Time series analysis of cumulative incidences of typhoid and paratyphoid fevers in China using both Grey and SARIMA models

Jiaqi Gao; Jiayuan Li; Mengqiao Wang

doi:10.1371/journal.pone.0241217

Abstract

Typhoid and paratyphoid fevers are common enteric diseases causing disability and death in China. Incidence data of typhoid and paratyphoid between 2004 and 2016 in China were analyzed descriptively to explore the epidemiological features such as age-specific and geographical distribution. Cumulative incidence of both fevers displayed significant decrease nationally, displaying a drop of 73.9% for typhoid and 86.6% for paratyphoid in 2016 compared to 2004. Cumulative incidence fell in all age subgroups and the 0–4 years-old children were the most susceptible ones in recent years. A cluster of three southwestern provinces (Yunnan, Guizhou, and Guangxi) were the top high-incidence regions. Grey model GM (1,1) and seasonal autoregressive integrated moving average (SARIMA) model were employed to extract the long-term trends of the diseases. Annual cumulative incidence for typhoid and paratyphoid were formulated by GM (1,1) as and respectively. SARIMA (0,1,7) × (1,0,1)₁₂ was selected among a collection of constructed models for high R² and low errors. The predictive models for both fevers forecasted cumulative incidence to continue the slightly downward trend and maintain the cyclical seasonality in near future years. Such data-driven insights are informative and actionable for the prevention and control of typhoid and paratyphoid fevers as serious infectious diseases.

Citation: Gao J, Li J, Wang M (2020) Time series analysis of cumulative incidences of typhoid and paratyphoid fevers in China using both Grey and SARIMA models. PLoS ONE 15(10): e0241217. https://doi.org/10.1371/journal.pone.0241217

Editor: Ka Chun Chong, The Chinese University of Hong Kong, HONG KONG

Received: March 16, 2020; Accepted: October 9, 2020; Published: October 28, 2020

Copyright: © 2020 Gao et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All incidence and death statistics files for typhoid and paratyphoid are available from the Chinese Public Health Science Database (http://www.phsciencedata.cn).

Funding: Research by Mengqiao Wang is funded by a Basic Research Grant from the Department of Science and Technology of Sichuan Province (Grant number: 2019YJ0035). 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.

Introduction

Typhoid and paratyphoid fevers, collectively referred to as enteric fever, are caused by systematic infection with the gram-negative bacterium Salmonella enterica serotype S. typhi and S. paratyphi (types of A, B, and C) [1]. The organisms enter the patients via the gastrointestinal tract and get into the bloodstream via the lymphatic channels, and a mouse model has been engineered [2]. Sanitary measures and personal hygiene play instrumental role as infections generally occur after intaking food or water contaminated by urine or feces [3]. The incubation period could generally last from 3 to 42 days, with on average 14 days for typhoid and 2–15 days for paratyphoid [4]. Clinical manifestations include high-temperature fever, prostration, fatigue, headache, and gastrointestinal reactions, with serious complications such as intestinal bleeding and perforation [5]. With symptoms not exclusive compared to other types of fevers, diagnosis of both typhoid and paratyphoid is conducted through clinical culture and test of patients’ blood, stool, or urine. Live-attenuated oral vaccine or capsular polysaccharide vaccine are currently available for prevention, and treatment options include ceftriaxone, ciprofloxacin, or azithromycin.

