https://orcid.org/0000-0002-2911-9895
Figures
Abstract
Background
Dog-mediated rabies disproportionately affects marginalised and socioeconomically disadvantaged communities. Tea estate (TE) communities in India exemplify one such vulnerable population. Despite their vulnerability, limited research has explored rabies epidemiology within TE settings. This retrospective study uses secondary data to evaluate the incidence of dog bite and their determinants amongst the TE communities in the Udalguri district of Assam state of India.
Methods
Secondary data from 17 to 29 months (January 2022 to May 2024) were retrieved from the hospitals and dispensaries of 11 TE of Udalguri district, Assam. The collected information included dog-bite victims’ demographics and adherence to post-exposure prophylaxis (PEP). Data were analysed using R software, employing descriptive statistics, chi-square tests, odds ratios and mixed-effect logistic regression. Administrative approval was obtained prior to data collection.
Results
A cumulative annual incidence of 11.8 bites per 1,000 individuals was recorded across 11 TE in Udalguri. Children aged ≤15 years accounted for 35% of cases, and dependents were the most affected occupational group (32%). Most exposures involved dogs (66%), and 76% of incidents were bites. Less than half (43%) of victims completed the full PEP regimen of five doses, although 71% received at least three doses. Chi-square analysis indicated that males and children aged ≤ 15 years were more likely to be bitten by dogs compared to other animals than females and the older residents. Children aged ≤ 15 years and non-workers had higher odds of receiving any PEP, while females and children aged ≤ 15 years are more likely to receive at least three doses. In multivariable analysis, females were less likely than males to be bitten by dogs compared to other animals (aOR = 0.4, 95% CI: 0.3–0.7), older individuals had higher odds of completing PEP (aOR = 1.8, 95% CI: 1.2–2.8), and children (≤15 years) were more likely to receive at least three doses of PEP (aOR = 1.9, 95% CI: 1.1–3.3). Temporal analysis showed no clear seasonal pattern, although spikes were observed during winter and monsoon months.
Conclusion
This retrospective study contributes to build the foundation for community-based approach to control dog-mediated rabies in TE by highlighting key epidemiological patterns, demographic vulnerabilities and limitations of the existing intervention implementation delivery among TE communities. We recommend further in-depth investigations to inform the context specific interventions designed to address the unique vulnerabilities, thereby reducing the risk of rabies specifically in tea -estate populations.
Author summary
This retrospective study analysed secondary data of hospital and dispensary records from 11 tea estates in Assam’s Udalguri district between January 2022 and May 2024 to estimate the incidence of dog bites, profile affected individuals, and assess access to and completion of post-exposure prophylaxis (PEP). The findings bring to light the gaps in rabies prevention, particularly in remote areas and underserved tea -estate communities, where delayed or incomplete PEP administration significantly increases the risk of rabies-related fatalities. The study highlights the pressing need to expand rabies surveillance to include marginalised populations, strengthen health systems in remote areas, improve PEP availability and adherence, and design community-specific strategies. By identifying barriers to timely and complete adherence to PEP, this study provides actionable insights such as expanding the research base to include vulnerable communities such as tea-estate communities, developing tailormade strategies, and better adherence to post-bite prophylaxis to guide policy-making and the implementation of targeted interventions to reduce the burden of rabies in such high-risk populations.
Citation: Tiwari HK, Mohanty P, Shirke RG, Kurup AU, Satwik C, Kindo SP, et al. (2025) Two-leaves and many bites: Profiling dog-bites and adherence to rabies prophylaxis in tea-estate communities of Udalguri District, Assam, India. PLoS Negl Trop Dis 19(12): e0013791. https://doi.org/10.1371/journal.pntd.0013791
Editor: Richard A. Bowen, Colorado State University, UNITED STATES OF AMERICA
Received: July 16, 2025; Accepted: November 21, 2025; Published: December 1, 2025
Copyright: © 2025 Tiwari et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are in the manuscript and its Supporting information files.
Funding: This work was supported by DBT Wellcome Trust India Alliance Intermediate Fellowship to HKT (IA/CPH/21/1/505957). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
Rabies is endemic to India, with approximately 99% of human rabies deaths resulting from dog bites [1]. The disease disproportionately affects marginalised and socioeconomically disadvantaged populations who often lack awareness of the consequences of untreated dog bites and have limited capacity to seek timely medical care [2–5]. Moreover, free-roaming dogs (FRD) often reside near the residential settlements of such vulnerable communities, significantly increasing the risk of dog bites [2,6–9]. Post-exposure prophylaxis (PEP) is the only effective practice to prevent the onset of rabies following a suspected exposure, provided it is administered promptly and appropriately [10]. However, the lack of availability, difficult accessibility, and poor affordability of PEP accentuate the challenging circumstances, leading to mortality [11]. These challenges get pronounced in remote and underserved regions, where health infrastructure is often inadequate or absent, leaving vulnerable communities without access to critical care [12,13].
Several countries have developed action plans for eliminating dog-mediated rabies in response to the ‘Zero by 30’ campaign, initiated under the tripartite agreement between WHO (World Health Organization), FAO (Food and Agriculture Organization), and the WOAH (World Organisation of Animal Health). While many of these plans address the disease in vulnerable populations, few explicitly identify or define these communities.
Addressing dog-mediated rabies in vulnerable populations requires confronting the broader social inequities in healthcare access before developing strategies that can be equitably scaled at the national level [14]. However, one of the major impediments to the judicious allocation of resources for implementing interventions for rabies control is the lack of reliable surveillance data, especially from the most affected communities [15,16]. While population-based data on rabies in humans and FRD provide some insight into communities’ attitude about the disease and FRD, a differentiated policy assessment grounded in principles of health equity is essential for designing context-sensitive interventions that respond to the diverse needs of affected populations [17].
