Spatial, temporal, and spatiotemporal analysis of mumps in Guangxi Province, China, 2005–2016

Mumps is an acute communicable disease caused by the mumps virus [1]. As a highly contagious disease, mumps is characterized by the clinical manifestations that swelling of the parotid glands, accompanied with pain and fever. Although this infection in the majority cases is mild, its complications, such as meningitis. Orchitis, ovarian inflammation, and deafness, affecting the nervous system, digestive system and causing other multiple organ damages are serious [2]. In most parts of the world, the annual incidence of mumps in the absence of immunization ranges from 100 to 1000 cases/100,000 population, with epidemic peaks every 2–5 years [3]. The incidence of mumps is rapidly declining because of the widespread use of vaccine, especially in combination with vaccines for measles and rubella (measles–mumps–rubella [MMR] vaccine). The United States, Australia, and many European countries such as the United Kingdom, Italy, etc. have conducted routine administration of 2 doses of MMR vaccine for children since the last century [4–7]. Nevertheless, mumps resurgences have been observed in many countries, even in those countries with higher immunization rates since the beginning of the new century [8–11]. This phenomenon may be due to variation in the virus strain, periodicity of the disease, attenuation of the body antibody and other factors [12–14]. Studies found that using three doses of MMR vaccination can significantly increase the crowd’s immunity to mumps [7, 15]. In addition, increasing the coverage of vaccination and timely supplementary immunization activities (SIA) in high-risk areas are also keys to preventing mumps outbreak. China, the world’s most populous country, has the largest number of reported mumps cases, which accounted for approximately 52.19% in Asia-Pacific countries and 30.01% in the world in 2016 [16]. Monovalent vaccine for mumps has been applied in China since the 1990s. In 2008, a vaccine containing the component of mumps was included in the expanded program on immunization (EPI). Children between the ages of 18 and 24 months can receive a free dose of MMR. In this way, the mumps epidemic has been greatly weakened due to the reduction of susceptible population. However, only a few regions have implemented two doses MMR immunization plan in China. The incidence of mumps in some provinces, such as Guangxi, Ningxia, and Hainan, is higher than the national average because of the low immunization rate. Previous studies on mumps in China had focused on its epidemiological characteristics rather than its spatial and temporal distribution. Spatiotemporal analysis on mumps in local areas mainly concentrated in the northern parts of China, such as Shandong, Gansu, and Shanxi [17, 18]. The economy, climate, and geography of Guangxi, as a high-incidence area of mumps and mountainous province in southern China, differ from those of the northern region of China. Therefore, Guangxi was chosen as the ideal site for spatiotemporal analysis in this study.

Many infectious diseases, including malaria, measles, Japanese encephalitis, and mumps, exhibit heterogeneity in spatial and temporal distribution [19–22]. Spatial statistical methods, based on the geographical information system (GIS), have been widely applied for their strong statistical power in the surveillance of communicable diseases and the evaluation of the effectiveness of preventive interventions in the past decades [23, 24]. Spatiotemporal analysis can not only visualize the epidemiological data and help observe the geographical distribution of disease from a more intuitive point of view, but identified spatial clusters and temporal clusters from a deeper level and determine the high- and low-risk areas. People inhabiting high-risk areas are the target population for health agencies to implement SIAs.

In this study, spatial autocorrelation and spatiotemporal scan analyses were conducted to statistically evaluate the significance of aggregation and determine the size of the range of hot-spots at the county level in 2005 and 2016. The results of these analysis could enhance the understanding of the epidemic characteristics and spatiotemporal patterns of mumps in Guangxi and provide a scientific basis for policy formulation and health resource allocation.

This study clarified the basic epidemiological characteristics of mumps and illustrated its substantial changes occurring with respect to the spatial, temporal and spatiotemporal trends and clusters in Guangxi, China during 2005–2016, using GIS and scan statistics. Results showed that the annual incidence rate and the annual reported cases of mumps were declining as a whole, from 41.84(per 100,000) (19,338) in 2005 to 19.13(per 100,000) (9176) in 2016. This result may be due to the introduction of mumps vaccine in China in the 1990s. However, two outbreaks occurred in 2006 and 2011. Similar phenomena can be observed in other parts of China [17, 18]. Likely because of the periodic changes in the incidence of mumps. Before 2007, mumps vaccination was based on the principle of voluntary inoculation and self-paying inoculation. Given the limitations of medical resources and economic conditions in different regions, the coverage of vaccination is low, especially in the underdeveloped province of Guangxi. An effective antibody barrier did not form in the crowd and susceptible populations continued to accumulate, thereby leading to a cyclical outbreak of mumps [29]. MMR was included in the EPI by the Chinese government in 2008. Since then, children aged 18 and 24 months have been receiving a free dose of mumps vaccine. Supplementary vaccination and the second dose of mumps vaccination have also been implemented in some areas with high incidence. All of these interventions led to a rapid decline in the incidence in high-risk group of 5–9 years old after 2013, and the overall number of cases simultaneously decreased. This result suggested that the prevention and control for mumps had been effective to some extent.