Both types of fevers are more common in developing than developed countries. In south Asia, southeast Asia, and sub-Saharan Africa areas with poor water supply and sanitation, they are a major cause of death and disability, especially among children, and have a significant impact on social and economic development [6–8]. Environmental factors, such as climate, have also been investigated to assess their influence on water-food-borne infections [9–12]. According to the estimate by the World Health Organization (WHO) in 2010, nearly 22 million cases of typhoid fever occur annually, with at least 200,000 deaths [10]. China has been among the high-incidence areas, with cumulative incidence about 10–50 per 100,000 before 1990 but gradually decreasing since then [13]. Nevertheless, typhoid and paratyphoid fevers remain important sporadic intestinal infectious diseases and are directly monitored nationwide. Descriptive epidemiology research has been conducted on specific provinces [14–17], but there remains a lack of national analysis on time series modeling which is key for the prevention and control measures [18]. Several mathematical models have been employed to predict the incidence of infectious diseases [19]. Based on the window size of the independent variable, the grey model (GM) predicts future value of the time series using only the most recent set of data. The GM assumes all data are positive (consistent with cumulative incidence) and sampling frequency of the time series is fixed (true for annual or monthly datasets). The grey model does not require errors in normal distribution and is hardly limited by small sample size [20]. Seasonal Autoregressive Integrated Moving Average (SARIMA) model, simultaneously taking general trend, periodic pattern, and random disturbance into consideration, has been widely applied in the research on various infectious diseases with periodic pattern, e.g. malaria [21, 22], hand-foot-mouth disease [23], AIDS [24], and tuberculosis [25]. Due to the infectivity and seasonality of infectious disease, SARIMA has more predictive power compared to other mathematical models [26, 27].

This study aims to conduct a thorough analysis of the long-term cumulative incidence of both typhoid and paratyphoid fevers in China at both national and regional levels, at both yearly and seasonal windows, and on the susceptibility of people of various ages. Statistical models of GM (1,1) and SARIMA would also be applied to fit historic yearly and monthly data for revealing the underlying structure of the infectious trends, and to forecast future incidence to provide data-driven knowledge for the prevention and control of these fevers.

Methods

Data sources

Typhoid or paratyphoid fevers were categorized as type-II infectious diseases by Centers for Disease Control and Prevention of China, and the historic incidence data from 2004 to 2016 were retrieved from the portal of Chinese Public Health Science Database (CPHSD, http://www.phsciencedata.cn/). Detailed data for 2017 and 2018 were not yet published in the database, but annual incidence for the combined fevers in both years was reported by official government news releases. Regional data covered 31 provincial regions in mainland China. All patients were diagnosed using criteria (GB 16001–1995) promulgated by the Ministry of Health of China: the patient had persistent fever (higher than 40°C) of unknown origin, accompanied by positive serum antibody (titer ≥ 1:80 by the Widal test, agglutination titer of typhoid or paratyphoid flagella ≥ 1:160), or the patient had unexplained fever, and S. typhi or S. paratyphi could be isolated in any specimen of serum, bone marrow, feces, or bile.

Time series models

GM (1,1) model—for GM (n, m), n is the order of the differential equation and m is the number of variables. GM (1,1) model can be used to predict incomplete or uncertain data. With original time series data x⁽⁰⁾(t) = (x⁽⁰⁾(1),x⁽⁰⁾(2),x⁽⁰⁾(3),…,x⁽⁰⁾(n)), define cumulative sum of . The adjacent mean is z(t) = (y(t)+y(t−1))/2.

With white-formed ordinary differential equation, D, a and u are estimated by the equations as follow: (1) (2) (3)

Substitute the values of a and u into the formula: then the estimated value of x⁽¹⁾(t) can be calculated as . The grey model can be used for extrapolative prediction only after passing the accuracy test of small error probability and posterior error ratio.

SARIMA model—in SARIMA(p, d, q) (P, D, Q)_s, p, d, and q annotate the order of auto-regression, the degree of trend difference, and the order of moving average respectively; P, D, and Q represent the seasonal auto-regression lag, the degree of seasonal difference, and the seasonal moving average; s annotates the length of the cyclical pattern. In SARIMA model fitting, the Augmented Dickey-Fuller (ADF) unit-root test evaluated the stationary status of time series; seasonal and non-seasonal differences were adopted to stabilize the term trend and periodicity. Parameters of SARIMA model were estimated by autocorrelation function graph and partial autocorrelation graph. Models of varying orders of p, q and P, Q were tested through Box-Jenkins test, and all models passing the residual test were evaluated for performance using criteria including mean absolute percentage error, root mean square error, Akaike Information Criterion (AIC), and R². For SARIMA, data from 2004 to 2015 were used in model fitting, and data in 2016 were used as out-of-sample set for prediction validation.

