Background
The occurrence and transmission patterns of infectious diseases are changing with the accelerated global integration and the impact of climatic, ecological and social environmental changes [
1,
2]. Recent outbreaks of infectious diseases such as Ebola, Zika and Corona Virus Disease 2019 (COVID-19) demonstrate that emerging and traditional infectious diseases are posing and will continue to pose a threat to human life and health, which brings new challenges for the emergency response capacity of the whole world, especially developing countries with limited available resources [
3‐
5]. Since the outbreak of the Severe acute respiratory syndrome (SARS) in 2003, China, as the world’s largest developing country, has made many efforts to build a response system for public health emergencies, promulgating a number of emergency plans, overhauling the national information system for disease prevention and control, and launching National Major Scientific and Technological Special Project for Prevention and Control. Yet the latest epidemic profile of notifiable infectious diseases released by the Chinese Centre for Disease Control and Prevention show that the number of reported cases of notifiable infectious diseases in mainland China remained as many as 699,466, and that 1531 people died between 00:00 on 1 May 2021 and 24:00 on 31 May 2021 [
6]. Thus China still remains under serious threat from infectious diseases.
The primary task for controlling infectious diseases is to understand the distribution characteristics of epidemiology. According to the existing research, about 80% of epidemiological data are spatial in nature [
7]. Therefore, the exact analysis of the spatial distribution characteristics of infectious diseases is a must for the effective study of the causes and other influencing factors of diseases and also for the formulation of effective prevention and control strategies. In recent years, with a mature statistical method, spatio-temporal statistics, researchers can not only conduct a dynamic analysis of the temporal and spatial distribution characteristics of infectious diseases, but also summarize the spatio-temporal transmission patterns by considering the three-dimensional environment in which they occur and are prevalent [
8,
9]. So far, extensive studies of the epidemiological characteristics of infectious diseases have been conducted by many scholars employing spatial statistical methods based on geographic information system (GIS) [
10‐
14].
Modelling spatio-temporal trends in infectious diseases is of great importance, because real-time epidemiological forecasting can help to predict the geographic expansion of diseases as well as the number of cases. What’s more, the exact prediction of the epidemic outbreak helps policy makers to prepare early so that they can better implement public health interventions. As the infectivity of pathogens and the availability of drugs and vaccines change over time, the application of updated shared data is much necessary for the evaluation and prediction of disease hazard. In China, the National Health Commission of the People’s Republic of China regularly releases on its official website the annual incidence of Class A, Class B and Class C notifiable infectious diseases, providing a platform for further study of the epidemiological characteristics of notifiable infectious diseases in China. In the Law of the PRC on the Prevention and Treatment of Infectious Diseases enacted in 1989, infectious diseases are classified into three categories, Class A, Class B and Class C, among which the former two can both cause large-scale severe epidemics within a short period of time, while the third is less infectious and causes only minor outbreaks. The number of notifiable infectious diseases in the three categories is ever-changing with the outbreak of emerging ones. For example, on October 2, 2020, the National Health Commission issued a draft of the revised Law on Prevention and Control of Infectious Diseases for consultation, which clearly states that two new types of infectious diseases, namely human H7N9 avian influenza and novel coronavirus, have been added to Class B. Currently, the three categories contain altogether 41 notifiable infectious diseases, as shown in Additional file
1: Annex 1. Epidemic incidence data have been shown to be valuable epidemiological tools for real-time assessment and prediction of trends and transmission potential [
15‐
18]. By means of the data, predictive models can be used to help provide timely forecasts of disease incidence and geographic spread of emerging epidemics. Auto Regressive Integrated Moving Average (ARIMA) is a time-domain tool for time series analysis that has been widely used for infectious disease prediction [
19,
20]. For example, many scholars have recently used ARIMA models to predict the incidence of COVID-19 [
21‐
23], as it was used to predict the incidence of other infectious diseases such as viral hepatitis [
24], malaria [
25,
26] and measles [
27].
Despite the accumulated findings, there are still limitations in the existing research. In general, most of the current researches on infectious diseases are the study of one specific disease, and the hotspots of research have been focused on a single study of the current epidemiological status, prediction of incidence trends and spatial attributes of the disease, lacking systematic research on the epidemiological characteristics and trend prediction of multi-species combinations. Since there is growing evidence of the complexity of disease interactions and disease etiology, a multi-species analysis to capture the overall trends of infectious diseases may be a worthwhile approach. Therefore, by using the method of spatial econometric analysis and time series forecasting combined and taking 31 Chinese provinces (Hong Kong, Macao and Taiwan are excluded) as the study unit, this study aims to focus on the current spatial and temporal distribution of the incidence of highly infectious Class B notifiable infectious diseases, and to predict the future development trend of the incidence of Class B notifiable infectious diseases. Hopefully, it will help Chinese public policy makers to formulate better health policy intervention measures, build better response of China’s public health epidemic prevention system to disease outbreaks, and provide for other countries valuable information to develop better intervention strategies to prevent and control the spread of emerging and re-emerging infectious diseases.