Assam, located in northeast India, is one of the country’s major tea-producing states where workers are employed in various capacities, including labour-intensive roles such as manually plucking tea leaves [18]. The tea estates (TE) are usually remotely located, and in many instances, far from accessible health centres equipped with adequate facilities [19]. However, some TE do have health centres for workers within the estate. Tea -estate workers are predominantly from socioeconomically disadvantaged backgrounds, many of whom are illiterate and work under gruelling circumstances that increase their vulnerability to animal attacks, including dog bites [20]. Despite representing a significant public health risk in these remote communities, the epidemiology of dog bites and rabies remains poorly documented. It is important to bridge such knowledge gaps to institute informed and targeted interventions by analysing existing data and adopting a spatial approach instead of a generalised one [17,20–22].
This study explores the epidemiology of dog bites and adherence to rabies prophylaxis in tea estate communities of the Udalguri district, Assam, India, using secondary data. The study hypothesized that:
- (a) Children and dependents have higher incidence of dog-bites compared to adults and workers
- (b) Females and children are less likely to complete the full regimen than males and older residents
- (c) Access to and completion of PEP vary across demographic and occupational groups within the tea-estate communities.
The study is guided by the above-mentioned hypotheses to evaluate dog-bite incidence, profile at risk populations, and assessment of PEP accessibility and adherence.
2. Materials and methods
Ethics statement: The study was approved by the Institutional Human Ethics Committee (IHEC) of the Indian Institute of Technology Guwahati (IHEC/2024/03). Administrative approval for data collection was obtained from the respective TE authorities before commencing the study.
2.1. Study area and approvals
This study was conducted in the Udalguri district in the North Assam division of Assam, northeast India (26.7452°N and 92.0962°E). The district shares an international border with Bhutan and a state boundary with Arunachal Pradesh (Fig 1). Udalguri is home to 22 TE, of which 11 consented to participate. These estates vary in size and workforce composition and are representative of the district’s tea-producing landscape.
2.2. Data collection
Secondary data were retrieved from TE dispensaries and hospital records from January 2022 to May 2024. Although these records were maintained in the dog-bite registers but also included bites from other animals suspected of transmitting rabies (cats, rats, or mongooses). Data were manually extracted from hardcopy registers and entered into Microsoft Excel. The collected data included the demographics of the bite victims, details of PEP, species of the biting animals, and the location of bites. Identifying information such as names and address of the bite victims were excluded. Three TEs maintained data from January 2022 (29 months), while eight provided records from January 2023 (17 months). The dataset was deemed adequate for the planned analyses.
2.3. Inclusion criteria and variable classification
Data entries containing at least two explanatory variables and three outcome variables were included for the regression analysis. Explanatory variables comprised age, gender, and employment status; outcome variables included adherence to the PEP regimen (complete/incomplete), the number of doses administered (≥3/ < 3), and species of the biting animal (dog/others). Age was categorised into five groups: ≤ 15 years, 16–30 years, 31–45 years, 46–60 years and >60 years. Employment status comprised ‘permanent workers’, ‘temporary workers’, ‘non-workers’ (adult family members or visitors not directly employed), and ‘dependents’ (family members ≤ 18 years). Average monthly temperature and relative humidity for the study period were obtained from Climate & Weather Averages in Udalguri, Assam, India [23].
2.4. Data analysis
Dog-bite incidence was calculated as the number of bites per 1,000 individuals per year. Chi-square goodness-of-fit tests compared observed and expected bite incidences for each explanatory variable. Chi-square tests of independence and corresponding odds ratios were used to examine associations between explanatory and outcome variables.
Three mixed-effects logistic regression analyses were conducted—each for one outcome variable (biting species, completion of PEP, and receipt of ≥ 3 doses)—with tea estate as a random effect to account for clustering. The model diagnostics were assessed through stepwise comparison of the mixed-effect models with generalised linear models (GLMs) with the same outcome and explanatory variables, using Akaike Information Criterion (AIC), Type-II Wald chi-square tests, and marginal and conditional R2 estimates. A Poisson regression model evaluated the effect of employment status on the bite incidences. Correlations between monthly dog-bite counts and average temperature and humidity were tested using Pearson’s correlation coefficient. All analyses were performed in R (version 4.3 or later) using the packages epiR, epitools, tidyverse, lme4, ggplot2, DHARMa, MuMIn, and patchwork [24–32].
3. Results
3.1. Incidence of dog-bites
Across the 11 TEs, a total of 1,187 bite cases were recorded. Annual incidence rates ranged from 1.5 to 36.9 bites/1000 individuals across the 11 TEs, with Hattigore TE recording the highest, and Betali TE the least (Table 1). Permanent staff and the dependents recorded the highest incidence (19.8/1000 individuals per year) while the temporary staff recorded the lowest.
3.2. Demographic and occupational distribution of bite victims and completion of PEP
Among children aged ≤ 15 years, 160 (38%) were ≤ 5 years, including one 9-month-old infant. Dependents constituted the most affected group (32%, n = 375), whereas temporary workers were the least affected (4%, n = 47). Bite counts differed significantly across most demographic categories under an equal-distribution assumption (p < 0.0001), except gender. Males (49%) and females (47%) were bitten at similar frequencies (p = 0.63) (Table 2).