Data analysis

Data analysis and visualization were performed in version 4.0.0 of the R statistical software (R Core Team, 2020). Significance level of 0.05 is used in the null-hypothesis test unless otherwise specified.

Results

Typhoid and paratyphoid fevers in China

Typhoid and paratyphoid fevers were by law monitored as type-II infectious diseases in China, and we tracked their recent long-term trends from 2004 to 2016 (Table 1). Nationally, incidence counts displayed a significant trend of decrease: almost 33000 cases of typhoid fever was reported in 2004 but the count fell below 10000 and stayed in the close range since 2010; more than 16000 cases of paratyphoid fever was reported in 2004, and the continual downfall reached to 2311 cases in 2016 (Fig 1A). With slightly expanding population during this period, the cumulative incidences mirrored and in fact better revealed the historic trends, observing a drop of 73.9% for typhoid fever and 86.6% for paratyphoid fever (Fig 1B). Cumulative incidence of typhoid fever was consistently higher (2–4 folds) than that of paratyphoid fever, and by 2016 they have respectively dropped to 0.63 and 0.17 cases per 100,000 population. Although the cumulative incidences were relatively low compared to other intestinal infectious diseases such as dysentery (9 cases per 100,000) and hepatitis (89 cases per 100,000), both types of fevers still appeared as infectious diseases of noticeable concern given their still high count of cases in the most populous country on earth. In contrast to the number of cases, deaths caused by both fevers were few (Fig 1A), and the very low mortality was different compared to other severe infectious diseases such as seasonal flu and severe acute respiratory syndrome (SARS). With characterization of cumulative incidence in months, both types of fevers displayed a clear seasonal pattern which peaked in the summer (mainly from June to August) and bottomed in the winter (interestingly, there appeared an “outlier” outbreak of both fevers in January of 2012) (Fig 1C).

Download:

Fig 1.

Annual incidence and death counts (A), annual cumulative incidence (B), and monthly cumulative incidences (C) of typhoid and paratyphoid fevers in China from 2004 to 2016. Note: sum of monthly incidence equals to the corresponding annual incidence.

https://doi.org/10.1371/journal.pone.0241217.g001

Download:

Table 1. Statistics of typhoid and paratyphoid fevers in China, 2004 to 2016.

https://doi.org/10.1371/journal.pone.0241217.t001

Age distribution of typhoid and paratyphoid fevers

People of different ages were substantially unalike with regards to physical, physiological, and psychological conditions, and consequently may be differentially susceptible to certain types of infectious diseases. In the earlier years of 2004 and 2005 when both typhoid and paratyphoid fevers were widespread, we observed higher cumulative incidence for young (15–29 years old) and middle-aged (30–44 years old) subgroups; however, as cases sequentially fell across all age groups throughout the following years, there no longer appeared a significant difference in cumulative incidence for different age groups except that the young children subgroup (0–4 years old) maintained the highest risk at double the level of other age groups (Fig 2A). Further analysis revealed that there was a quite consistent composition of typhoid fever and paratyphoid fever in 4:1 ratio across all age groups (more evident in recent than earlier year) (Fig 2B).

Download:

Fig 2.