Discussions
This study uses the incidence rates of Class B notifiable infectious diseases published on the official website of the National Health Commission of the People’s Republic of China from 2007 to 2020 as a data source to analyse the characteristics of the spatial and temporal distribution of the incidence rates of Class B notifiable infectious diseases in China, and to forecast their development trends. The results of the study show that, in terms of spatial distribution, the overall incidence of Class B notifiable infectious diseases in China during the period 2007–2020 showed an increasing distribution of “East-Central-West”, indicating that the incidence of Class B notifiable infectious diseases has a certain correlation with the level of economic development of the region. The incidence rate of Class B infectious diseases is correlated with the economic development level of the region. Previous studies have confirmed that higher levels of economic development can be beneficial to the reduction of likelihood of infectious disease outbreaks, as well as the disease hazards associated with infectious disease outbreaks [
32]. For example, Wu et al. [
33] explored the impact of climate change on human infectious diseases and found that developing countries face greater health risks from infectious disease outbreaks than developed countries because their public health systems lack the resources and capacity to respond effectively to the challenges. The present study confirms that the incidence of Class B notifiable infectious diseases is lower in the economically developed eastern region of China than the central and western regions, which is easily explained by the fact that the eastern region has more medical and educational resources and better infrastructure than the central and western regions, and will suffer less from the burden of disease caused by infectious diseases, especially vaccine-accessible infectious diseases such as measles, whooping cough and tuberculosis [
34]. However, the incidence of Class B notifiable infectious diseases decreased the most in the west, and the regional differences in incidence rates was on a decreasing trend. The analysis suggests that although the western region is relatively lagging behind in economic development, it has benefited from the Western Development Policy in recent years, making it possible to achieve significant results in the prevention and treatment of infectious diseases in the western region. The identification of regional differences in the incidence of Class B notifiable infectious diseases is conducive to the adoption of differentiated infectious disease prevention and control intervention strategies for different regions. Since effective prevention and control of infectious diseases in the central and western regions is crucial to reducing the overall incidence of infectious diseases nationwide, it is much necessary to continue to increase the investment and talent introduction policies for the central and western regions. Through the “policy dividend” [
35], the economic development of the central and western regions and the level of health prevention and control can be improved. Meanwhile, the eastern region, as the driving force in the improvement of infectious disease prevention and control, should interact well with the central and western regions and play a leading role in order to reduce the health hazards of infectious diseases to the Chinese population.
A lot of studies have shown significant spatial clustering in the incidence of Class B notifiable infectious diseases such as typhoid and paratyphoid fever, dengue fever and novel coronavirus [
36‐
39]. According to the current stuy, the Moran’s
I index for the incidence of Class B notifiable infectious diseases in China from 2007 to 2020 was significantly positive, indicating that the incidence of Class B notifiable infectious diseases showed a strong clustering in spatial distribution, which is consistent with the results of previous studies. Besides, the Moran’s
I index values show an increasing trend year by year, indicating a gradual increase in the degree of clustering in the spatial distribution of the incidence of Class B notifiable infectious diseases. In other words, there is a spatial correlation between the incidence of Class B notifiable infectious diseases in each province of China and that in neighbouring provinces, and the degree of correlation is gradually increasing. This result confirms the fact that no provincial unit can control infectious diseases without cooperating with other provinces in the fight against them, and no region can protect itself from an infectious disease crisis alone. Therefore, in order to reduce the incidence of Class B notifiable infectious diseases in a particular province, the level of economic development and the level of sanitary and epidemiological protection in the province as well as in the surrounding provinces need to be taken into account as important factors affecting the effectiveness of infectious disease prevention and control [
40‐
42]. What’s more, in order to reduce the overall incidence of Class B notifiable infectious diseases in China, it is recommended that a regional community for improvement in infectious disease prevention and control be formed by breaking down geographical and administrative barriers and delineating a multi-level regional framework, and that inter-provincial mechanisms for joint prevention and control of infectious diseases be established so as to give full play to the positive spatial spillover effects of the positive prevention and control of infectious diseases across the region.
According to the incidence rates of Class B notifiable infectious diseases in China from 2007 to 2020, the national overall rates as well as the rates in each province are decreasing except a certain rise in four provinces (Guangdong, Hainan, Hunan, and Tibet). This indicates that we have achieved notable results in the prevention and control of infectious diseases in China, which is consistent with the results of previous studies [
43]. According to the research analysis, the fact that the four provinces have not seen a decline in the incidence of Class B notifiable infectious diseases may be mainly related to their special geographical location and climatic conditions [
44‐
49]. In spite of a certain economic boost in Tibet, a border region in the southwest, the increasing trade flows results in the bigger size of the mobile population, which leads to the problem of new and traditional infectious diseases and their increasingly prominent cross-border transmission. Guangdong, Hunan and Hainan all have tropical or subtropical monsoon climates, which is temperate and conducive to the prevalence of climate-sensitive mosquito-borne viral diseases such as dengue, malaria and encephalitis B.
