and coincided with the highest temperatures, a maximum of 30 degrees Celsius and a minimum of 20 degrees. Spring, covering March, April and May, also saw a notable 14 outbreaks, while winter and autumn had significantly fewer, at approximately 6 and 4 respectively. The authors note that similar spring and summer peaks have been reported across Europe and Asia regardless of farm size, and they highlight the importance of initiating prevention and preparedness activities before the high-risk summer period.

The decline in reported outbreaks after 2022 coincided with intensified control efforts by Nepal’s Department of Livestock Services. In 2022, despite awareness campaigns and movement control measures, widespread outbreaks occurred, which the authors attribute at least in part to limited coordination and awareness among farmers and other stakeholders in the early stages. The department then conducted intensified campaigns and coordinated with farmers and stakeholders from March to August 2023, after which fewer outbreaks were reported. However, the study is explicit that these findings rest on passive surveillance data, so the observed decline cannot be attributed solely to control measures; changes in surveillance coverage, reporting practices and underreporting may also have contributed. The absence of reported outbreaks from February to July 2024, which coincided with continued control efforts, should likewise not be interpreted as evidence of program effectiveness.

Cluster detection sharpened the spatial picture. Space-time permutation scan statistics identified eight outbreak clusters between 2022 and July 2024, two of which were statistically significant. The most likely cluster lay in Sudurpaschim Province, centered at coordinates 28.998970 N, 80.275381 E, with a radius of 70.26 km encompassing Kailali, Kanchanpur, Doti and Dadeldhura districts. It was active from January to April 2023, with six observed cases against an expected count of 0.84, an observed-to-expected ratio of 7.17 and a p-value of 0.00070. The second most likely cluster was in Bagamati Province, covering the Kathmandu, Bhaktapur and Lalitpur districts, with a radius of 12.87 km. The authors caution that these clusters indicate a nonrandom concentration of reported outbreaks but may reflect differences in pig density, farm connectivity, surveillance coverage, reporting practices or other unmeasured factors rather than specific patterns of disease transmission, so they should be treated as priority areas for field investigation, enhanced surveillance and targeted control rather than as evidence of particular transmission routes.

Environmental associations were examined through an exploratory multivariable Poisson regression using daily rainfall data from the CHIRPS dataset and temperature, dewpoint temperature and relative humidity data from the ERA5-Land dataset. In the univariable stage, rainfall, with a p-value of 0.00339, and relative humidity, with a p-value of 0.1753, met the predefined inclusion threshold, while temperature and dewpoint temperature did not. In the multivariable model, rainfall remained the only variable significantly associated with outbreak counts, at p = 0.0118. The authors propose that increased rainfall may facilitate viral spread through environmental contamination, the movement of people and vehicles, or other mechanical transmission routes, an observation consistent with earlier reports of positive associations between precipitation and African swine fever incidence. Because only 42 outbreaks were available for modeling, statistical power was limited and coefficient estimates may be unstable, so the findings are presented as exploratory and hypothesis-generating, with no causal relationships inferred and confirmation required in larger studies before informing policy.

Several limitations frame the whole analysis. The study relied on passive surveillance data and field reports, meaning the true incidence of African swine fever in Nepal may be underestimated owing to underreporting by farmers and undetected cases, particularly in remote or resource-limited areas. The absence of farm-level biosecurity data, pig movement records and trade network information limited the ability to identify transmission pathways, so the apparent sequence from Kathmandu Valley to eastern and then western Nepal cannot be definitively distinguished from independent introductions through the trade of live pigs and pig products; phylogenetic and animal movement network analyses would be needed to clarify this. The study also covered only the early phase from March 2022 to July 2024, which may not capture long-term trends or later changes in the epidemiological situation. Even so, the authors argue the findings support risk-based surveillance, prioritization of high-risk areas, strengthened farm-level biosecurity and preventive measures ahead of the rainy season, alongside longitudinal studies with active surveillance to refine national prevention and control programs in Nepal.

Subject of Research: Spatiotemporal patterns of early-phase African swine fever outbreaks in Nepal

Article Title: Understanding the spread of African swine fever in Nepal: spatiotemporal analysis of early-phase outbreaks, March 2022–July 2024

Article References: Manandhar, S., Tang, H., Karki, S., Dejyong, T., Charoenlarp, W., & Upadhyaya, M. (2026). Understanding the spread of African swine fever in Nepal: spatiotemporal analysis of early-phase outbreaks, March 2022–July 2024. Animal Diseases, 6(1), Article 60. https://doi.org/10.1186/s44149-026-00272-4

Image Credits: AI Generated

DOI: 10.1186/s44149-026-00272-4

Keywords: African swine fever, Nepal, spatiotemporal analysis, outbreak clusters, rainfall, pig farming, surveillance, Poisson regression, SaTScan, biosecurity, livestock disease, smallholder farms