Age-specific annual cumulative incidence (A) and composition (B) of typhoid and paratyphoid fevers in China from 2004 to 2016.

https://doi.org/10.1371/journal.pone.0241217.g002

Geographical cluster of typhoid and paratyphoid fevers

National statistics in China generally covered 31 provincial regions, and accordingly we analyzed the annual cumulative incidence of typhoid and paratyphoid fevers by region. The sequential downward trend of cumulative incidence was observed universally in almost all provinces (S1 Fig). There also existed a geo-separation effect: cumulative incidences were clearly higher in southern provinces than northern provinces (Fig 3, S2 and S3 Figs). The top high-incidence provinces were Yunnan, Guizhou, and Guangxi (Fig 3), and interestingly, all three provinces were geographically adjacent and located to the southwestern part of China (bordering with Vietnam, Laos, and Myanmar). The top provinces also matched the similar decreasing trend and the seasonal pattern observed in the nation for both fevers (Fig 4A and 4B). Standing out from all others, Yunnan was the province of the highest cumulative incidence for both typhoid and paratyphoid fevers, at levels as 10–12 times high as national average and even 3–4 times high as the second highest province of Guizhou.

Download:

Fig 3. Region-specific annual cumulative incidence of typhoid and paratyphoid fevers in China from 2004 to 2016.

Red marker annotates the median value. The regions are ordered by the median cumulative incidence of the combined typhoid and paratyphoid (sum of the red markers).

https://doi.org/10.1371/journal.pone.0241217.g003

Download:

Fig 4.

Annual (A) and monthly (B) cumulative incidences of typhoid and paratyphoid fevers in a cluster of three provinces (A. Yunnan, B. Guizhou, C. Guangxi).

https://doi.org/10.1371/journal.pone.0241217.g004

Prediction of annual incidence by GM (1, 1)

GM (1,1), a basic type of grey model with first-order equation and single variable, was applied on both typhoid and paratyphoid fevers, and resulted in smooth exponential models that effectively tracked the downward trend of incidences (Fig 5). The GM (1,1) model for typhoid fever formulated cumulative incidence as (with t annotating year) and thus modelled annual cumulative incidence (per 100,000) as ((2004) at 2.54); similarly, the GM (1,1) model for paratyphoid fever formulated cumulative incidence as and thus modelled annual cumulative incidence (per 100,000) as ((2004) at baseline level at 1.26). Evaluation of the GM (1,1) performance revealed posterior error ratio at 0.17 and 0.12 respectively (both significantly smaller than the commonly used threshold of 0.35), and small error probability both higher than the commonly used criteria of 0.95 [26]. Therefore, the fitted models were statistically robust and thus may be applicable for the forecast of annual incidence in near future.

Download:

Fig 5.

Annual cumulative incidences of typhoid and paratyphoid fevers as observed (solid line) and fitted by GM (1, 1) (dotted line).

https://doi.org/10.1371/journal.pone.0241217.g005

Indeed, the established grey models predicted cumulative incidence (per 100,000) in 2017 and 2018 at 0.43 and 0.39 for typhoid fever, and at 0.09 and 0.07 for paratyphoid fever (S1 Table). Most recent government news releases reported a combined “typhoid + paratyphoid” cumulative incidence of 0.78 in 2017 and 0.78 in 2018, which were close to the sum of our predicted incidence of 0.52 cases per 100,000 in 2017 and 0.46 cases per 100,000 in 2018. Admittedly, there appeared to be a slight underestimation of both fevers by the grey models for 2017 and 2018, which could be revealed by the fitting curve falling below the actual curve in years from 2013 to 2016 (Fig 5).

