Spatiotemporal Distribution of Human Rabies and Identification of Predominant Risk Factors in China from 2004 to 2020

Weiwei Meng; Tianren Shen; Okugbe Ebiotubo Ohore; Susan Christina Welburn; Guojing Yang

doi:10.1371/journal.pntd.0012557

Abstract

Human rabies is a prevalent issue in China, posing a significant public health concern in the country. This study fitted the Bayesian model of separable in spatial and temporal variation and inseparable spatiotemporal variation in disease risk respectively based on Integrated Nested Laplace Approximation (INLA) to investigate the spatiotemporal characteristics of human rabies across 31 provinces in China from 2004 to 2020. It also investigated the influence of natural and socio-environmental factors on the incidence of the disease. Within the study period, a total of 26,807 cases of human rabies were reported, with the highest risk of incidence occurring in 2007, followed by a steady annual decline to the lowest risk in 2020. Guangxi Province exhibited the highest risk, while Jilin Province had the lowest, with the southern, central, and eastern regions reporting higher risks than the northern and western areas. By 2020, most provinces such as Guangxi and Guizhou had significantly reduced their relative risk (RR) of human rabies from historical highs. However, some provinces like Hunan, Henan, and Jiangsu experienced an increase in RR compared to previous years. As the annual average temperature increases, the risk of human rabies incidence in China correspondingly rises. Conversely, with increases in the annual average daily sunshine duration, per capita disposable income of urban residents, and local government healthcare expenditures, the risk of human rabies incidence declines. We conclude that the risk of human rabies in China initially increased and then decreased annually from 2004 to 2020. Future efforts should continuously increase financial investments in rabies prevention and control, focusing particularly on Hunan, Henan, Jiangsu, and provinces characterized by higher temperatures, shorter sunshine durations, and lower economic levels.

Author summary

Rabies is a Neglected Tropical Disease (NTD), predominantly prevalent in Africa and Asia, with a nearly 100% fatality rate upon onset. China is striving to achieve the WHO’s global goal of zero deaths from dog-mediated human rabies by 2030. Therefore, exploring the spatial and temporal evolution of human rabies in China and the possible influencing factors can provide a scientific basis for the development of more effective human rabies prevention and control programs in China. The results of the study showed that from 2004 to 2020, the overall relative risk of human rabies incidence in China showed an increasing and then decreasing trend. The relative risk of most provinces, such as Guangxi and Guizhou, declined significantly, but there were still some provinces, such as Hunan and Henan, where the relative risk showed a rising trend year by year. The results of the analysis of influencing factors indicate that lower per capita disposable income of urban residents, local government healthcare expenditures and sunshine hours, higher temperatures are associated with a higher risk of human rabies.

Citation: Meng W, Shen T, Ohore OE, Welburn SC, Yang G (2024) Spatiotemporal Distribution of Human Rabies and Identification of Predominant Risk Factors in China from 2004 to 2020. PLoS Negl Trop Dis 18(10): e0012557. https://doi.org/10.1371/journal.pntd.0012557

Editor: Sergio Recuenco, Universidad Nacional Mayor de San Marcos, PERU

Received: January 6, 2024; Accepted: September 20, 2024; Published: October 31, 2024

Copyright: © 2024 Meng 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: The human rabies data that support the findings of this study are publicly available from the Data-Center of China Public Health Science and can be downloaded after registration. The identifier URL is: https://www.phsciencedata.cn/Share/en/data.jsp?id=915b15c5-48db-462e-8108-3ff2ba155b52&show=0#.

Funding: GY received funding from the National Natural Science Foundation of China (82260655), and Hainan Provincial Natural Science Foundation, China (821CXTD440). 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, commonly manifested as acute encephalitis, is a zoonotic disease caused by the rabies virus (RABV) genotype 1 and is also categorized as a neglected tropical disease (NTD) [1]. The disease is predominantly transmitted through bites or scratches from animals infected with rabies virus [2], with 99% of human cases resulting from dog bites. [3]. The incubation period typically ranges from one to three months, and once symptoms appear, the fatality rate is nearly 100% [4]. While rabies is present on all continents except Antarctica, more than 95% of deaths occur in Asia and Africa [3]. In China, over 95% of human rabies cases are due to dog bites [5].

Recognizing the ongoing prevalence of rabies in many countries, the World Health Organization (WHO), the World Organisation for Animal Health (WOAH), the Food and Agriculture Organization of the United Nations (FAO), and the Global Alliance for Rabies Control (GARC) formed the United Against Rabies (UAR) collaboration in 2015. This alliance is dedicated to eradicating human rabies transmitted by dogs and aims to achieve zero human rabies deaths by 2030 [6,7]. Historically, China’s incidence of human rabies during the early 21st century was second globally only to India, with three epidemic peaks recorded in the mid-1950s, mid-1980s, and again at the turn of the century [8]. Despite a decline in reported cases following the third peak, achieved through measures such as training health workers in post-exposure prophylaxis (PEP) and increasing access to PEP in rural areas, rabies remains a significant public health issue in China [9].

McIntyre et al. has shown that zoonotic pathogens are more sensitive to climate change than non-zoonotic pathogens [10]. Additionally, changes in land use can alter the transmission risk of infectious diseases by modifying the living conditions for hosts and vectors, thereby increasing the opportunities for contact between wildlife and humans [11]. Previous studies have indicated that higher temperatures and rainfall are associated with increased incidence rates of human rabies [12–15], whereas higher humidity levels tend to correlate with lower incidence rates [12]. This suggests that environmental factors significantly influence rabies outcomes. Infectious diseases are intricately linked to social determinants; factors such as inadequate sanitation, restricted access to healthcare, and insufficient health education significantly contribute to the prevalence of infectious diseases. Additionally, sociodemographic factors influence exposure to infectious diseases; for instance, human rabies is predominantly observed in children aged 5–14 years [3]. In the countries affected by NTDs, more than 70% are low or lower-middle-income nations [16]. NTDs, along with HIV/AIDS, tuberculosis, and malaria, are collectively referred to as infectious diseases of poverty (IDoP) [17]. Previous studies have indicated that rabies is a disease closely associated with social determinants. A study in Vietnam indicated that higher human population density and illiteracy rates are associated with an increased risk of human rabies [12]. Similarly, research from Yunnan Province in China found a positive correlation between human rabies incidence and population density, and a negative correlation with economic levels [15]. Additionally, a study from Curitiba in Brazil demonstrated that higher economic levels are associated with lower rates of dog bites [18].

In the present study, the Bayesian spatiotemporal model was developed to analyze the spatiotemporal patterns of human rabies incidence in China from 2004 to 2020. Compared to frequentist methods, Bayesian approaches utilize Bayes’ theorem to integrate prior distributions with likelihood functions, resulting in posterior distributions. This method effectively quantifies the uncertainties of estimates and offers more reliable outcomes when handling data from small samples [19,20]. The Bayesian spatiotemporal model leverages this approach to consider the relationships between spatial and temporal correlations, thereby enabling more precise identification of high-risk areas and facilitating the creation of disease risk maps [21]. It also enables the analysis of spatiotemporal trends and potential influencing factors. This provides a scientific basis for formulating effective strategies for the prevention and control of human rabies in China, aiming to achieve the goal of zero human rabies deaths by 2030. The study utilizes Integrated Nested Laplace Approximations (INLA) for the computation of the joint posterior distribution of model parameters, a method proposed by Rue et al. [22], which effectively addresses the time-consuming and slow convergence issues associated with Markov Chain Monte Carlo (MCMC) methods.

2. Data sources and methods

2.1 Variable selection

Human rabies is a disease linked with economic and social factors [12,23]. In China, human rabies predominantly affects economically disadvantaged rural areas where both cultural and economic levels are generally low, with farmers constituting the primary affected group. There is also a seasonal pattern, with higher incidence rates observed during the summer and autumn seasons [24]. Additionally, human rabies cases are more prevalent in the warm and humid southern regions of China, whereas the northeastern regions, characterized by lower population density and longer winters, exhibit lower incidence rates [25]. Consequently, this study incorporates socioeconomic indicators: per capita Gross Domestic Product (GDP) (in ten thousand yuan), per capita disposable income of urban residents (in ten thousand yuan), and per capita disposable income of rural residents (in ten thousand yuan), expenditure on healthcare by local governments (in 100 million yuan), population density (people/sq km), urbanization rate (%), density of healthcare institutions (per 10,000 people), and the percentage of the illiterate population among those aged 15 and over (%). Meteorological factors considered are: annual average temperature (°C), annual average relative humidity (%), annual sunshine duration (hours), and annual average precipitation (mm).