3.3. Association tests
All 1,187 records were analysed for associations between the available predictor variables (gender, age, employment status) and outcomes (biting species, PEP completion, and ≥ 3 doses). Female victims were less likely than males to be bitten by dogs compared to other species (79% vs 88%) but were more likely to receive ≥ 3 PEP doses compared to less than three doses (Table 3). Children ≤ 15 years had higher odds of being bitten by dogs compared to other species, completing the full regimen, and receiving ≥ 3 doses compared with older individuals. Non-workers showed reduced odds of completing the PEP regimen, and along with temporary workers, were less likely to receive ≥ 3 doses compared to dependents.
3.4. Mixed effects logistic regression analysis and adjusted odd ratios
A total of 824 entries that included all explanatory and outcome variables were considered to construct the generalised mixed effects logistic regression models. Model comparison and diagnostic statistics consistently supported the use of a generalised linear mixed-effects modelling framework over a standard generalised linear model. Across all three outcomes, inclusion of tea estate as a random effect substantially improved model fit and accounted for intra-estate clustering. For the biting species model, the AIC decreased from 716.2 (GLM) to 708.7 (GLMM), with the conditional R2 rising from 0.045 to 0.27, indicating that estate-level differences explained roughly a quarter of the total variance. For PEP completion, the AIC reduction was more pronounced (1140.2 to 1089.1), and the conditional R2 increased from 0.013 to 0.54, suggesting that over half of the variability in regimen completion was attributable to estate-level factors rather than individual covariates. The strongest clustering was observed for the ≥ 3 vaccine doses model, where AIC dropped sharply from 728.1 to 592.9 and the conditional R2 rose from 0.009 to 0.88, indicating dominant contextual influences at the estate level. Collectively, these metrics demonstrate that mixed-effects models more accurately captured the hierarchical structure of the data, with tea-estate-specific contexts exerting a substantial influence on all three outcomes. The multivariable models and the adjusted odds ratios are depicted in Fig 2).
- (a) Being bitten by a dog compared to other animals: In the univariable analysis, females had lower odds than males of being bitten by a dog compared other species (OR = 0.4, 95% CI: 0.3–0.6; p = 0.0004). After adjusting for age and employment status in the multivariable GLMM, this association remained significant (aOR = 0.4, 95% CI: 0.3–0.7; p < 0.001). Neither age group nor employment status showed a significant association with the biting species. The random intercept variance for tea estate was 1.02, and the intra-class correlation coefficient (ICC) was 0.24 (Table 4).
- (b) Completing the PEP regimen: In the univariable analysis, age and employment status were significant in univariable analysis, but only age remained so in multivariable models. However, in the mixed -effects model, after adjusting for gender and employment status, only age remained a significant predictor. Children aged ≤15 years had twice the odds of completing the vaccine regimen compared with adults (aOR = 2.0, 95% CI 1.2–3.1; p = 0.003). No strong evidence emerged to suggest an association between employment categories and completion of PEP after adjusting for covariates (Table 5). The random intercept variance was 3.7, with an intra-class correlation coefficient (ICC) of 0.53.
- (c) At least three doses of vaccines: In the mixed-effects model, younger age and employment status were associated with receiving ≥3 vaccine doses. Children aged ≤15 years had higher adjusted odds (aOR = 2.2, 95% CI 1.2–4.3; p = 0.01), whereas permanent workers had lower odds (aOR = 0.3, 95% CI 0.1–0.8; p = 0.02) compared with dependents. The high random-effect variance (ICC = 0.88) (Table 6).
3.5. Distribution of bites and employment status
Dependents had the highest odds of getting bitten. Compared with dependents, non-workers (OR 0.5, 95%CI 0.4-0.7, p < 0.0001), and temporary workers (OR 0.2 95%CI 0.1 -0.3, p < 0.0001) had significantly lower odds, while permanent staff did not differ significantly.
3.6. Monthly and seasonal variation in bite incidence
The monthly distribution of bite cases across TEs is shown in Fig 3. Hattigore TE consistently reported higher bite counts than the combined total of all other TEs. Spikes were observed during winter (Dec–Feb) in both 2023 and 2024, with a secondary rise during the monsoon (Jun–Aug). Post-monsoon months (Sep–Nov) showed a contrasting decline. Weak negative correlations were observed between monthly bite counts and average temperature (r = –0.18) and humidity (r = –0.20).
The highlighted points are peaks of bite cases above the respective means.
4. Discussion
4.1. Overview
Tea-estate -communities face unique health challenges including rabies-related deaths resulting from animal bites. This study, the first in this domain comprehensively evaluates incidences of animal bites in the TE community in a district of Assam state in India. It demonstrates high vulnerabilities to potential dog-mediated rabies and highlights the disparities in adherence to preventive practices attributable to spatial, demographic and behavioural community practices, while underscoring the need for robust animal bite reporting systems that encompasses One Health programming and aggressive rabies awareness campaigns for vulnerable communities such as TE workers.
A markedly high annual bite incidence rate (36.9/1000 individuals) was reported in Hattigiore, TE, more than twice the overall annual incidence rate (16.0/1000 individuals). A consistently elevated monthly bite number (Fig 3) suggests the presence of a persistent risk due to possibly a high FRD population supported by inefficient waste management similar to other rural and semi-rural milieu of the country. A better reporting and record-keeping in Hattigore TE, compared to other sites, where records of bite care were poorly maintained is also a reason for a comparative high incidence along with the facilitation of easy and no-cost availability of PEP by the TE administration, as anecdotally corroborated by the estate welfare manager. While the availability of PEP in one TE promotes better record maintenance, it emphasizes the challenges of uniform provision of prophylaxis across other TEs. A structured bite-reporting system linked with PEP uptake is recommended for all TEs to help ascertain the true prevalence of bite [33]. The variability of reporting, and the proclivity of TE workers to avoid seeking medical assistance following a bite could also be due to variability in the intensity of sensitising the worker communities to compulsorily report exposures, even if it is just a scratch. It was evident from records at Hattigore and Dimakuchi TEs that the number of scratches reported was comparatively higher than in other estates. Enhancing sensitivity about fatal consequences of ignored dog-bite exposures has shown improved reporting [34–37].