Prediction of monthly incidence by SARIMA

Given the apparent seasonal pattern of monthly cumulative incidence, we conducted further time series analysis: a pseudo-forecast scenario was designed to use data from 2004 to 2015 for fitting SARIMA models and data in 2016 for validating model prediction (S2 Table). The incidence of both typhoid and paratyphoid fevers displayed apparent seasonal pattern (s = 12), and by a differential of non-seasonal (d = 1) and seasonal (D = 0) effects, the time series passed the ADF test (P < 0.05). A spectrum of parameter settings were compared in parallel in the goodness-of-fit test statistics, and an optimal model of SARIMA (0,1,7) × (1,0,1)₁₂ was chosen given its top performance (high R², low mean absolute percentage error, low AIC) for both typhoid and paratyphoid fevers (Table 2). Selection of the same SARIMA model also suggested similar seasonal and annual trends for both fevers. SARIMA (0,1,7) × (1,0,1)₁₂ fitted the cumulative incidences well for the train set of 2004–2015 period, and even the “outlier” outbreak in January 2012 was uncovered; more importantly, for the independent validation set, SARIMA-fitted cumulative incidence closely matched the actual data in 2016, demonstrating the robustness of the predictive model (Fig 6 and S2 Table). For following years of 2017 and 2018, actual monthly incidence data were not yet published, but our SARIMA (0,1,7) × (1,0,1)₁₂ model forecasted similar seasonal pattern for both fevers, which should be informative for the future prevention and control of these diseases (Fig 6).

Download:

Fig 6. Monthly cumulative incidences of typhoid and paratyphoid fevers as observed (solid line) and fitted by SARIMA (0,1,7) × (1,0,1)₁₂ (dashed line).

A shaded 95% prediction interval was shown for 2016–2018.

https://doi.org/10.1371/journal.pone.0241217.g006

Download:

Table 2. Goodness-of-fit summary of plausible SARIMA models.

https://doi.org/10.1371/journal.pone.0241217.t002

Discussion

Typhoid and paratyphoid are associated with high infectiousness, large number of hidden carriers, and heavy disease burden. The trend in seasonality potentially suggested season-related factors (including but not limited to temperature, precipitation, humidity etc.) as key environment variables associated with these fevers [28–30]. Our analysis indicated that the incidence of typhoid and paratyphoid fevers continually decreased in China, and several factors potentially contributed to this well-demonstrated example of the prevention and control of infectious disease. First, higher living condition and better hygiene (water, food, toilet etc.) prevents the infectious pathways of the bacteria; second, available vaccination, though not mandatory but still recommended, provides extra protection to individuals particularly in those high-risk regions [16, 31]. People in all ages were susceptible to both diseases (typhoid fever presented higher cumulative incidence than paratyphoid fever), and the risk is particularly higher in the young children group, in agreement with findings from previous study [32].

While susceptibility of young children may be related to their lack of hygiene, an option not to receive the non-mandatory immunization, short-lasting immunity from the vaccines even when immunized, and disapproval of the vaccines for children under 2 years of age all contribute to the risk of typhoid and paratyphoid to this most vulnerable subgroup. As a countermeasure, the WHO in 2017 prequalified a new typhoid conjugate vaccine with longer lasting immunity for use in children from the age of 6 months, and the Strategic Advisory Group of Experts (SAGE) on immunization recommended such vaccines for routine use in children over 6 months of age in typhoid endemic countries [33].

The seasonal trends of typhoid and paratyphoid were apparent and consistent throughout years with higher incidence in the summer, implicating that intensive public health efforts should be applied in the spring and summer to contain any potential outbreak and epidemic. Geographical analysis demonstrated significant disparity in the incidence of typhoid and paratyphoid among provincial regions. The cluster in the southwestern region and particularly the hotspot of Yunnan potentially suggested that exclusive local features including geology, climate, and economy may contribute to the incidence of both fevers. The highest-incidence cluster in the southwestern part of China could be a result of pathogenic bacteria growing faster under high temperature and subtropical precipitation climate [34], and suboptimal status of economic development and public health condition likely leading to water pollution and food contamination [32, 35]. Further suppression of the incidence of these fevers in China depends on the effective prevention and control actions in such hotspot regions.