2.2 Data sources

The Internet-based Nationwide Notifiable Infectious Disease Reporting Information System was launched in 2004, hence this study spans from 2004 onwards. Data on human rabies cases from 2004 to 2020 across 31 provinces in China (excluding Taiwan, Hong Kong, and Macao Special Administrative Regions) were obtained from the Data-center of Public Health Science (https://www.phsciencedata.cn). Economic, social, and demographic data were sourced from the annual China Statistical Yearbook (https://www.stats.gov.cn/sj/ndsj/). The year-end population figures for each province were derived from the China Statistical Yearbooks spanning 2016 to 2021, with census data specifically for the years 2010 and 2020, and estimates from official population sampling surveys used for the intervening years. Provincial area data were obtained from the Ministry of Civil Affairs of the People’s Republic of China (https://www.mca.gov.cn/). Meteorological data were sourced from the China Meteorological Data Service Centre’s China’s surface climate data daily value data set (V3.0).

2.3 Data processing and model building

2.3.1 Data processing.

The per capita GDP of each province was calculated by dividing the provincial GDP by the end-of-year population count. Population density was determined by dividing the total end-of-year population by the provincial area. The density of healthcare institutions is calculated as the number of healthcare institutions at the end of the year divided by the population at year-end [26]. Meteorological data are initially averaged annually from daily records and then interpolated using the inverse distance weighting method. Subsequently, the data are amalgamated according to administrative divisions to produce province-specific datasets.

2.3.2 Variable screening.

Initially, we employed Spearman’s rank correlation analysis to identify variables associated with the incidence of human rabies in China, excluding those with a P-value > 0.05. Subsequently, a multicollinearity analysis was performed, and variables with a Variance Inflation Factor (VIF) ≥ 10 were excluded. The final variables included in the Bayesian spatiotemporal model were standardized by subtracting the mean and dividing by the standard deviation.

2.3.3 Spatial autocorrelation analysis.

For each year, we conducted a global spatial autocorrelation analysis of human rabies incidence rates across China to understand the spatial clustering patterns of the disease nationwide. This analysis utilized Moran’s I index as the statistic for assessing spatial autocorrelation, with significance determined through Z-value and P-value. Moran’s I value range from -1 to 1, with values > 0 indicating positive spatial autocorrelation (closer to 1 implies stronger autocorrelation), values < 0 indicating negative spatial autocorrelation (closer to -1 implies stronger negative autocorrelation), and values close to 0 or P-value > 0.05 indicating a random distribution of human rabies incidence [27].

2.3.4 Establishment and evaluation of the Bayesian spatiotemporal models.

It is generally accepted that human rabies case data follow a Poisson distribution [28]. The model was expressed as follows for

i = 1,..,31 for the 31 provinces and t = 1,..,17 for the years considered. Y_it represents the number of human rabies cases in the i^th province in year t, and E_it denotes the expected number of cases, which is the product of the overall incidence rate and the population of the province [29,30]. θ_it represents the relative risk (RR) of incidence in the i^th province in year t. Bayesian spatiotemporal model with separable spatial and temporal variation and spatio-temporal interaction models with inseparable spatiotemporal variation in disease risk are constructed [31]. These interaction models enhance the simple Bayesian spatiotemporal model framework by including spatio-temporal interaction effects, thus providing the model with the flexibility to accommodate temporal and spatial variations RR of human rabies. This adaptability ensures that the model captures the varying trends across different provinces and times more accurately [32]. The expressions for these models are shown in Table 1.

Download:

Table 1. Types of Bayesian spatiotemporal models.

https://doi.org/10.1371/journal.pntd.0012557.t001

Here, b₀ represents the intercept, u_i is the spatial structured effect reflecting spatial dependency, indicating that neighboring provinces have similar incidence rates and follows an intrinsic conditional autoregressive (iCAR) process, specified as ; υ_i is the spatial unstructured effect indicating spatial heterogeneity and follows a normal distribution, specified as ; γ_t is the temporal structured effect following a first-order autoregressive (AR1) process, specified as ,where |ρ|<1 is the autoregressive coefficient; ϕ_t is the temporal unstructured effect following a normal distribution, specified as ; δ_it represents the spatio-temporal interaction effect. Since the spatial and temporal effects are divided into structured and unstructured effects, the Bayesian spatio-temporal interaction model is further divided into four types [31]. Type I model is the interaction of temporal and spatial unstructured effects, so δ_it follows a normal distribution; Type II model is the interaction of temporal structured and spatial unstructured effects, so δ_it follows AR1⊗Normal; Type III model is the interaction of spatial structured and temporal unstructured effects, so δ_it follows iCAR⊗Normal; Type IV model is the interaction of temporal and spatial structured effects, so δ_it follows iCAR⊗AR1. For the Bayesian spatiotemporal model, where disease risk is considered to have separable spatial and temporal variations, the spatial effect, denoted as , and the temporal effect, denoted as ; in models where disease risk features inseparable spatiotemporal interaction, the spatiotemporal interaction effect is denoted as . We assign each random effect the penalized complexity priors proposed by Simpson et al[33], with prior parameter values using the default settings in R-INLA.

In assessing the fit of spatiotemporal interaction models within the Bayesian framework, two commonly used metrics are employed: the Deviance Information Criterion (DIC), the Watanabe-Akaike Information Criterion (WAIC)[34]. Lower values of these criteria indicate a better model fit, taking into account the complexity of the model.

2.3.5 Cold and hot spot analysis.

Based on the results of the Bayesian spatiotemporal model, hot spot, cold spot and non-cold and non-hot regions can be delineated according to the posterior probability of the spatial effects [35]. Hot spots are defined as provinces where the posterior probability . Cold spots are defined as provinces where the posterior probability . Regions that are neither hot spots nor cold spots are those where the posterior probability .

2.3.6 Ecological regression.

Following variable selection, covariates were incorporated into spatiotemporal and spatio-temporal interaction models. The model with the lowest DIC and WAIC was selected to explore the determinants of human rabies incidence in China. The expressions for these models are as follows:

Here, β_z represents the coefficient for covariate z, and X_zit represents the value of the z^th influencing factor in province i in year t. The RR of the influencing factors is given by exp (β) [36].

2.3.7 Statistical software.

Spearman correlation analyses were conducted using the ‘psych’ package in R software version 4.3.0. Multicollinearity analysis were performed using the ‘car’ package. Calculations of the expected number of disease cases were carried out with the ‘SpatialEpi’ package[29, 37]. The Bayesian spatiotemporal models were constructed using the ‘INLA’ package (version: 23.11.26). Data visualization and global spatial autocorrelation analyses were performed using ArcGIS 10.8 software. The significance level was set at α = 0.05.

3. Results

3.1 Epidemiological characteristics of human rabies in China’s provinces

From 2004 to 2020, China’s 31 provinces reported a total of 26,807 human rabies cases, with 26,165 fatalities. The average annual incidence rate was 0.119 per 100,000 population, and the average annual mortality rate was 0.116 per 100,000. The highest number of reported cases and deaths occurred in the years 2006–2007, with a subsequent yearly decline. The annual average incidence rates by province are shown in Fig 1, with the highest rates observed in Guangxi Province (0.52 per 100,000) and Guizhou Province (0.51 per 100,000). Guangxi, Guizhou, Hainan, and Hunan provinces had notably higher average annual incidence rates compared to other provinces. See Fig 1 for details.

Download:

Fig 1. Annual average incidence rates of human rabies in provinces from 2004–2020.