The onset of mumps was mainly observed in students and childcare children in 5–9 years old. This finding is consistent with previous results [9, 22, 30]. Mumps is a respiratory disease, and adolescents and children are generally susceptible. Children aged 5–9 years old should attend kindergarten or elementary school in China. The low immunity of children, dense population, crowded environment and personal hygiene are important factors that cause its transmission. The increasing incidence of mumps among the elderly population is another matter of concern in this study. On the one hand, the vaccination coverage of mumps in this age group has remained low. Although the circulation incidence can provide immunity from natural infection to people, the elderly population has less opportunity to obtain such natural immunity because the occurrence of mumps epidemic gradually decreases as a result of the implementation of the national immunization planning policy and effective control of infectious diseases [31]. On the other hand, a constant attenuation of mumps antibodies is observed in human serum. Differences in antibody levels between populations may be induced by several factors, including uptake of mumps vaccine, different immunogenicity of mumps vaccines or variation in schedules, (including spacing between doses), and previous exposure to mumps infection. Even two doses of mumps vaccine will not achieve a 100% seroconversion; hence, a group of susceptible individuals remains in the population [32, 33]. In addition, previous studies found that mumps virus genotypes in different Chinese provinces vary, but the present mumps vaccine mainly targets the F type. The immunization effect between different F subtypes and other types of virus has yet to be evaluated [34]. All of these factors may increase the risk of infection in the elderly and children, suggesting that the enhanced immunization and supplementary vaccination should be implemented for these special populations.

In the current study, mumps cases were reported in all 12 months, but most occurred between April and July, with a small peak in October and January. This bimodal seasonal distribution was consistent with that of northern China [17, 18, 34] and prompts that there may be a relationship between the mumps incidence and the climate. A study in Guangzhou found that mean temperature, relative humidity, and atmospheric pressure can increase the risk of mumps, whereas wind speed has a protective effect [35]. Another study in Jining showed that temperature, humidity, and sunshine have a linear relationship with mumps incidence when exceeding a certain threshold [36]. Our previous research also indicated that meteorological factors, such as temperature and wind speed, can exert a significant impact on the mumps incidence, which can be modified by several socioeconomic factors (submitted). According to the seasonality of mumps incidence, local medical institutions should formulate and implement targeted intervention strategies before the advent of disease peak period.

The spatial analysis in this study demonstrated a significant spatiotemporal heterogeneity throughout Guangxi in 2005–2016. The results of the global autocorrelation analysis revealed a spatial positive correlation in the study period, whereas the annual analysis did not show the same trend. This phenomenon indicated that the high-incidence areas of mumps are unstable. The results of the LISA and purely spatial cluster analysis showed a remarkable variation in the spatial distribution of mumps in Guangxi, with most high-risk counties located in the cities of Liuzhou, Nanning, and Chongzuo. The counties where mumps constantly emerged were scattered in remote mountainous areas and in economically developed cities. With the increasing number of young people born in mountainous areas but leaving their hometowns to seek job opportunities in urban areas, many children are left at home supported by their grandparents. Given the remote distance, traffic congestion, economic backwardness, and serious brain drain, mountainous areas are characterized by poor access to health services, scarce health resources, and inadequate health care services for children. The timeliness and completeness of mumps vaccination are also not guaranteed [37]. This condition explains why some clusters are found in poor counties located in the western and northern parts of Guangxi, such as Jiangzhou, Xilin, and Luocheng. In addition, only a few children who migrate to the urban or suburban areas with their parents can timely receive free mumps vaccination [21, 38]. Meanwhile, the current immunization strategies do not allow local children to develop high antibody concentration. Hence, a disease outbreak may occur easily in labor-importing regions like Xingning, Wuming, Yufeng, and Liunan. Increasing the coverage of health services and implementing two or three doses of immunizations may be the focus of local health agencies.

The first three likely clusters identified using the spatiotemporal clustering analysis were located in the western, northern, and central parts of Guangxi, respectively, which were consistent with the hot spots of LISA. However, these high-risk areas were not the same as those of the purely spatial scan analysis possibly because of the difference in the analysis variables included in the two scanning methods. The temporal clusters identified using the spatiotemporal scan statistics were mainly recorded between 2009 and 2012. The trend of annual incidence and the seasonal pattern also indicated that the incidence of mumps was higher in this period than in other periods, suggesting a resurgence, which has yet to be examined for specific causes.

In addition, this study found that spatiotemporal analysis has unique advantages in detecting the high incidence of disease and determining the scope of prevention and control compared with the traditional methods of epidemiological analysis [39–41]. LISA has a better chance of detecting true cluster areas with low false-positive rates especially performing well on outlier detection [42]. Simultaneously, Kulldorff’s statistical analysis has a strong power to detect mumps spatial distribution and focus on the extent of mumps [43]. These two spatial analysis methods have been widely used in the field of public health. The use of a combination of these methods can enhances the comprehensive understanding of the spatial distribution patterns and characteristics of the disease [44–46].

The results of this study should be interpreted in consideration of the following limitations. First, data are obtained from a passive surveillance system in China, which means that some cases of mumps may have been underreported because of their sub-clinical symptoms, even though quality control with data collection has been an important component of the disease surveillance system in China for all reportable diseases [47]. Second, a circular scan window is used in the purely spatial and spatiotemporal scan statistics. Although a circular scan window requires less computational time and performs well at detecting for large clusters, it cannot effectively identify the narrow, long, and noncontiguous areas [48]. Finally, this study only analyzed the spatial distribution pattern and epidemiological characteristics of mumps between 2005 and 2016. The environmental factors, meteorological data, urbanization, and sociological factors that may drive the mumps epidemic have yet to be examined in detail. Thus, a flaw may exist in the interpretation of the causal relationship of the disease.

This study, to our knowledge, is the first to described the epidemiological characteristics of mumps by using the spatiotemporal analysis at the county level in Guangxi, in southern China and identified the spatial and spatiotemporal high-risk clusters from 2005 to 2016. Though, the mumps incidence generally decreased to a low level, there still exist some high-risk areas of mumps, where mainly concentrated in the western, northern and central regions of Guangxi. Remote urban districts and densely populated urban areas may become mumps epidemic areas. Male children and children aged 5–9 years are susceptible to the mumps incidence in this study. Measures targeting these vulnerable groups and high-risk areas should be developed to prevent future mumps outbreak and reduce geographical inequities in the burden of mumps and allocation of health resources.