The grey model and the SARIMA model, two commonly applied statistical models to fit and predict the incidence of infectious diseases, were applied in parallel on typhoid and paratyphoid in this study. Based on the grey theory, GM (1,1) model constructed a new cumulative time series from the original one, and the trends of new series could be approximated by the solution of the first-order linear differential equation [36]. Unlike some parametric models requiring large sample size and unfounded assumption of variable distributions, GM (1,1) model needs a small sample size to achieve high prediction accuracy, and is mainly distribution-independent [37]. However, the establishment of grey model requires non-fluctuating data, so GM (1,1) is not suitable for the prediction of monthly cumulative incidence with seasonal pattern. In comparison, the SARIMA model fitted the seasonal periodicity and long-term trend of time series, and iterative comparison of parameter settings would fine-tune and select the optimal model [38, 39]. The final SARIMA model fitted the long-term trend and annual seasonality for typhoid and paratyphoid fevers both for the available time range and for the near future.

To the best of our knowledge, this is the first study to explore in detail factors of age groups and regional distribution for the cumulative incidence of typhoid and paratyphoid fevers in China, and apart from descriptive analysis, statistical learning models of both GM (1,1) and SARIMA are applied to formulate the underlying incidence pattern as well as to predict the trend in a prospective manner. Such methods benefit from epidemiology-based retrospective databank and result in evidence-driven knowledge for the prevention and control of both fevers: with a general drop of incidence in recent years, typhoid and paratyphoid remain risky infectious diseases in China, and should be closely monitored as potential threats to the public health particularly in the summer season and in the high-risk southwestern regions (most notably Yunnan province).

Supporting information

S1 Fig. Cumulative incidences of typhoid and paratyphoid from 2004 to 2016 in 31 provincial regions.

https://doi.org/10.1371/journal.pone.0241217.s001

(DOCX)

S2 Fig. Geographical map of China.

https://doi.org/10.1371/journal.pone.0241217.s002

(DOCX)

S3 Fig. Cumulative incidences of typhoid and paratyphoid in four segments of China.

Note: 31 provinces were divided into 4 segments as following: Southwest (Sichuan, Chongqing, Tibet, Guangxi, Guizhou, Yunnan), Northwest (Shaanxi, Qinghai, Shanxi, Xinjiang, Ningxia, Inner Mongolia, Gansu), Northeast (Heilongjiang, Jilin, Liaoning, Beijing, Tianjin, Hebei, Anhui, Jiangsu, Shandong, Henan), and Southeast (Shanghai, Zhejiang, Jiangxi, Fujian, Hunan, Hubei, Guangdong, Hainan).

https://doi.org/10.1371/journal.pone.0241217.s003

(DOCX)

S1 Table. Comparison of actual and predicted incidence of typhoid and paratyphoid fevers by GM (1,1).

https://doi.org/10.1371/journal.pone.0241217.s004

(DOCX)

S2 Table. Comparison of actual and predicted incidence of typhoid and paratyphoid fevers in 2016 by SARIMA (0,1,7) × (1,0,1) ₁₂.

https://doi.org/10.1371/journal.pone.0241217.s005

(DOCX)

Acknowledgments

The authors would like to thank the Chinese Public Health Science Database for making the epidemiologic data on typhoid and paratyphoid publicly accessible to the academia and the civil society.