This map was created using ArcGIS 10.8 software. The base layer of the map is from the National Catalogue Service For Geographic Information of the Ministry of Natural Resources of the People’s Republic of China(https://www.webmap.cn/mapDataAction.do?method=forw&resType=5&storeId=2&storeName=%E5%9B%BD%E5%AE%B6%E5%9F%BA%E7%A1%80%E5%9C%B0%E7%90%86%E4%BF%A1%E6%81%AF%E4%B8%AD%E5%BF%83&fileId=BA420C422A254198BAA5ABAB9CAAFBC1).

https://doi.org/10.1371/journal.pntd.0012557.g001

3.2 Spatial autocorrelation analysis of human rabies

A global spatial autocorrelation analysis of human rabies incidence rates was conducted. The results indicated positive spatial autocorrelation in the incidence of human rabies in China’s 31 provinces from 2004 to 2019, manifesting as "high-high" and "low-low" clustering patterns (Moran’s I > 0 and P < 0.05). In 2020, the incidence of human rabies was found to be randomly distributed in space (Moran’s I = 0.102, P = 0.117). See Table 2.

Download:

Table 2. Global spatial autocorrelation analysis results of human rabies incidence rates in China’s 31 provinces from 2004–2020.

https://doi.org/10.1371/journal.pntd.0012557.t002

3.3 Results of the Bayesian spatiotemporal model

3.3.1 Analysis of spatial effects.

The RR for spatial effects was highest in Guangxi Province (13.46) and lowest in Jilin Province (0.026). The hot spot and cold spot analysis revealed 17 hot spot areas (54.84%), predominantly in the southern, central and eastern provinces, and 11 cold spot areas (35.48%), mainly in the northern and western provinces. The 3 areas (9.68%) were neither hot spots nor cold spots, namely Ningxia Province, Shanxi Province, and Tianjin. See Fig 2 for details.

Download:

Fig 2. RR of the spatial effect and distribution of hot-cold spots.

This figure illustrates the posterior means of spatial relative risks obtained from spatial random effects and the classification into hot and cold spots based on these posterior means. The size of the spheres indicates the magnitude of the spatial relative risks, with larger spheres denoting higher risks. This map was created using ArcGIS 10.8 software. The base layer of the map is from the National Catalogue Service For Geographic Information of the Ministry of Natural Resources of the People’s Republic of China(https://www.webmap.cn/mapDataAction.do?method=forw&resType=5&storeId=2&storeName=%E5%9B%BD%E5%AE%B6%E5%9F%BA%E7%A1%80%E5%9C%B0%E7%90%86%E4%BF%A1%E6%81%AF%E4%B8%AD%E5%BF%83&fileId=BA420C422A254198BAA5ABAB9CAAFBC1).

https://doi.org/10.1371/journal.pntd.0012557.g002

3.3.2 Analysis of temporal effects.

The RR for temporal effects showed an initial increase followed by a decline from 2004 to 2020. The RR increased yearly from 2004 to 2007, reaching their peak in 2007. The period of 2006–2007 showed overall higher RR. Starting from 2008, the RR began to decrease annually, reaching their lowest level in 2020. See Fig 3 for details.

Download:

Fig 3. RR of the temporal effect for human rabies in China from 2004–2020.

This figure illustrates the posterior means of the temporal relative risks for human rabies derived from temporal random effects from 2004 to 2020, presented as black lines. The 95% credible intervals (CI) are shown as blue areas.

https://doi.org/10.1371/journal.pntd.0012557.g003

3.4 Results of the Bayesian spatio-temporal interaction model

The results show that the fitting effect of spatio-temporal interaction model II and IV is better than the other two, and the WAIC and DIC of model II are the lowest, so model II is finally selected for analysis. See Table 3 for the evaluation indexes of fitting performance of spatio-temporal interaction model.

Download:

Table 3. Evaluation of the fitting performance of spatio-temporal interaction models.

https://doi.org/10.1371/journal.pntd.0012557.t003

The spatio-temporal interaction effects of RR are shown in Fig 4. The results show that from 2004 to 2007, the high-risk areas were concentrated in the southern, central, and eastern provinces, with notably higher RR in Guangxi, Guizhou, and Hunan provinces compared to other provinces. Starting from 2009, the epidemic area expanded gradually, but the overall RR showed a year-by-year declining trend. In 2020, only 2 provinces had an RR greater than 5, an 80% decrease from 2005 (10 provinces). These two provinces were Hunan (16.65) and Henan (5.31). The RR of human rabies in Hunan Province has increased yearly since 2015 and has become the province with the highest RR in China from 2017. The RR in Henan Province has shown a fluctuating upward trend since 2011. Besides, the RR in Sichuan and Jiangsu provinces has increased year by year since 2017 and 2018, respectively. The RR of Guangxi Province decreased by 90.73% from its historical high, while that of Guizhou Province decreased by 90.41%. Additionally, significant declines were also observed in Chongqing, Hainan, and Guangdong provinces.

Download:

Fig 4. RR of the spatio-temporal interaction effect for Human Rabies in China from 2004–2020.

This figure illustrates the posterior means of the spatio-temporal RR for human rabies, derived from the spatio-temporal interaction effects covering the period from 2004 to 2020. This map was created using ArcGIS 10.8 software. The base layer of the map is from the National Catalogue Service For Geographic Information of the Ministry of Natural Resources of the People’s Republic of China(https://www.webmap.cn/mapDataAction.do?method=forw&resType=5&storeId=2&storeName=%E5%9B%BD%E5%AE%B6%E5%9F%BA%E7%A1%80%E5%9C%B0%E7%90%86%E4%BF%A1%E6%81%AF%E4%B8%AD%E5%BF%83&fileId=BA420C422A254198BAA5ABAB9CAAFBC1).

https://doi.org/10.1371/journal.pntd.0012557.g004

3.5 Ecological regression

The collinearity analysis revealed that the per capita GDP, per capita disposable income of rural residents and annual average humidity exhibited VIF greater than 10, indicating significant multicollinearity. Consequently, these variables were excluded from further analysis. The final variables incorporated in model for analysis included: per capita disposable income of urban residents, expenditure on healthcare by local governments, percentage of illiterate population to total aged 15 and over, density of healthcare institutions, urbanization rate, population density, annual sunshine duration, annual average temperature, and annual average precipitation.

The results indicate that, after incorporating covariates, the Type II spatio-temporal interaction model exhibited the lowest DIC and WAIC, suggesting optimal model fit. Consequently, this model was selected to investigate the influencing factors of human rabies incidence in China. See Table 4 for details.

Download:

Table 4. Ecological regression model fitting results.

https://doi.org/10.1371/journal.pntd.0012557.t004

The results of the influencing factor analysis indicated a positive correlation between the risk of human rabies and annual average temperature (RR = 3.381, 95% CI: 2.275–4.838), implying a higher risk with increasing temperature. A negative correlation was observed with per capita disposable income of urban residents (RR = 0.625, 95% CI: 0.410–0.917), expenditure on healthcare by local governments (RR = 0.763, 95% CI: 0.572–0.996) and annual sunshine duration (RR = 0.668, 95% CI: 0.533–0.826), indicating that higher income levels, greater local governments healthcare spending, and longer sunshine duration are associated with lower risks. Other variables did not show a statistically significant correlation with the risk of human rabies (95% CI included 1). See Table 5 for details.

Download:

Table 5. RR and 95% Credible Intervals of influencing factors in Type II spatio-temporal interaction model.

https://doi.org/10.1371/journal.pntd.0012557.t005

4. Discussion

This study investigated the spatiotemporal distribution and potential influencing factors of human rabies in China from 2004 to 2020. The results indicate that from 2004 to 2019, cases of human rabies in China exhibited spatial clustering, with higher risks observed in the southern, central, and eastern provinces compared to the northern and western provinces. In 2020, the spatial distribution of human rabies cases in China appeared random, which may be attributed to reduced human mobility due to the COVID-19 pandemic, subsequently diminishing exposure risks. From 2004 to 2020, the RR of human rabies in China initially increased, followed by a decline, with most provinces reaching a lower level of RR by 2020. Notably, the many provinces such as Guizhou, Guangxi, and Hainan experienced significant declines from their peak years. China’s human rabies control measures have been effective.

