VL holds significant importance for global public health, with broad distribution throughout the world, leading to substantial morbidity and mortality each year [30, 31]. Because of the diversity and complexity of VL transmission, timely VL-associated epidemic estimates and assessments among a large-scale population over a long period are necessary to inform a country’s preventative and control measures. In this study, we explored the epidemic trend and spatial–temporal distribution characteristics of VL cases in mainland China over a 16-year study period. The results revealed a declining trend in China from 2004 to 2019. We also found several distinct endemic patterns in different provinces. The northwestern regions, including Gansu, Sichuan, and especially Xinjiang, experienced a decreasing trend in general, while the central areas, including Shanxi and Shaanxi, witnessed an increasing trend, especially in recent years. Significant spatial–temporal aggregation areas of VL were detected during our study period. The clustering regions were initially distributed in northwestern China, and then shifted to central China in the last 2 years. The epidemiological feature analysis indicated that there were significant changes in the seasonal patterns and age structure of VL cases. The changes found in our study could inform the design of targeted preventative and control strategies against VL in China.
VL incidence in China during 2004–2019 exhibited substantial fluctuation, with peaks observed in 2008–2009 and 2015. This trend was basically in accord with that of Xinjiang. As one of the most highly affected VL endemic areas in China, Kashi Prefecture in Xinjiang witnessed two outbreaks, in 2008–2009 and 2015. The case numbers in Kashi Prefecture were very high during these two periods and therefore significantly affected the trend for mainland China [32]. However, according to Joinpoint regression analysis, an overall downward trend was detected. This might be related to the decreasing trends occurring in Gansu, Sichuan, and especially Xinjiang, which were previously the most seriously affected epidemic provinces.
Among the eight epidemic provinces, substantially different trends were discovered. Joinpoint regression indicated that Shanxi and Shaanxi witnessed significantly increasing trends. This is consistent with previous studies, which also revealed a resurgence of MT-ZVL in Shanxi and Shaanxi provinces and showed that cases of infection clustered mainly in the Shanxi-Shaanxi border areas and eastern Shanxi province [33]. This resurgence might be associated with several factors. Firstly, Shanxi and Shaanxi are both MT-ZVL endemic provinces, with their mountainous areas being the epidemic regions. Canine animals are the main reservoir host and infectious source of MT-ZVL. Thanks to the control program conducted in the 1950s, the number of stray and domestic dogs in these areas declined considerably. However, VL transmission in these regions was not completely interrupted. With the rapid economic and societal development, a growing number of rural residents flooded the cities seeking better job opportunities. As a result, young children and elderly people were left behind, and they began keeping dogs for safety reasons. Combined with other suitable transmission factors, namely vectors, climate, and environment, a resurgence in VL therefore occurred in these two provinces. Secondly, the predominant control strategy in MT-ZVL epidemic regions was patient treatment. However, this measure cannot effectively reduce VL incidence. Thirdly, although the application of insecticide-impregnated dog collars had been proven to be effective in other countries [34], there are some difficulties in large-scale implementation in China. Fourthly, spraying of insecticides had limited effects on vector control because of the exophilic behavior of the sand fly in MT-ZVL endemic areas. Additionally, Shanxi and Shaanxi provinces border Hebei and Henan, where VL cases also re-emerged. This might be related to latent infection or imported infectious sources, which need to be further investigated. On the contrary, despite an initial upsurge, VL incidence in Gansu and Sichuan experienced significantly decreasing trends. Xinjiang was initially the most heavily affected VL epidemic region in China, and remained an epidemic province for AVL and DT-ZVL, with AVL distributed in the oases of the plains of Kashi Prefecture and DT-ZVL located in the desert regions in southern Xinjiang. However, in the last 3 years, VL incidence in Xinjiang declined dramatically, with only three cases reported in 2019. In these three provinces, the majority of VL cases were mainly concentrated in a few counties, all of which are located in underdeveloped areas, where almost no funds were previously reserved for disease prevention [35]. The reduction revealed in this study might be thanks to the national program of fighting against poverty. With strong support from the government, the residential environmental and sanitary conditions in rural areas were improved considerably. For instance, a large-scale project was undertaken in epidemic areas to replace earthen walls in residential housing with concrete walls, which effectively reduced the sand fly breeding areas and therefore reduced the contact frequency between sand flies and humans. However, several challenges still exist in these regions. Vectors are widely distributed in this area and difficult to control. Latent infection and wild animal infectious sources might still exist in these regions, which makes VL resurgence possible. The results of spatial autocorrelation and spatial–temporal aggregation analysis indicated that the cluster areas of VL shifted from northwestern China (including Gansu, Xinjiang, and Sichuan) to central China (including Shaanxi and Shanxi), which is in accord with that of Joinpoint regression. It also coincides with a previous study in mountainous areas, which revealed a gradual shift in VL cluster areas from southern Gansu and northern Sichuan province to the Shanxi-Shaanxi border areas and eastern Shanxi province [33]. These changes found in our study illustrate that MT-ZVL endemic regions in central China require more attention, and integrated control measures need to be implemented.