References

1. Obaro SK, Iroh Tam PY, Mintz ED. The unrecognized burden of typhoid fever. Expert Rev Vaccines. 2017;16(3):249–60. pmid:27797598
- View Article
- PubMed/NCBI
- Google Scholar
2. Mathur R, Oh H, Zhang D, Park SG, Seo J, Koblansky A, et al. A mouse model of Salmonella typhi infection. Cell. 2012;151(3):590–602. pmid:23101627
- View Article
- PubMed/NCBI
- Google Scholar
3. Crump JA, Mintz ED. Global trends in typhoid and paratyphoid Fever. Clin Infect Dis. 2010;50(2):241–6. pmid:20014951
- View Article
- PubMed/NCBI
- Google Scholar
4. Crump JA. Progress in Typhoid Fever Epidemiology. Clin Infect Dis. 2019;68(Suppl 1):S4–S9. pmid:30767000
- View Article
- PubMed/NCBI
- Google Scholar
5. Buckle GC, Walker CL, Black RE. Typhoid fever and paratyphoid fever: Systematic review to estimate global morbidity and mortality for 2010. J Glob Health. 2012;2(1):010401. pmid:23198130
- View Article
- PubMed/NCBI
- Google Scholar
6. Stanaway JD, Reiner RC, Blacker BF. The global burden of typhoid and paratyphoid fevers: a systematic analysis for the Global Burden of Disease Study 2017. The Lancet Infectious Diseases. 2019;19(4):369–81. pmid:30792131
- View Article
- PubMed/NCBI
- Google Scholar
7. Arya SC, Agarwal N. Comment on: The diagnosis of typhoid fever in the Democratic Republic of the Congo. Trans R Soc Trop Med Hyg. 2012;106(12):776; author reply 7. pmid:23062607
- View Article
- PubMed/NCBI
- Google Scholar
8. Lunguya O, Phoba MF, Mundeke SA, Bonebe E, Mukadi P, Muyembe JJ, et al. The diagnosis of typhoid fever in the Democratic Republic of the Congo. Trans R Soc Trop Med Hyg. 2012;106(6):348–55. pmid:22551639
- View Article
- PubMed/NCBI
- Google Scholar
9. Liu Z, Lao J, Zhang Y, Liu Y, Zhang J, Wang H, et al. Association between floods and typhoid fever in Yongzhou, China: Effects and vulnerable groups. Environ Res. 2018;167:718–24. pmid:30241731
- View Article
- PubMed/NCBI
- Google Scholar
10. Naheed A, Ram PK, Brooks WA, Hossain MA, Parsons MB, Talukder KA, et al. Burden of typhoid and paratyphoid fever in a densely populated urban community, Dhaka, Bangladesh. Int J Infect Dis. 2010;14 Suppl 3:e93–9. pmid:20236850
- View Article
- PubMed/NCBI
- Google Scholar
11. McMichael AJ, Woodruff RE, Hales S. Climate change and human health: present and future risks. Lancet. 2006;367(9513):859–69. pmid:16530580
- View Article
- PubMed/NCBI
- Google Scholar
12. McIver LJ, Imai C, Buettner PG, Gager P, Chan VS, Hashizume M, et al. Diarrheal Diseases and Climate Change in Cambodia. Asia Pac J Public Health. 2016;28(7):576–85. pmid:27485898
- View Article
- PubMed/NCBI
- Google Scholar
13. Liu HX, Zhang J. [Analysis of reported infectious diarrhea (other than cholera, dysentery, typhoid and paratyphoid) in China in 2011]. Zhonghua Yu Fang Yi Xue Za Zhi. 2013;47(4):328–32. pmid:23928638
- View Article
- PubMed/NCBI
- Google Scholar
14. Yang X, Zou J, Kong D, Jiang G. The analysis of GM (1, 1) grey model to predict the incidence trend of typhoid and paratyphoid fevers in Wuhan City, China. Medicine (Baltimore). 2018;97(34):e11787. pmid:30142765
- View Article
- PubMed/NCBI
- Google Scholar
15. Safaeian M, Gao YT, Sakoda LC, Quraishi SM, Rashid A, Wang BS, et al. Chronic typhoid infection and the risk of biliary tract cancer and stones in Shanghai, China. Infect Agent Cancer. 2011;6:6. pmid:21535882
- View Article