After verification and adoption of the selected best model using the data of 2020, the short-term forecast for 5 years from 2020 to 2024 was carried out using the best model. The results show that, firstly, compared with the predicted 2020incidence assuming no COVID-19 epidemic broke out, the actual 2020 incidence of Class B infectious diseases are smaller for both China as a whole and all the provinces (except Hubei Province), suggesting that the prevention and control measures taken in response to the COVID-19 epidemic are conducive to controlling the occurrence and development of other Class B notifiable infectious diseases, which is consistent with the results of existing studies [
50‐
52]. The study suggests that it is the government’s mandatory prevention and control strategies as well as the public’s increasing awareness of personal health that works. On the one hand, studies have shown that non-pharmaceutical interventions implemented by the government during the COVID-19 epidemic (e.g. school closures, movement restrictions and social distancing) contributed to a decline in the incidence of infectious diseases such as pertussis, scarlet fever and hand, foot and mouth disease (HFMD) [
53]. On the other hand, both the initial period of rigorous epidemic prevention and control and the subsequent period of normal practice of them saw an increase in public awareness of personal health. Such intervention measures as wearing masks, social distancing, hand washing and ventilation also effectively prevent the spread of other infectious diseases transmitted through respiratory tract, intestinal tract or intimate contact, for example, whooping cough, scarlet fever, tuberculosis, and brucellosis, etc. Secondly, the forecast of the incidence of Class B notifiable infectious diseases in China from 2021 to 2024 shows that the incidence of infectious diseases is still on a downward trend nationwide and in most provinces, but the predicted incidence of Class B notifiable infectious diseases in such provinces as Guangdong, Guizhou, Hunan, Hainan, Tibet and Guangxi is on an upward trend, which suggests that the government should focus more attention upon provinces where the incidence of infectious diseases is on an upward trend, and tailor specific measures to the actual situation of each province to avoid the re-emergence of certain infectious diseases in a province or even nationwide and the outbreak of new infectious diseases due to relaxed vigilance. The ARIMA model is highly accurate (within 10%) in predicting the incidence of Class B notifiable infectious diseases, and can effectively compensate for the current lack of capacity to develop, evaluate, manufacture, distribute and manage effective medical countermeasures (e.g. vaccines, diagnostics, etc.), effectively address the unmet disease burden associated with outbreaks or prevalence of traditional and emerging infectious diseases, and effectively guide policy decisions such as rational allocation of health resources and pre-deployment of emergency supplies [
54].
There are, of course, certain limitations in this study. The first limitation is about the prediction accuracy of the ARIMA model. On the one hand, the prediction accuracy of the ARIMA model is easily affected by the sample size. The bigger the sample size, the higher the prediction accuracy of the model, with the number of variables unchanged. In this study, based on the principle of indicator data availability, 13 years of data (2007–2019) were finally selected for fitting the ARIMA model, using 2020 data for model validation and forecasting the incidence of Class B notifiable infectious diseases in China as well as in each province from 2021 to 2024. Although the relative errors of the selected models were less than 10% for China as a whole and for each province (except Hubei Province), the prediction accuracy could be further improved in the future by increasing the sample size. On the other hand, it is worth mentioning that while this study recognises that notifiable infectious diseases in Class B are more infectious and may lead to outbreaks or epidemics with relatively high sensitivity and specificity compared with notifiable infectious diseases in category C. The use of incidence data for Class B notifiable infectious diseases can effectively reduce the reduction in predictive accuracy of ARIMA models due to missing reports. However, it is worth mentioning that the prediction results may deviate from the actual values over time due to the changing external environment of the host, e.g. policy instability and vaccine availability [
55]. Therefore, it is much necessary to update the forecast of the incidence of Class B notifiable infectious diseases in accordance with the availability of data. Secondly, this study uses the annual incidence of Class B notifiable infectious diseases, a multi-disease joint data, for the prediction study. While this is good for getting a full picture of infectious diseases, it also has shortcomings because even if problems can be identified, it still takes time to determine the specific situation. Future applications will need to be tailored to the specific characteristics of each disease. Finally, the study on the incidence of Class B notifiable infectious diseases is mainly an exploratory spatial data analysis. Future studies can be made on the basis of the existing studies by combining other variables such as economic factors, transport factors and population mobility factors to conduct empirical spatial data analysis through the establishment of spatio-temporal regression models and to explore the causes and driving mechanisms of the formation and changes in the spatio-temporal patterns of the incidence of Class B notifiable infectious diseases.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.