In 2020, the provinces with a RR greater than 5 included Hunan and Henan provinces. Hunan has been the province with the highest RR since 2017, showing a year-over-year increase. Additionally, the RR in Henan Province has exhibited fluctuating upward trends since 2011. Sichuan and Jiangsu provinces have seen a rebound in RR in recent years. Lower coverage of standardized post-exposure prophylaxis and canine immunization, and lack of public awareness of rabies are potential reasons for the continued high risk of rabies in humans in these provinces [38–40].

The results regarding influencing factors show that China’s RR of human rabies is positively correlated with annual average temperature, consistent with previous studies [14,41] and aligning with the higher incidence of rabies during the summer and autumn seasons in China [42,43]. In warmer climates, people tend to wear lighter clothing and increase outdoor activities. Additionally, dogs are more irritable [13] and have a wider range of activities in hot weather [14], increasing the risk of bites. The study also found a negative correlation between the RR of human rabies and annual sunshine duration, which aligns with the higher average annual sunshine duration in northern provinces compared to the southern provinces [44], where rabies risk is higher. This association between lower risk and longer sunshine duration may reflect changes in human and vector behavior rather than the direct effects of UV light on the rabies virus. Although the rabies virus is sensitive to UV light [45], transmission primarily occurs through direct contact between humans and vectors. Sunshine duration may influence the risk of human rabies by altering the behaviors of humans and vectors, thereby impacting the likelihood of exposure.

There is a negative correlation between the RR of human rabies and the per capita disposable income of urban residents in China, consistent with previous findings [13], Human rabies is closely associated with economic levels, predominantly affecting impoverished populations [3]. As economic standards improve, there is typically an enhanced focus on health and an increased awareness of disease prevention, which may lead to reduced exposure risks to human rabies. Additionally, in this study, local government healthcare expenditures were found to be negatively correlated with the RR of human rabies in China. With the economic growth, there has been a consistent increase in government spending on healthcare, which has supported various initiatives. These include health education campaigns targeting rabies [46], inclusion of outpatient costs for dog bite victims under medical insurance [43], and provision of free rabies vaccinations for dogs and cats [47]. Additionally, human rabies vaccines, human rabies immunoglobulin, and anti-rabies serum have been included in China’s national medical insurance drug list [48], effectively promoting rabies control. A study showed that over the past 20 years, post-exposure vaccination rates and antibody injection rates have significantly increased in China, along with rising canine immunization coverage [49]. However, as of 2020, the average canine immunization coverage rate at Chinese monitoring sites was only 30% [50], while the WHO considers a coverage rate of 70% sufficient to eliminate human rabies [51]. Vaccinating dogs is considered the most cost-effective method for eliminating human rabies [52], and this method has contributed significantly to the effective control of human rabies in Guangxi Province [53].

In China, human rabies predominantly occurs in rural areas [50], but this study found no significant correlation between urbanization rate and RR of human rabies, aligning with the findings of Hangyu Li et al. [41]. Some provinces with lower urbanization rates, like Xizang and Xinjiang, are located in high-altitude areas characterized by low temperatures and longer sunshine duration, reducing dog activity and human exposure risk. Studies on the influencing factors of human rabies in Yunnan Province [15] and Vietnam [12] by Jing Yu and Dung Phung, respectively, identified human population density and annual average precipitation as risk factors, and annual average relative humidity as a protective factor. However, in this study, these factors did not show a significant correlation with the RR of human rabies, possibly because areas with high population density, such as Beijing, Shanghai, and Tianjin, are not hotspots for human rabies in China. Additionally, the spatial scale of this study was at the provincial level, and correlations may exist at smaller spatial scales.

This study has limitations. The analysis was conducted at the provincial-level spatial scale and annual temporal scale, which may introduce biases in identifying influencing factors; moreover, how each influencing factor affects the risk of human rabies requires deeper investigation. The human rabies case data were sourced from the Data-center of Public Health Science and lacked detailed individual case investigations, which could lead to biases and underreporting issues. The absence of comprehensive monitoring of canine immunization coverage and dog populations in China restricted the inclusion of these variables, limiting further exploration of potential factors influencing the incidence of human rabies. These limitations highlight areas for future research. Provinces in China should systematically monitor canine coverage and dog populations, and manage this data through a dedicated database, to better assess risks and devise more effective prevention strategies.

In summary, this study employs Bayesian spatiotemporal models to analyze the epidemiological characteristics of rabies in China from 2004 to 2020, exploring potential influencing factors. The results reveal spatial clustering of human rabies cases, with higher risks in the southern, central, and eastern provinces than in the northern and western provinces, and initially increased and then decreased in RR over the years. The annual average temperature was identified as a risk factor for human rabies in China, while per capita disposable income of urban residents, expenditure on healthcare by local governments and annual sunshine duration were protective factors.

Moving forward, efforts should concentrate on human rabies prevention and control in Hunan, Henan, Jiangsu provinces and other areas with high temperatures, low sunshine hours and low economic levels. It is imperative to increase financial investment in rabies control, enhance standardized treatment rates for exposed populations, strengthen surveillance, and improve canine immunization coverage to effectively manage and mitigate the risk of rabies in China.

5. Conclusion

This study reveals the spatial and temporal dynamics of human rabies in China from 2004 to 2020, while examining its association with socio-economic and environmental factors. Over the observed period, the risk of human rabies in China initially increased and then showed a consistent decline. Regions with higher risk were mainly located in the southern, central and eastern provinces. While most provinces in China have achieved commendable control over human rabies, there are still provinces such as Hunan and Henan where rabies cases are on the rise, highlighting them as critical areas for enhanced prevention and control measures. The analysis identifies annual average temperature as a risk factor, annual sunshine duration, per capita disposable income of urban residents and local government expenditure on healthcare as protective factors for human rabies in China. These findings suggest that increasing financial commitments to rabies prevention and focusing on provinces with high temperatures, short sunshine durations, and lower economic levels will be pivotal for future control strategies in China.

Disclaimer

The authors alone are responsible for the views expressed in this article and they do not necessarily represent the views, decisions, or policies of the institutions with which they are affiliated.

Supporting information

S1 Fig. Correlation analysis.

https://doi.org/10.1371/journal.pntd.0012557.s001

(TIF)

S1 Table. Multicollinearity analysis.

https://doi.org/10.1371/journal.pntd.0012557.s002

(DOCX)