Elevated annual incidence in Dimakuchi (26.5/1000 individuals), Paneery (14.7/1000 individuals), and Bhutiachang (11.6/1000 individuals) is likely attributable to the high density of FRD closer to urban centre, which support a greater carrying capacity due to dense human populations, as compared to TE along national and international borders that have sparse human habitation [38]. Interestingly, Corramore TE, located along the district periphery, also exhibited high incidence (15.1/1,000), contradicting the assumption that remoteness from human habitats implies reduced risk. This trend may be explained by the transboundary movement of FRDs, especially given Corramore’s proximity to the international border with Bhutan [39,40]. These findings highlight the need for systematic studies assessing FRD movement and densities in border regions to elucidate spatial heterogeneities in bite risk. The high incidence among permanent workers (19.0/1000 individuals per year) may be attributed to the outdoor nature of their work, which requires frequent movements and prolonged interaction with environments shared by FRD, thereby increasing their vulnerability [41]. In contrast, the similar high incidence among the ‘dependent’ category - which includes comprises children, housewives, and the elderly may be explained by multiple factors. While an increased exposure of children and elderly is plausible, this category may also have a higher likelihood if reporting bite cases compared to other groups.
The disproportionate representation of children aged ≤15 (37%) is consistent with previous studies globally and nationally. Children are particularly vulnerable to dog bites due to their spontaneous affection for animals, smaller stature and lack of awareness about safe interactions [42,43]. However, a significant concern is that 38% of these children are under 5 years’ age, including a 9-month-old baby. One likely reason for such a high proportion of young children exposed to animal bites is the lack of supervision when parents leave for work [44]. Many of these bites, which could potentially result in rabies, may go unreported [44–48]. Introducing creche facilities in TE could help reduce dog bites among young children by ensuring supervision.
4.2. Determinants of bite exposure and PEP adherence
Significant associations emerged between the demographic variables (age and gender) with exposure to dog bites and adherence to PEP (Table 3). When accounting for intra-estate clustering using mixed-effects logistics regression, gender and age emerged as significant predictors of exposure and PEP adherence. A moderate to substantial TE level heterogeneity indicates strong estate-oriented context in rabies prevention strategies (Tables 4–6). Higher exposure of males to bites by dogs can be due to the heightened exposure to dogs because of outdoor work patterns [49]. A moderate clustering effect (ICC = 0.24) indicated estate level variations such as dog population density and waste-management influencing dog-human interactions [50,51]. While females and children aged ≤ 15 years and non-workers show higher odds of receiving post-bite PEP, lack of adherence in males, permanent employees, and dependents reflect potential barriers and differences in healthcare access or compliance [11,52].
The low completion rate among the ‘dependents’ may be due to the financial burden faced by parents or heads of families of losing daily wages to accompany children or other dependents to the hospital for PEP. Nonetheless, the higher odds of children <15 years for receiving >3 doses is a contrasting finding and hence, the speculation of parents’ apprehension of losing wages warrants further investigation. On the contrary, ‘non-workers’ are also at higher odds to receive at least three doses of vaccines than ‘dependent’ counterparts. It is a surprising finding because, as opposed to the ‘non-workers’, the ‘dependents’ are provided the vaccines free of cost and adherence to PEP generally improves when supplied at reduced or no cost [10,53]. However, Hattigore TE is an exception where PEP was free even to the ‘non-workers’, explaining the increased overall odds. The ‘non-workers’ category includes the unemployed population which may have greater flexibility in attending vaccination appointments without the concern of wage loss, hence the high odds compared to other categories. Alternatively, this category may be employed in informal or off-estate employment potentially earning more than tea -estate workers thus being better able to pay for the vaccination costs. However, these reasonings remain speculative and call for exploratory studies to shed more light on the motivation of ‘non-workers’ in this context.
The ‘age’ variable is an important determinant affecting PEP regimen compliance or receiving at least three doses, especially when adjusted for gender and employment status. This finding emphasizes the need for age-specific interventions, such as targeted health messaging and strategies promoting PEP adherence to ensure that children and their caregivers are adequately informed about rabies prevention and management. The demographic variables especially the ‘employment status’ evinced nuanced interdependencies post multivariable analyses adjustment. A structured investigation such as questionnaire surveys on awareness about rabies and post dog-bite practices are recommended to further explore the mutual interdependence of socio-demographic variables influencing exposure to dog bites by using multivariable framework.
4.3. Estate-level clustering and model implications
The use of mixed effects models appears justified as it revealed moderate (ICC = 0.24) to strong clustering effect (ICC = 0.88). While moderate clustering suggests influence of estate-level factors like dog density, the high ICC for PEP completion underscores contextual barriers such as health care access, counselling quality or vaccine availability among the different TE. Further, a very high clustering effect indicates that estate-specific contextual factors such as location of vaccination points, and health care facilitation, and awareness enhancing measures strongly influence adherence to PEP. These findings demonstrate that more than individual influences, determinants at the estate level that fall in the domain of the estate management substantially shape the PEP completion rates. It is recommended that the interventions aimed at rabies elimination in TE settings should incorporate both individual- and system-level components: targeted communication for adult men and workers, flexible vaccination delivery mechanisms that accommodate labour schedules, and estate-specific micro-planning to address local barriers.