Significant changes in seasonal patterns were revealed in 2018–2019, which might be predominantly associated with the decline that occurred in Xinjiang. DT-ZVL was previously seriously endemic in Xinjiang, with the epidemic peak typically seen in October–November. However, after the reduction in Xinjiang, these peaks have disappeared in recent years, leaving only one peak in March to May. Children aged 0–2 remained the most at-risk population throughout our study period, although the age structure of VL cases changed in 2018–2019. This might be associated with the following. Firstly, the most at-risk population for DT-ZVL was children aged 0–2. With the decline of DT-ZVL in Xinjiang, case numbers in this age group dropped significantly. Secondly, among MT-ZVL cases, the proportion of children aged 0–2 has also decreased in recent years, probably because the age structure of rural residents changed in endemic regions [12, 36]. These findings indicate that attention should be focused on all age groups when performing preventative measures, and comprehensive control strategies are required before and during March to May.
Based on the trends and variations in VL prevalence in mainland China, some measures should be taken to achieve further reductions: (1) Controlling infectious sources. During the epidemic seasons, large-scale screening of infectious sources should be conducted among at-risk population in endemic areas. Once cases are diagnosed, medical treatment should be provided immediately. Meanwhile, measures including killing infected dogs and restricting domestic dog-keeping should also be taken. (2) Reducing VL transmission. Insecticide should be sprayed in vector habitats before their breeding season and during transmission season to reduce the density of sand flies. (3) Protecting vulnerable populations. Health education should be popularized to encourage residents to use insecticide-impregnated nets and sand fly repellent incense, reduce outdoor activities, avoid traveling to epidemic areas, and prevent sand fly bites. (4) Strengthening surveillance. The most effective approach for tracking the changes in VL prevalence is facility-based surveillance in the outpatient or hospital setting. (5) Improving the diagnostic capacity of clinicians. In non-epidemic areas, there is a general lack of awareness about VL diagnosis among clinicians, which could lead to serious outcomes among patients and the spread of VL. Therefore, training programs among clinicians are required to achieve early detection and treatment.
Despite the many important discoveries revealed in this study, it is not without limitations. Firstly, there are three types of VL endemic in eight provinces in mainland China, which demonstrate substantial differences in epidemiological characteristics. However, because of the limited surveillance data collected by the NDRIS of China, at present there is a lack of information on precise epidemiological type and indigenous/imported classification of individual VL cases within mainland China. Therefore, it was not possible to conduct trend and spatial cluster analysis stratified by VL epidemiological type and indigenous/imported classification in this study. However, based on the distinctive endemic regions of these three types of VL, we analyzed and described them according to their respective epidemic provinces. We hope this paper can provide broader insight into the epidemic trends and spatial–temporal distribution of VL in China. Secondly, for multiple reasons, differentiating the clinical characteristics of VL from those of similar diseases is difficult, which might result in some VL cases being misdiagnosed. Additionally, some deaths might be classified into other categories, which may underestimate the true level of VL incidence and mortality. Ongoing studies in this area will hopefully raise awareness among clinicians and the public regarding VL, which might alleviate this situation. This study was able to identify the endemic trends of VL in different regions and identify its spatial–temporal clustering over an extended period, which revealed that, although decreased at the national level, VL incidence saw a significantly increasing trend in Shanxi and Shaanxi provinces. The focus of prevention and control efforts should therefore be shifted from northwestern to central China, especially Shanxi, Shaanxi, and Gansu provinces. Future investigations are needed to collect information on epidemiological type and indigenous/imported classification of individual VL cases. Additionally, stratified analysis of trends and spatial clusters should be performed at the city or county level in relatively small regions to further explore epidemic patterns of VL and provide timely information on prevention and control.