References

1. WHO. Ending the neglect to attain the sustainable development goals: a road map for neglected tropical diseases 2021–2030. Geneva: World Health Organization; 2020 2020.
2. Hemachudha T, Ugolini G, Wacharapluesadee S, Sungkarat W, Shuangshoti S, Laothamatas J. Human rabies: neuropathogenesis, diagnosis, and management. Lancet Neurol. 2013;12(5):498–513. Epub 2013/04/23. pmid:23602163.
- View Article
- PubMed/NCBI
- Google Scholar
3. WHO. Rabies 2023-1-19 [2023 Sep 22]. Available from: https://www.who.int/news-room/fact-sheets/detail/rabies.
4. Jackson AC. Rabies virus infection: an update. J Neurovirol. 2003;9(2):253–8. Epub 2003/04/23. pmid:12707856.
- View Article
- PubMed/NCBI
- Google Scholar
5. Miao F, Li N, Yang J, Chen T, Liu Y, Zhang S, et al. Neglected challenges in the control of animal rabies in China. One Health. 2021;12:100212. Epub 2021/02/09. pmid:33553562; PubMed Central PMCID: PMC7843516.
- View Article
- PubMed/NCBI
- Google Scholar
6. WHO, FAO, WOAH. Zero by 30: the global strategic plan to end human deaths from dog-mediated rabies by 2030. Geneva: World Health Organization; 2018.
7. WHO, FAO, WOAH. Global elimination of dog-mediated human rabies: report of the rabies global conference, 10–11 December 2015, Geneva, Switzerland. Geneva: World Health Organization, 2016 2016. Report No.: Contract No.: WHO/HTM/NTD/NZD/2016.02.
8. China C. Technical guidelines for human rabies prevention and control (2016). Chin J Viral Dis. 2016;6(03):161–88(in Chinese).
- View Article
- Google Scholar
9. Yin W, Dong J, Tu C, Edwards J, Guo F, Zhou H, et al. Challenges and needs for China to eliminate rabies. Infect Dis Poverty. 2013;2(1):23. Epub 2013/10/04. pmid:24088366; PubMed Central PMCID: PMC3851465.
- View Article
- PubMed/NCBI
- Google Scholar
10. McIntyre KM, Setzkorn C, Hepworth PJ, Morand S, Morse AP, Baylis M. Systematic Assessment of the Climate Sensitivity of Important Human and Domestic Animals Pathogens in Europe. Sci Rep. 2017;7(1):7134. Epub 2017/08/05. pmid:28769039; PubMed Central PMCID: PMC5541049.
- View Article
- PubMed/NCBI
- Google Scholar
11. Gottdenker NL, Streicker DG, Faust CL, Carroll CR. Anthropogenic land use change and infectious diseases: a review of the evidence. Ecohealth. 2014;11(4):619–32. Epub 2014/05/24. pmid:24854248.
- View Article
- PubMed/NCBI
- Google Scholar
12. Phung D, Nguyen HX, Nguyen HLT, Luong AM, Do CM, Tran QD, et al. The effects of socioecological factors on variation of communicable diseases: A multiple-disease study at the national scale of Vietnam. PLoS One. 2018;13(3):e0193246. Epub 2018/03/02. pmid:29494623; PubMed Central PMCID: PMC5832231.
- View Article
- PubMed/NCBI
- Google Scholar
13. Yao HW, Yang Y, Liu K, Li XL, Zuo SQ, Sun RX, et al. The spatiotemporal expansion of human rabies and its probable explanation in mainland China, 2004–2013. PLoS Negl Trop Dis. 2015;9(2):e0003502. Epub 2015/02/19. pmid:25692883; PubMed Central PMCID: PMC4334667.
- View Article
- PubMed/NCBI
- Google Scholar
14. Guo D, Yin W, Yu H, Thill JC, Yang W, Chen F, et al. The role of socioeconomic and climatic factors in the spatio-temporal variation of human rabies in China. BMC Infect Dis. 2018;18(1):526. Epub 2018/10/24. pmid:30348094; PubMed Central PMCID: PMC6198482.
- View Article
- PubMed/NCBI
- Google Scholar
15. Yu J, Xiao H, Yang W, Dellicour S, Kraemer MUG, Liu Y, et al. The impact of anthropogenic and environmental factors on human rabies cases in China. Transbound Emerg Dis. 2020;67(6):2544–53. Epub 2020/04/30. pmid:32348020.
- View Article
- PubMed/NCBI
- Google Scholar
16. Manderson L, Aagaard-Hansen J, Allotey P, Gyapong M, Sommerfeld J. Social research on neglected diseases of poverty: continuing and emerging themes. PLoS Negl Trop Dis. 2009;3(2):e332. Epub 2009/02/25. pmid:19238216; PubMed Central PMCID: PMC2643480.
- View Article
- PubMed/NCBI
- Google Scholar
17. Bhutta ZA, Sommerfeld J, Lassi ZS, Salam RA, Das JK. Global burden, distribution, and interventions for infectious diseases of poverty. Infect Dis Poverty. 2014;3:21. Epub 2014/08/12. pmid:25110585; PubMed Central PMCID: PMC4126350.
- View Article
- PubMed/NCBI
- Google Scholar
18. Constantino C, Da Silva EC, Dos Santos DM, Paploski IAD, Lopes MO, Morikawa VM, et al. One Health Approach on Dog Bites: Demographic and Associated Socioeconomic Factors in Southern Brazil. Trop Med Infect Dis. 2023;8(4). Epub 2023/04/27. pmid:37104315; PubMed Central PMCID: PMC10146334.
- View Article
- PubMed/NCBI
- Google Scholar
19. Shoari N, Heydari S, Blangiardo M. A decade of child pedestrian safety in England: a bayesian spatio-temporal analysis. BMC Public Health. 2023;23(1):215. Epub 2023/02/02. pmid:36721178; PubMed Central PMCID: PMC9889245.
- View Article
- PubMed/NCBI
- Google Scholar
20. van de Schoot R, Kaplan D, Denissen J, Asendorpf JB, Neyer FJ, van Aken MAG. A gentle introduction to bayesian analysis: applications to developmental research. Child Dev. 2014;85(3):842–60. Epub 2013/10/15. pmid:24116396; PubMed Central PMCID: PMC4158865.
- View Article
- PubMed/NCBI
- Google Scholar
21. Ugarte MD, Adin A, Goicoa T, Militino AF. On fitting spatio-temporal disease mapping models using approximate Bayesian inference. (1477–0334 (Electronic)).
22. Rue H, Martino S, Chopin N. Approximate Bayesian inference for latent Gaussian models by using integrated nested Laplace approximations. Journal of the Royal Statistical Society Series B (Statistical Methodology). 2009;71(2):319–92.
- View Article
- Google Scholar
23. Taylor E, George K, Johnson E, Whitelegg H, Prada JM, Horton DL. Quantifying the interconnectedness between poverty, health access, and rabies mortality. PLoS Negl Trop Dis. 2023;17(4):e0011204. Epub 2023/04/20. pmid:37079553; PubMed Central PMCID: PMC10118163.
- View Article
- PubMed/NCBI
- Google Scholar
24. Zhou H, Vong S, Liu K, Li Y, Mu D, Wang L, et al. Human Rabies in China, 1960–2014: A Descriptive Epidemiological Study. PLoS Negl Trop Dis. 2016;10(8):e0004874. Epub 2016/08/09. pmid:27500957; PubMed Central PMCID: PMC4976867.
- View Article
- PubMed/NCBI
- Google Scholar
25. Wu H, Song M, Shen XX, Tang Q, Wang SM. Epidemiology of rabies in China from 1996 to 2009. Disease Surveillance. 2011;26(06):427–30+34(in Chinese).
- View Article
- Google Scholar
26. WHO. Density of hospitals (per 100 000 population) [2024 Apr 3]. Available from: https://www.who.int/data/gho/indicator-metadata-registry/imr-details/3361.
27. Moran PA. Notes on continuous stochastic phenomena. 1950;(0006–3444 (Print)). D—CLML: 5019:10193:26:230 OTO—NLM.
28. Zhou XN. Spatial Epidemiology: China Science Publishing & Media Ltd; 2009(in Chinese).
29. Moraga P. Geospatial Health Data: Modeling and Visualization with R-INLA and Shiny: Chapman & Hall/CRC Biostatistics Series; 2019.
30. Wang Y, Lai Y, Du Z, Zhang W, Feng C, Li R, et al. Spatiotemporal Distribution of Hand, Foot, and Mouth Disease in Guangdong Province, China and Potential Predictors, 2009⁻2012. Int J Environ Res Public Health. 2019;16(7). Epub 2019/04/17. pmid:30987085; PubMed Central PMCID: PMC6480297.
- View Article
- PubMed/NCBI
- Google Scholar
31. Knorr-Held L. Bayesian modelling of inseparable space-time variation in disease risk. Stat Med. 2000;19(17–18):2555–67. Epub 2000/08/29. pmid:10960871.
- View Article
- PubMed/NCBI
- Google Scholar
32. Vicente G, Goicoa T, Ugarte MD. Bayesian inference in multivariate spatio-temporal areal models using INLA: analysis of gender-based violence in small areas. Stochastic Environmental Research and Risk Assessment. 2020;34(10):1421–40.
- View Article
- Google Scholar
33. Daniel S, Håvard R, Andrea R, Thiago GM, Sigrunn HS. Penalising Model Component Complexity: A Principled, Practical Approach to Constructing Priors. Statistical Science. 2017;32(1):1–28.
- View Article
- Google Scholar
34. Gomez-Rubio V. Bayesian inference with INLA: Chapman and Hall/CRC; 2020.
35. Richardson S, Thomson A, Best N, Elliott P. Interpreting posterior relative risk estimates in disease-mapping studies. Environ Health Perspect. 2004;112(9):1016–25. Epub 2004/06/17. pmid:15198922; PubMed Central PMCID: PMC1247195.
- View Article
- PubMed/NCBI
- Google Scholar
36. Marta B, Michela C. Spatial modeling. Spatial and Spatio-temporal Bayesian Models with R-INLA2015. p. 173–234.
- View Article
- Google Scholar
37. Naing NN. Easy way to learn standardization: direct and indirect methods. Malays J Med Sci. 2000;7(1):10–5. Epub 2000/01/01. pmid:22844209; PubMed Central PMCID: PMC3406211.
- View Article
- PubMed/NCBI
- Google Scholar
38. Sun J W., Huang X Y., Su J, Li X, L., Xu B, L. Epidemiology and control strategy on human rabies in Henan Province, China. Chinese Journal of Zoonoses. 2016;32(10):939–43(in Chinese).
- View Article
- Google Scholar
39. Yang H, Chen S, Dai Z, Luo L, Sun Q, Zhao S, et al. Integrated Rabies Surveillance—Hunan Province, China, 2020. China CDC Wkly. 2022;4(44):986–9. Epub 2022/12/10. pmid:36483987; PubMed Central PMCID: PMC9709299.
- View Article
- PubMed/NCBI
- Google Scholar
40. Wang XC, Zhang N, Zhou WZ, Qin YF, Li ZF, Hu JL, et al. Disease burden of rabies in Jiangsu province, 2008–2018: a surveillance data-based analysis. Chin J Public Health. 2021;37(01):134–7(in Chinese).
- View Article
- Google Scholar
41. Li H, Li Y, Chen Y, Chen B, Su Q, Hu Y, et al. Mapping rabies distribution in China: a geospatial analysis of national surveillance data. Int J Infect Dis. 2023;131:140–6. Epub 2023/04/09. pmid:37030651.
- View Article
- PubMed/NCBI
- Google Scholar
42. Yue Y, Chen Q, Mu D, Li Y, Yin W. A Descriptive Analysis of Human Rabies in Mainland China, 2005–2020. Int J Environ Res Public Health. 2022;20(1). Epub 2023/01/09. pmid:36612701; PubMed Central PMCID: PMC9819004.
- View Article
- PubMed/NCBI
- Google Scholar
43. Zhang N, Song C, Tao X, Zhu W. Epidemiologic Features of Human Rabies in China from 2015–2021. Zoonoses. 2023;3(1)
- View Article
- Google Scholar
44. Xiao FJ, Zhang XG, Liao YM, LIU QF. Sunshine Duration in China: Variation Characteristics and Its Influence. Chinese Agricultural Science Bulletin. 2020;36(20):92–100(in Chinese).
- View Article
- Google Scholar
45. WOAH. Rabies 2009 [2023 Sep 23]. Available from: https://www.woah.org/app/uploads/2021/03/rabies-1.pdf.
46. Qiao YC, Wang F, He YL, Yang Q, Yang J, Wei YS. Regional and age difference of human rabies prevalence of the past fourteen years in China. (1873–1716 (Electronic)).
- View Article
- Google Scholar
47. Wang XJ, Qiu DR, Chang YH. Analysis on epidemic characteristics of rabies in Luzhou City 2014 to 2020. J Prev Med Inf. 2022;38(07):942–6(in Chinese).
- View Article
- Google Scholar
48. The People’s Republic of China basic medical insurance drug catalog 2023 [2023 Sep 22]. Available from: https://www.gov.cn/zhengce/zhengceku/2023-01/18/content_5737840.htm.
49. Shen T, Welburn SC, Sun L, Yang GJ. Progress towards dog-mediated rabies elimination in PR China: a scoping review. Infect Dis Poverty. 2023;12(1):30. Epub 2023/04/07. pmid:37024944; PubMed Central PMCID: PMC10077633.
- View Article
- PubMed/NCBI
- Google Scholar
50. Liu Z, Liu M, Tao X, Zhu W. Epidemic Characteristics of Human Rabies—China, 2016–2020. China CDC Wkly. 2021;3(39):819–21. Epub 2021/10/02. pmid:34594999; PubMed Central PMCID: PMC8477051.
- View Article
- PubMed/NCBI
- Google Scholar
51. Coleman PG, Dye C. Immunization coverage required to prevent outbreaks of dog rabies. Vaccine. 1996;14(3):185–6. Epub 1996/02/01. pmid:8920697.
- View Article
- PubMed/NCBI
- Google Scholar
52. Fitzpatrick MC, Hampson K, Cleaveland S, Mzimbiri I, Lankester F, Lembo T, et al. Cost-effectiveness of canine vaccination to prevent human rabies in rural Tanzania. Ann Intern Med. 2014;160(2):91–100. Epub 2014/03/05. pmid:24592494; PubMed Central PMCID: PMC4084874.
- View Article
- PubMed/NCBI
- Google Scholar
53. Wei XK, Xiong Y, Li XN, Zheng M, Pan Y, He XX, et al. Vaccination demonstration zone successfully controls rabies in Guangxi Province, China. (1471–2334 (Electronic)).
- View Article
- Google Scholar