4.4. Seasonality of bites in tea-estates
Although no clear seasonal trend emerged, notable spikes during the winter (Dec- Feb) and monsoon (Jun-Aug) seasons suggest changes in canine activity which presumably increases the bite risk. Nonetheless, the potential utility of seasonally tailored community intervention programs, enhanced surveillance during high-risk periods, and timely rabies vaccination drives in canines and human populations prior to high-risk periods cannot be undermined [54]. As the data were collected from only a few estates in a single district of Assam, a key limitation of this study, a further exploration of seasonality of dog-bites is recommended [55].
4.5. Operational barriers and health seeking behaviour
The informal interactions with the health staff at a TE during the data collection revealed, a persistent challenge highlighting inconsistent availability of PEP in healthcare facilities. Bite victims are often directed to private pharmacies, where costs and travel distances become deterrents to completing the full regimen. Similar barriers such as remoteness and limited vaccine availability have been widely cited in many studies [56–58]. The recent recommended shift to a reduced dosage of PEP administered intradermally instead of the intramuscular route could be a solution to economise PEP usage in such high-risk communities [10,11,59]. However, feasibility studies considering bite severity and injury categories are necessary before broader implementation. Another concern raised by healthcare staff is the reliance on traditional healers resulting in underreporting and non-completion of PEP. Engaging the traditional healers through citizen science initiatives may enhance case reporting and even encourage timely referral to healthcare facilities, an initiative [60–62]. Such strategies, though speculative, hold promise and warrant field-testing.
A strong anecdotal perception persists in the TE that the number of FRD, semi-owned and feral, is high among communities, and residents are unaware of anti-rabies immunisation in dogs. Tea-estates are rarely included in rabies awareness campaigns, and no Animal Birth Control (ABC) programs are rarely conducted. Such interventions are cardinal to prevent and control dog-mediated rabies. We recommend initiating these interventions along with need assessment studies to tailor interventions to the specific needs of TE communities [63–65].
4.6. Limitations and future directions
Although limited by retrospective nature and deficient data due to weak bite recording mechanisms in the TE, this study highlights the need for a comprehensive, focused, demographically sensitive approach to rabies prevention in the TE communities. We recommend investigative research exploring the factors influencing the high incidence of dog-mediated rabies in such vulnerable communities, as few such studies exist. The likely logistical barriers hindering PEP access to healthcare facilities must be removed [58,66]. At the same time, rabies awareness and relevance to One Health need emphasis to devise robust strategies that reduce dog bites and potential rabies deaths [67–69]. Educational campaigns should be prioritised for vulnerable populations, caregivers and parents [67]. A robust dog-bite surveillance system sensitive to spatial and temporal trends is recommended [70,71].
These findings can be a harbinger for future research to expand its geographic scope to include a broader representation of tea -estate communities across Assam and other parts of India. Mixed-method approaches including household surveys, structured interviews with tea -estate workers, and focus group discussions should be employed to uncover the socio-economic determinants and behavioural patterns contributing to high dog-bite incidence and poor PEP adherence. Studies focusing on the ecology of FRD populations, their density, movement, and interactions with tea -estate residents can help design more effective ABC and vaccination interventions. Socio-economic variables, such as household income, rabies-related awareness, and access to healthcare, should be systematically explored to identify barriers and facilitators of rabies control. Additionally, longitudinal studies are needed to assess seasonal variations in bite incidents and evaluate the long-term impact of community-based interventions like rabies education drives and child-safe creche facilities. This study lays the foundation for deeper inquiry into the vulnerability of tea -estate communities to dog-mediated rabies and the development of integrated, One Health-aligned solutions to mitigate this public health burden.
Supporting information
S1 Data. Excel sheet with tea -estate retrospective data_Udalguri.
https://doi.org/10.1371/journal.pntd.0013791.s001
(XLSX)
Acknowledgments
We acknowledge the authorities of the participating TEs, the medical officers and other ancillary staff for their invaluable support in providing information and medical records. We want to acknowledge the contributions of Dipankar Gogoi, Aasib Daimari, and Sumi Basumatary for data collection.
References
- 1. Radhakrishnan S, Vanak AT, Nouvellet P, Donnelly CA. Rabies as a Public Health Concern in India-A Historical Perspective. Trop Med Infect Dis. 2020;5(4):162. pmid:33096767
- 2. Daigle L, Ravel A, Rondenay Y, Simon A, Mokoush KN, Aenishaenslin C. Knowledge, attitudes, and practices regarding dogs and dog bites in Indigenous northern communities: A mixed methods study. Front Vet Sci. 2023;10:1080152. pmid:36891468
- 3. Tiwari HK, O’Dea M, Robertson ID, Vanak AT. Knowledge, attitudes and practices (KAP) towards rabies and free-roaming dogs (FRD) in Shirsuphal village in western India: A community based cross-sectional study. PLoS Negl Trop Dis. 2019;13(1):e0007120. pmid:30682015
- 4. Ćetković J, Žarković M, Knežević M, Cvetkovska M, Vujadinović R, Rutešić S, et al. Financial and Socio-Economic Effects of Investment in the Context of Dog Population Management. Animals (Basel). 2022;12(22):3176. pmid:36428403
- 5. Raghavan M, Martens PJ, Burchill C. Exploring the relationship between socioeconomic status and dog-bite injuries through spatial analysis. Rural Remote Health. 2014;14(3):2846. pmid:25124792
- 6. Tiwari HK, Robertson ID, O’Dea M, Vanak AT. Knowledge, attitudes and practices (KAP) towards rabies and free roaming dogs (FRD) in Panchkula district of north India: A cross-sectional study of urban residents. PLoS Negl Trop Dis. 2019;13(4):e0007384. pmid:31034474
- 7. De la Puente-León M, Levy MZ, Toledo AM, Recuenco S, Shinnick J, Castillo-Neyra R. Spatial Inequality Hides the Burden of Dog Bites and the Risk of Dog-Mediated Human Rabies. Am J Trop Med Hyg. 2020;103(3):1247–57. pmid:32662391
- 8. Warembourg C, Fournié G, Abakar MF, Alvarez D, Berger-González M, Odoch T, et al. Predictors of free-roaming domestic dogs’ contact network centrality and their relevance for rabies control. Sci Rep. 2021;11(1):12898. pmid:34145344
- 9. Gutiérrez-Zapata S, Santoro S, Gegundez-Arias ME, Selva N, Calzada J. Dog invasions in protected areas: A case study using camera trapping, citizen science and artificial intelligence. Global Ecol Conserv. 2024;54:e03109.
- 10. Nadal D, Bote K, Masthi R, Narayana A, Ross Y, Wallace R, et al. Rabies post-exposure prophylaxis delivery to ensure treatment efficacy and increase compliance. IJID One Health. 2023;1:100006. pmid:38152594
- 11. Changalucha J, Steenson R, Grieve E, Cleaveland S, Lembo T, Lushasi K, et al. The need to improve access to rabies post-exposure vaccines: Lessons from Tanzania. Vaccine. 2019;37(Suppl 1):A45–53. pmid:30309746
- 12. Evans MV, Andréambeloson T, Randriamihaja M, Ihantamalala F, Cordier L, Cowley G, et al. Geographic barriers to care persist at the community healthcare level: Evidence from rural Madagascar. PLOS Glob Public Health. 2022;2(12):e0001028. pmid:36962826
- 13. Baazeem M, Kruger E, Tennant M. Current status of tertiary healthcare services and its accessibility in rural and remote Australia: A systematic review. Health Sci Rev. 2024;11:100158.
- 14. Miranda MEG, Miranda NLJ. Rabies Prevention in Asia: Institutionalizing Implementation Capacities. Rabies Rabies Vacc. 2020:103–16.
- 15. Changalucha J, Hampson K, Jaswant G, Lankester F, Yoder J. Human rabies: prospects for elimination. CAB Rev. 2021;16:039. pmid:34765015
- 16. Brownson RC, Fielding JE, Green LW. Building Capacity for Evidence-Based Public Health: Reconciling the Pulls of Practice and the Push of Research. Annu Rev Public Health. 2018;39:27–53. pmid:29166243
- 17. Arias Caicedo MR, Xavier D de A, Arias Caicedo CA, Andrade E, Abel I. Epidemiological scenarios for human rabies exposure notified in Colombia during ten years: A challenge to implement surveillance actions with a differential approach on vulnerable populations. PLoS One. 2019;14(12):e0213120. pmid:31881039
- 18.
Government of Assam. About Tea Industries | Industries & Commerce. n.d. [cited December 5, 2024]. https://industries.assam.gov.in/portlet-innerpage/about-tea-industries
- 19. Kalarivayil R, Chattaraj B, Nair SS. Space and Place precarity in the Global South: The Case of Tea Garden Workers in Assam. Soc Inclusion. 2024;12. Available from: https://www.cogitatiopress.com/socialinclusion/article/view/7776
- 20. Khayli M, Lhor Y, Bengoumi M, Zro K, El Harrak M, Bakkouri A, et al. Using geostatistics to better understand the epidemiology of animal rabies in Morocco: what is the contribution of the predictive value? Heliyon. 2021;7(1):e06019. pmid:33537478
- 21. Li D, Liu Q, Chen F, Jiang Q, Wang T, Yin X, et al. Knowledge, attitudes and practices regarding to rabies and its prevention and control among bite victims by suspected rabid animals in China. One Health. 2021;13:100264. pmid:34036144
- 22. Zhao J, Luo M, Tan X, Zhu Z, Zhang M, Liu J, et al. Spatial accessibility and inequality analysis of rabies-exposed patients to rabies post-exposure prophylaxis clinics in Guangzhou City, China. Int J Equity Health. 2024;23(1):122. pmid:38877457
- 23.
Udalguri Weather Averages - Assam, IN. n.d. [cited July 4, 2025]. Available from: https://www.worldweatheronline.com/en-in/udalguri-weather-averages/assam/in.aspx
- 24.
Stevenson M, Sergeant E. epiR: Tools for the Analysis of Epidemiological Data. 2025. https://doi.org/10.32614/cran.package.epir
- 25.
Aragon TJ, Fay MP, Wollschlaeger D, Omidpanah A. Epitools: Epidemiology Tools. V. 0.5-10.1. 2020. Released March 22. Available from: https://cran.r-project.org/web/packages/epitools/index.html
- 26.
Aragon TJ, Fay MP, Wollschlaeger D, Omidpanah A. Epitools: Epidemiology Tools. 2020. Available from: https://cran.r-project.org/web/packages/epitools/index.html
- 27. Wickham H, Averick M, Bryan J, Chang W, McGowan L, François R, et al. Welcome to the Tidyverse. JOSS. 2019;4(43):1686.