[ref1] 1. WHO. Ending the neglect to attain the sustainable development goals: a road map for neglected tropical diseases 2021–2030. Geneva: World Health Organization; 2020 2020.

[ref2] 2. Hemachudha T, Ugolini G, Wacharapluesadee S, Sungkarat W, Shuangshoti S, Laothamatas J. Human rabies: neuropathogenesis, diagnosis, and management. Lancet Neurol. 2013;12(5):498–513. Epub 2013/04/23. pmid:23602163.
View Article
PubMed/NCBI
Google Scholar

[3] View Article

[4] PubMed/NCBI

[5] Google Scholar

[ref3] 3. WHO. Rabies 2023-1-19 [2023 Sep 22]. Available from: https://www.who.int/news-room/fact-sheets/detail/rabies.

[ref4] 4. Jackson AC. Rabies virus infection: an update. J Neurovirol. 2003;9(2):253–8. Epub 2003/04/23. pmid:12707856.
View Article
PubMed/NCBI
Google Scholar

[8] View Article

[9] PubMed/NCBI

[10] Google Scholar

[ref5] 5. Miao F, Li N, Yang J, Chen T, Liu Y, Zhang S, et al. Neglected challenges in the control of animal rabies in China. One Health. 2021;12:100212. Epub 2021/02/09. pmid:33553562; PubMed Central PMCID: PMC7843516.
View Article
PubMed/NCBI
Google Scholar

[12] View Article

[13] PubMed/NCBI

[14] Google Scholar

[ref6] 6. WHO, FAO, WOAH. Zero by 30: the global strategic plan to end human deaths from dog-mediated rabies by 2030. Geneva: World Health Organization; 2018.

[ref7] 7. WHO, FAO, WOAH. Global elimination of dog-mediated human rabies: report of the rabies global conference, 10–11 December 2015, Geneva, Switzerland. Geneva: World Health Organization, 2016 2016. Report No.: Contract No.: WHO/HTM/NTD/NZD/2016.02.