- 28. Bates D, Mächler M, Bolker B, Walker S. Fitting Linear Mixed-Effects Models Usinglme4. J Stat Soft. 2015;67(1):1–48.
- 29.
Wickham H. Ggplot2: Elegant Graphics for Data Analysis. New York: Springer-Verlag; 2016. released. Available from: https://ggplot2.tidyverse.org
- 30.
Florian H. DHARMa: Residual Diagnostics for Hierarchical (Multi-Level/ Mixed) Regression Models. R Package Version 0.4.7. 2024. Released. Available from: https://CRAN.R-project.org/package=DHARMa
- 31.
Kamil B. MuMIn: Multi-Model Inference. 2025. https://doi.org/10.32614/CRAN.Package.MuMIn. Released.
- 32.
Thomas LP. “Patchwork: The Composer of Plots.” R package version 1.3.1. 2025. Available from: https://doi.org/10.32614/CRAN.package.patchwork
- 33. Swedberg C, Mazeri S, Mellanby RJ, Hampson K, Chng NR. Implementing a One Health Approach to Rabies Surveillance: Lessons From Integrated Bite Case Management. Front Trop Dis. 2022;3:829132. pmid:36945698
- 34.
Lakestani NN. A Study of Dog Bites and Their Prevention. 2007. Available from: https://era.ed.ac.uk/handle/1842/2668
- 35. Barrios CL, Aguirre V, Parra A, Pavletic C, Bustos-López C, Perez S, et al. Systematic Review: Comparison of the Main Variables of Interest in Publications of Canine Bite Accidents in the Written Press, Gray and Scientific Literature in Chile and Spain, between the Years 2013 and 2017. Animals (Basel). 2021;11(3):893. pmid:33800962
- 36. Kisaka S, Makumbi FE, Majalija S, Bangirana A, Thumbi SM. Epidemiology and preclinical management of dog bites among humans in Wakiso and Kampala districts, Uganda: Implications for prevention of dog bites and rabies. PLoS One. 2020;15(9):e0239090. pmid:32956373
- 37. Ghosh S, Chowdhury S, Haider N, Bhowmik RK, Rana MS, Prue Marma AS, et al. Awareness of rabies and response to dog bites in a Bangladesh community. Vet Med Sci. 2016;2(3):161–9. pmid:29067191
- 38. Guilloux AGA, Panachão LI, Alves AJS, Zetun CB, Cassenote AJF, Dias RA. Stray dogs in urban fragments: relation between population’s perception of their presence and socio-demographic factors. Pesq Vet Bras. 2018;38(1):89–93.
- 39. Hughes J, Macdonald DW. A review of the interactions between free-roaming domestic dogs and wildlife. Biol Conserv. 2013;157:341–51.
- 40. Raynor B, De la Puente-León M, Johnson A, Díaz EW, Levy MZ, Recuenco SE, et al. Movement patterns of free-roaming dogs on heterogeneous urban landscapes: Implications for rabies control. Prev Vet Med. 2020;178:104978. pmid:32302776
- 41. Owczarczak-Garstecka SC, Christley R, Watkins F, Yang H, Bishop B, Westgarth C. Dog bite safety at work: An injury prevention perspective on reported occupational dog bites in the UK. Safety Sci. 2019;118:595–606.
- 42. Monti L, Kotzalidis GD, Arcangeli V, Brozzi C, Iacovino R, Giansanti C, et al. Psychological Sequelae of Dog Bites in Children: A Review. Children (Basel). 2024;11(10):1218. pmid:39457183
- 43. Vučinić M, Vučićević M. Children are victims of dog bites due to irresponsible dog ownership, parenthood, and managers of school institutions in Serbia. J Veterinary Behav. 2019;30:61–8.
- 44. Morrongiello BA, Schwebel DC, Stewart J, Bell M, Davis AL, Corbett MR. Examining parents’ behaviors and supervision of their children in the presence of an unfamiliar dog: does The Blue Dog intervention improve parent practices? Accid Anal Prev. 2013;54:108–13. pmid:23499982
- 45. Jakeman M, Oxley JA, Owczarczak-Garstecka SC, Westgarth C. Pet dog bites in children: management and prevention. BMJ Paediatr Open. 2020;4(1):e000726. pmid:32821860
- 46. Bernardo LM, Gardner MJ, Rosenfield RL, Cohen B, Pitetti R. A comparison of dog bite injuries in younger and older children treated in a pediatric emergency department. Pediatr Emerg Care. 2002;18(3):247–9. pmid:12066018
- 47. Shields WC, McDonald EM, Stepnitz R, McKenzie LT, Gielen AC. Dog bites: an opportunity for parent education in the pediatric emergency department. Pediatr Emerg Care. 2012;28(10):966–70. pmid:23023457
- 48. Cavalcanti AL, Porto E, Dos Santos BF, Cavalcanti CL, Cavalcanti AFC. Facial dog bite injuries in children: A case report. Int J Surg Case Rep. 2017;41:57–60. pmid:29035774
- 49. Thahaby N, Akand AH, Hamdani SA, Bhat AH, Hussain SA, Shiekh I, et al. Epidemiological pattern of dog bites and the occurrence of rabies in humans within Srinagar district of Kashmir Valley, India. Comp Immunol Microbiol Infect Dis. 2020;73:101556. pmid:33035770
- 50. Wright N, Subedi D, Pantha S, Acharya KP, Nel LH. The Role of Waste Management in Control of Rabies: A Neglected Issue. Viruses. 2021;13(2):225. pmid:33535718
- 51. Tiwari HK, Bruce M, O’Dea M, Robertson ID. Utilising Group-Size and Home-Range Characteristics of Free-Roaming Dogs (FRD) to Guide Mass Vaccination Campaigns against Rabies in India. Vaccines (Basel). 2019;7(4):136. pmid:31575061
- 52. Castillo-Neyra R, Buttenheim AM, Brown J, Ferrara JF, Arevalo-Nieto C, Borrini-Mayorí K, et al. Behavioral and structural barriers to accessing human post-exposure prophylaxis and other preventive practices in Arequipa, Peru, during a canine rabies epidemic. PLoS Negl Trop Dis. 2020;14(7):e0008478. pmid:32692739
- 53. Hampson K, Cleaveland S, Briggs D. Evaluation of cost-effective strategies for rabies post-exposure vaccination in low-income countries. PLoS Negl Trop Dis. 2011;5(3):e982. pmid:21408121
- 54. Bashar M, Duggal M. Seasonal Variation in Incidence and Severity of Dog Bites in a Union Territory of Northern India. JCDR. 2019.