[ref8] 8. China C. Technical guidelines for human rabies prevention and control (2016). Chin J Viral Dis. 2016;6(03):161–88(in Chinese).
View Article
Google Scholar

[18] View Article

[19] Google Scholar

[ref9] 9. Yin W, Dong J, Tu C, Edwards J, Guo F, Zhou H, et al. Challenges and needs for China to eliminate rabies. Infect Dis Poverty. 2013;2(1):23. Epub 2013/10/04. pmid:24088366; PubMed Central PMCID: PMC3851465.
View Article
PubMed/NCBI
Google Scholar

[21] View Article

[22] PubMed/NCBI

[23] Google Scholar

[ref10] 10. McIntyre KM, Setzkorn C, Hepworth PJ, Morand S, Morse AP, Baylis M. Systematic Assessment of the Climate Sensitivity of Important Human and Domestic Animals Pathogens in Europe. Sci Rep. 2017;7(1):7134. Epub 2017/08/05. pmid:28769039; PubMed Central PMCID: PMC5541049.
View Article
PubMed/NCBI
Google Scholar

[25] View Article

[26] PubMed/NCBI

[27] Google Scholar

[ref11] 11. Gottdenker NL, Streicker DG, Faust CL, Carroll CR. Anthropogenic land use change and infectious diseases: a review of the evidence. Ecohealth. 2014;11(4):619–32. Epub 2014/05/24. pmid:24854248.
View Article
PubMed/NCBI
Google Scholar

[29] View Article

[30] PubMed/NCBI

[31] Google Scholar

[ref12] 12. Phung D, Nguyen HX, Nguyen HLT, Luong AM, Do CM, Tran QD, et al. The effects of socioecological factors on variation of communicable diseases: A multiple-disease study at the national scale of Vietnam. PLoS One. 2018;13(3):e0193246. Epub 2018/03/02. pmid:29494623; PubMed Central PMCID: PMC5832231.
View Article
PubMed/NCBI
Google Scholar

[33] View Article

[34] PubMed/NCBI

[35] Google Scholar

[ref13] 13. Yao HW, Yang Y, Liu K, Li XL, Zuo SQ, Sun RX, et al. The spatiotemporal expansion of human rabies and its probable explanation in mainland China, 2004–2013. PLoS Negl Trop Dis. 2015;9(2):e0003502. Epub 2015/02/19. pmid:25692883; PubMed Central PMCID: PMC4334667.
View Article
PubMed/NCBI
Google Scholar

[37] View Article

[38] PubMed/NCBI

[39] Google Scholar

[ref14] 14. Guo D, Yin W, Yu H, Thill JC, Yang W, Chen F, et al. The role of socioeconomic and climatic factors in the spatio-temporal variation of human rabies in China. BMC Infect Dis. 2018;18(1):526. Epub 2018/10/24. pmid:30348094; PubMed Central PMCID: PMC6198482.
View Article
PubMed/NCBI
Google Scholar

[41] View Article

[42] PubMed/NCBI

[43] Google Scholar

[ref15] 15. Yu J, Xiao H, Yang W, Dellicour S, Kraemer MUG, Liu Y, et al. The impact of anthropogenic and environmental factors on human rabies cases in China. Transbound Emerg Dis. 2020;67(6):2544–53. Epub 2020/04/30. pmid:32348020.
View Article
PubMed/NCBI
Google Scholar

[45] View Article

[46] PubMed/NCBI

[47] Google Scholar

[ref16] 16. Manderson L, Aagaard-Hansen J, Allotey P, Gyapong M, Sommerfeld J. Social research on neglected diseases of poverty: continuing and emerging themes. PLoS Negl Trop Dis. 2009;3(2):e332. Epub 2009/02/25. pmid:19238216; PubMed Central PMCID: PMC2643480.
View Article
PubMed/NCBI
Google Scholar

[49] View Article

[50] PubMed/NCBI

[51] Google Scholar

[ref17] 17. Bhutta ZA, Sommerfeld J, Lassi ZS, Salam RA, Das JK. Global burden, distribution, and interventions for infectious diseases of poverty. Infect Dis Poverty. 2014;3:21. Epub 2014/08/12. pmid:25110585; PubMed Central PMCID: PMC4126350.
View Article
PubMed/NCBI
Google Scholar

[53] View Article

[54] PubMed/NCBI

[55] Google Scholar

[ref18] 18. Constantino C, Da Silva EC, Dos Santos DM, Paploski IAD, Lopes MO, Morikawa VM, et al. One Health Approach on Dog Bites: Demographic and Associated Socioeconomic Factors in Southern Brazil. Trop Med Infect Dis. 2023;8(4). Epub 2023/04/27. pmid:37104315; PubMed Central PMCID: PMC10146334.
View Article
PubMed/NCBI
Google Scholar

[57] View Article

[58] PubMed/NCBI

[59] Google Scholar

[ref19] 19. Shoari N, Heydari S, Blangiardo M. A decade of child pedestrian safety in England: a bayesian spatio-temporal analysis. BMC Public Health. 2023;23(1):215. Epub 2023/02/02. pmid:36721178; PubMed Central PMCID: PMC9889245.
View Article
PubMed/NCBI
Google Scholar

[61] View Article

[62] PubMed/NCBI

[63] Google Scholar

[ref20] 20. van de Schoot R, Kaplan D, Denissen J, Asendorpf JB, Neyer FJ, van Aken MAG. A gentle introduction to bayesian analysis: applications to developmental research. Child Dev. 2014;85(3):842–60. Epub 2013/10/15. pmid:24116396; PubMed Central PMCID: PMC4158865.
View Article
PubMed/NCBI
Google Scholar

[65] View Article

[66] PubMed/NCBI

[67] Google Scholar

[ref21] 21. Ugarte MD, Adin A, Goicoa T, Militino AF. On fitting spatio-temporal disease mapping models using approximate Bayesian inference. (1477–0334 (Electronic)).

[ref22] 22. Rue H, Martino S, Chopin N. Approximate Bayesian inference for latent Gaussian models by using integrated nested Laplace approximations. Journal of the Royal Statistical Society Series B (Statistical Methodology). 2009;71(2):319–92.
View Article
Google Scholar

[70] View Article

[71] Google Scholar

[ref23] 23. Taylor E, George K, Johnson E, Whitelegg H, Prada JM, Horton DL. Quantifying the interconnectedness between poverty, health access, and rabies mortality. PLoS Negl Trop Dis. 2023;17(4):e0011204. Epub 2023/04/20. pmid:37079553; PubMed Central PMCID: PMC10118163.
View Article
PubMed/NCBI
Google Scholar

[73] View Article

[74] PubMed/NCBI

[75] Google Scholar

[ref24] 24. Zhou H, Vong S, Liu K, Li Y, Mu D, Wang L, et al. Human Rabies in China, 1960–2014: A Descriptive Epidemiological Study. PLoS Negl Trop Dis. 2016;10(8):e0004874. Epub 2016/08/09. pmid:27500957; PubMed Central PMCID: PMC4976867.
View Article
PubMed/NCBI
Google Scholar

[77] View Article

[78] PubMed/NCBI

[79] Google Scholar

[ref25] 25. Wu H, Song M, Shen XX, Tang Q, Wang SM. Epidemiology of rabies in China from 1996 to 2009. Disease Surveillance. 2011;26(06):427–30+34(in Chinese).
View Article
Google Scholar

[81] View Article

[82] Google Scholar

[ref26] 26. WHO. Density of hospitals (per 100 000 population) [2024 Apr 3]. Available from: https://www.who.int/data/gho/indicator-metadata-registry/imr-details/3361.

[ref27] 27. Moran PA. Notes on continuous stochastic phenomena. 1950;(0006–3444 (Print)). D—CLML: 5019:10193:26:230 OTO—NLM.

[ref28] 28. Zhou XN. Spatial Epidemiology: China Science Publishing & Media Ltd; 2009(in Chinese).

[ref29] 29. Moraga P. Geospatial Health Data: Modeling and Visualization with R-INLA and Shiny: Chapman & Hall/CRC Biostatistics Series; 2019.

[ref30] 30. Wang Y, Lai Y, Du Z, Zhang W, Feng C, Li R, et al. Spatiotemporal Distribution of Hand, Foot, and Mouth Disease in Guangdong Province, China and Potential Predictors, 2009⁻2012. Int J Environ Res Public Health. 2019;16(7). Epub 2019/04/17. pmid:30987085; PubMed Central PMCID: PMC6480297.
View Article
PubMed/NCBI
Google Scholar

[88] View Article

[89] PubMed/NCBI

[90] Google Scholar

[ref31] 31. Knorr-Held L. Bayesian modelling of inseparable space-time variation in disease risk. Stat Med. 2000;19(17–18):2555–67. Epub 2000/08/29. pmid:10960871.
View Article
PubMed/NCBI
Google Scholar

[92] View Article

[93] PubMed/NCBI

[94] Google Scholar

[ref32] 32. Vicente G, Goicoa T, Ugarte MD. Bayesian inference in multivariate spatio-temporal areal models using INLA: analysis of gender-based violence in small areas. Stochastic Environmental Research and Risk Assessment. 2020;34(10):1421–40.
View Article
Google Scholar

[96] View Article

[97] Google Scholar

[ref33] 33. Daniel S, Håvard R, Andrea R, Thiago GM, Sigrunn HS. Penalising Model Component Complexity: A Principled, Practical Approach to Constructing Priors. Statistical Science. 2017;32(1):1–28.
View Article
Google Scholar

[99] View Article

[100] Google Scholar

[ref34] 34. Gomez-Rubio V. Bayesian inference with INLA: Chapman and Hall/CRC; 2020.

[ref35] 35. Richardson S, Thomson A, Best N, Elliott P. Interpreting posterior relative risk estimates in disease-mapping studies. Environ Health Perspect. 2004;112(9):1016–25. Epub 2004/06/17. pmid:15198922; PubMed Central PMCID: PMC1247195.
View Article
PubMed/NCBI
Google Scholar

[103] View Article

[104] PubMed/NCBI

[105] Google Scholar

[ref36] 36. Marta B, Michela C. Spatial modeling. Spatial and Spatio-temporal Bayesian Models with R-INLA2015. p. 173–234.
View Article
Google Scholar

[107] View Article

[108] Google Scholar

[ref37] 37. Naing NN. Easy way to learn standardization: direct and indirect methods. Malays J Med Sci. 2000;7(1):10–5. Epub 2000/01/01. pmid:22844209; PubMed Central PMCID: PMC3406211.
View Article
PubMed/NCBI
Google Scholar

[110] View Article

[111] PubMed/NCBI

[112] Google Scholar

[ref38] 38. Sun J W., Huang X Y., Su J, Li X, L., Xu B, L. Epidemiology and control strategy on human rabies in Henan Province, China. Chinese Journal of Zoonoses. 2016;32(10):939–43(in Chinese).
View Article
Google Scholar

[114] View Article

[115] Google Scholar

[ref39] 39. Yang H, Chen S, Dai Z, Luo L, Sun Q, Zhao S, et al. Integrated Rabies Surveillance—Hunan Province, China, 2020. China CDC Wkly. 2022;4(44):986–9. Epub 2022/12/10. pmid:36483987; PubMed Central PMCID: PMC9709299.
View Article
PubMed/NCBI
Google Scholar

[117] View Article

[118] PubMed/NCBI

[119] Google Scholar

[ref40] 40. Wang XC, Zhang N, Zhou WZ, Qin YF, Li ZF, Hu JL, et al. Disease burden of rabies in Jiangsu province, 2008–2018: a surveillance data-based analysis. Chin J Public Health. 2021;37(01):134–7(in Chinese).
View Article
Google Scholar

[121] View Article

[122] Google Scholar

[ref41] 41. Li H, Li Y, Chen Y, Chen B, Su Q, Hu Y, et al. Mapping rabies distribution in China: a geospatial analysis of national surveillance data. Int J Infect Dis. 2023;131:140–6. Epub 2023/04/09. pmid:37030651.
View Article
PubMed/NCBI
Google Scholar

[124] View Article

[125] PubMed/NCBI

[126] Google Scholar

[ref42] 42. Yue Y, Chen Q, Mu D, Li Y, Yin W. A Descriptive Analysis of Human Rabies in Mainland China, 2005–2020. Int J Environ Res Public Health. 2022;20(1). Epub 2023/01/09. pmid:36612701; PubMed Central PMCID: PMC9819004.
View Article
PubMed/NCBI
Google Scholar

[128] View Article

[129] PubMed/NCBI

[130] Google Scholar

[ref43] 43. Zhang N, Song C, Tao X, Zhu W. Epidemiologic Features of Human Rabies in China from 2015–2021. Zoonoses. 2023;3(1)
View Article
Google Scholar

[132] View Article

[133] Google Scholar

[ref44] 44. Xiao FJ, Zhang XG, Liao YM, LIU QF. Sunshine Duration in China: Variation Characteristics and Its Influence. Chinese Agricultural Science Bulletin. 2020;36(20):92–100(in Chinese).
View Article
Google Scholar

[135] View Article

[136] Google Scholar

[ref45] 45. WOAH. Rabies 2009 [2023 Sep 23]. Available from: https://www.woah.org/app/uploads/2021/03/rabies-1.pdf.

[ref46] 46. Qiao YC, Wang F, He YL, Yang Q, Yang J, Wei YS. Regional and age difference of human rabies prevalence of the past fourteen years in China. (1873–1716 (Electronic)).
View Article
Google Scholar

[139] View Article

[140] Google Scholar

[ref47] 47. Wang XJ, Qiu DR, Chang YH. Analysis on epidemic characteristics of rabies in Luzhou City 2014 to 2020. J Prev Med Inf. 2022;38(07):942–6(in Chinese).
View Article
Google Scholar

[142] View Article

[143] Google Scholar

[ref48] 48. The People’s Republic of China basic medical insurance drug catalog 2023 [2023 Sep 22]. Available from: https://www.gov.cn/zhengce/zhengceku/2023-01/18/content_5737840.htm.

[ref49] 49. Shen T, Welburn SC, Sun L, Yang GJ. Progress towards dog-mediated rabies elimination in PR China: a scoping review. Infect Dis Poverty. 2023;12(1):30. Epub 2023/04/07. pmid:37024944; PubMed Central PMCID: PMC10077633.
View Article
PubMed/NCBI
Google Scholar

[146] View Article

[147] PubMed/NCBI

[148] Google Scholar

[ref50] 50. Liu Z, Liu M, Tao X, Zhu W. Epidemic Characteristics of Human Rabies—China, 2016–2020. China CDC Wkly. 2021;3(39):819–21. Epub 2021/10/02. pmid:34594999; PubMed Central PMCID: PMC8477051.
View Article
PubMed/NCBI
Google Scholar

[150] View Article

[151] PubMed/NCBI

[152] Google Scholar

[ref51] 51. Coleman PG, Dye C. Immunization coverage required to prevent outbreaks of dog rabies. Vaccine. 1996;14(3):185–6. Epub 1996/02/01. pmid:8920697.
View Article
PubMed/NCBI
Google Scholar

[154] View Article

[155] PubMed/NCBI

[156] Google Scholar

[ref52] 52. Fitzpatrick MC, Hampson K, Cleaveland S, Mzimbiri I, Lankester F, Lembo T, et al. Cost-effectiveness of canine vaccination to prevent human rabies in rural Tanzania. Ann Intern Med. 2014;160(2):91–100. Epub 2014/03/05. pmid:24592494; PubMed Central PMCID: PMC4084874.
View Article
PubMed/NCBI
Google Scholar

[158] View Article

[159] PubMed/NCBI

[160] Google Scholar

[ref53] 53. Wei XK, Xiong Y, Li XN, Zheng M, Pan Y, He XX, et al. Vaccination demonstration zone successfully controls rabies in Guangxi Province, China. (1471–2334 (Electronic)).
View Article
Google Scholar

[162] View Article

[163] Google Scholar

Figures

Abstract

Author summary

1. Introduction

2. Data sources and methods

2.1 Variable selection

2.2 Data sources

2.3 Data processing and model building

2.3.1 Data processing.

2.3.2 Variable screening.

2.3.3 Spatial autocorrelation analysis.

2.3.4 Establishment and evaluation of the Bayesian spatiotemporal models.

2.3.5 Cold and hot spot analysis.

2.3.6 Ecological regression.

2.3.7 Statistical software.

3. Results

3.1 Epidemiological characteristics of human rabies in China’s provinces

3.2 Spatial autocorrelation analysis of human rabies

3.3 Results of the Bayesian spatiotemporal model

3.3.1 Analysis of spatial effects.

3.3.2 Analysis of temporal effects.

3.4 Results of the Bayesian spatio-temporal interaction model

3.5 Ecological regression

4. Discussion

5. Conclusion

Disclaimer

Supporting information

S1 Fig. Correlation analysis.

S1 Table. Multicollinearity analysis.

References