- 55. Yılmaz S, Delice O, İba Yılmaz S. Epidemiological characteristics, seasonality, trends of dog bite injuries, and relationship with meteorological data. Ann Agric Environ Med. 2023;30(2):229–34. pmid:37387371
- 56. Leblanc C, Kassié D, Ranaivoharimina M, Rakotomanana EFN, Mangahasimbola RT, Randrianarijaona A, et al. Mixed methods to evaluate knowledge, attitudes and practices (KAP) towards rabies in central and remote communities of Moramanga district, Madagascar. PLoS Negl Trop Dis. 2024;18(3):e0012064. pmid:38551968
- 57. N’Guessan RD, Heitz-Tokpa K, Amalaman DM, Tetchi SM, Kallo V, Ndjoug Ndour AP, et al. Determinants of Rabies Post-exposure Prophylaxis Drop-Out in the Region of San-Pedro, Côte d’Ivoire. Front Vet Sci. 2022;9:878886. pmid:35873685
- 58. Tiwari HK, Gogoi-Tiwari J, Robertson ID. Eliminating dog-mediated rabies: challenges and strategies. Animal Diseases. 2021;1(1).
- 59. Surendran J, Hs R, Kumari N, M Prasanth R, Fotedar N. A prospective study on health seeking behaviour and post exposure prophylaxis received by animal bite victims at the anti-rabies clinic in a tertiary care centre of urban Bangalore. F1000Res. 2024;13:175. pmid:39015143
- 60. Beasley EA, Wallace RM, Coetzer A, Nel LH, Pieracci EG. Roles of traditional medicine and traditional healers for rabies prevention and potential impacts on post-exposure prophylaxis: A literature review. PLoS Negl Trop Dis. 2022;16(1):e0010087. pmid:35051178
- 61. Bonaparte SC, Adams L, Bakamutumaho B, Barbosa Costa G, Cleaton JM, Gilbert AT, et al. Rabies post-exposure healthcare-seeking behaviors and perceptions: Results from a knowledge, attitudes, and practices survey, Uganda, 2013. PLoS One. 2021;16(6):e0251702. pmid:34077427
- 62. Kabeta T, Deresa B, Tigre W, Ward MP, Mor SM. Knowledge, Attitudes and Practices of Animal Bite Victims Attending an Anti-rabies Health Center in Jimma Town, Ethiopia. PLoS Negl Trop Dis. 2015;9(6):e0003867. pmid:26114573
- 63. Dutta A, Hiremath RN, Hasure SY. A KAP Study on Dog Bite and Its Management in a Rural Community — Need for Increasing the Awareness. J Med Sci Health. 2022;8(2):119–26.
- 64. Roy SK, Kar Chakraborty S, Mozumdar A. Health: Cognition and Threshold among the Oraon Tea Garden Labourers of Jalpaiguri District, West Bengal. J Anthropol. 2013;2013:1–9.
- 65. Rajput S, Hense S, Thankappan KR. Healthcare utilisation: a mixed-method study among tea garden workers in Indian context. J Healthe Res. 2021;36(6):1007–17.
- 66. Yamabhai J, Cusripituck P, Mingbualuang T, Sangkachai N, Sakchainanon W, Tungwongjulaniam C, et al. A persona-based exploration of rabies post-exposure prophylaxis seeking behavior and its implication for communication strategic planning: Evidence from Thailand. One Health. 2025;20:100980. pmid:39931353
- 67. Acharya KP, Acharya N, Phuyal S, Upadhyaya M, Lasee S. One-health approach: A best possible way to control rabies. One Health. 2020;10:100161. pmid:33117875
- 68. Zha R, Lu J, Chen J, Guo C, Lu J. Exploring one health-based strategies for rabies elimination: Overview and future prospects. PLoS Negl Trop Dis. 2025;19(6):e0013159.
- 69. Mohanty P, Boro PK, Heydtmann S, Durr S, Tiwari HK. Rabies in rural northeast India: A case report emphasising the urgency of the One Health approach. One Health. 2024;19:100850. pmid:39802066
- 70. Gborie SR, Issahaku GR, Bonful HA, Bandoh DA, Squire J, Ameme DK, et al. Analysis of dog bite surveillance data, Volta Region, Ghana, 2020. Front Trop Dis. 2023;4.
- 71. Chouhan CS, Raihan A, Mia MM, Banerjee S, Shahriar I, Nath P, et al. Demographic, temporal, and spatial analysis of human animal bite cases in Mymensingh District, Bangladesh. PLoS Negl Trop Dis. 2025;19(6):e0012204. pmid:40489552