- PubMed/NCBI
- Google Scholar
16. Gu H, Yan C, Jiang Z, Li X, Chen E, Jiang J, et al. Epidemiological Trend of Typhoid and Paratyphoid Fevers in Zhejiang Province, China from 1953 to 2014. Int J Environ Res Public Health. 2018;15(11). pmid:30388758
- View Article
- PubMed/NCBI
- Google Scholar
17. Gu H, Fan W, Liu K, Qin S, Li X, Jiang J, et al. Spatio-temporal variations of typhoid and paratyphoid fevers in Zhejiang Province, China from 2005 to 2015. Sci Rep. 2017;7(1):5780. pmid:28720886
- View Article
- PubMed/NCBI
- Google Scholar
18. Zhang X, Liu Y, Yang M, Zhang T, Young AA, Li X. Comparative study of four time series methods in forecasting typhoid fever incidence in China. PLoS One. 2013;8(5):e63116. pmid:23650546
- View Article
- PubMed/NCBI
- Google Scholar
19. Olsson GE, Hjertqvist M, Lundkvist A, Hornfeldt B. Predicting high risk for human hantavirus infections, Sweden. Emerg Infect Dis. 2009;15(1):104–6. pmid:19116065
- View Article
- PubMed/NCBI
- Google Scholar
20. Zhang L, Wang L, Zheng Y, Wang K, Zhang X, Zheng Y. Time Prediction Models for Echinococcosis Based on Gray System Theory and Epidemic Dynamics. Int J Environ Res Public Health. 2017;14(3). pmid:28273856
- View Article
- PubMed/NCBI
- Google Scholar
21. Anwar MY, Lewnard JA, Parikh S, Pitzer VE. Time series analysis of malaria in Afghanistan: using ARIMA models to predict future trends in incidence. Malar J. 2016;15(1):566. pmid:27876041
- View Article
- PubMed/NCBI
- Google Scholar
22. Hui-Yu H, Hua-Qin S, Shun-Xian Z, Lin AI, Yan LU, Yu-Chun C, et al. [Application of ARIMA model to predict number of malaria cases in China]. Zhongguo Xue Xi Chong Bing Fang Zhi Za Zhi. 2017;29(4):436–40. pmid:29508575
- View Article
- PubMed/NCBI
- Google Scholar
23. Fu T, Chen T, Dong ZB, Luo SY, Miao Z, Song XP, et al. Development and comparison of forecast models of hand-foot-mouth disease with meteorological factors. Sci Rep. 2019;9(1):15691. pmid:31666565
- View Article
- PubMed/NCBI
- Google Scholar
24. Li K, Liu M, Feng Y, Ning C, Ou W, Sun J, et al. Using Baidu Search Engine to Monitor AIDS Epidemics Inform for Targeted intervention of HIV/AIDS in China. Sci Rep. 2019;9(1):320. pmid:30674890
- View Article
- PubMed/NCBI
- Google Scholar
25. Yan CQ, Wang RB, Liu HC, Jiang Y, Li MC, Yin SP, et al. [Application of ARIMA model in predicting the incidence of tuberculosis in China from 2018 to 2019]. Zhonghua Liu Xing Bing Xue Za Zhi. 2019;40(6):633–7. pmid:31238610
- View Article
- PubMed/NCBI
- Google Scholar
26. Wang YW, Shen ZZ, Jiang Y. Comparison of ARIMA and GM(1,1) models for prediction of hepatitis B in China. PLoS One. 2018;13(9):e0201987. pmid:30180159
- View Article
- PubMed/NCBI
- Google Scholar
27. Liu Q, Li Z, Ji Y, Martinez L, Zia UH, Javaid A, et al. Forecasting the seasonality and trend of pulmonary tuberculosis in Jiangsu Province of China using advanced statistical time-series analyses. Infect Drug Resist. 2019;12:2311–22. pmid:31440067
- View Article
- PubMed/NCBI
- Google Scholar
28. Kelly-Hope LA, Alonso WJ, Thiem VD, Anh DD, Canh DG, Lee H, et al. Geographical distribution and risk factors associated with enteric diseases in Vietnam. Am J Trop Med Hyg. 2007;76(4):706–12. pmid:17426175
- View Article
- PubMed/NCBI
- Google Scholar
29. Kovats RS, Edwards SJ, Hajat S, Armstrong BG, Ebi KL, Menne B. The effect of temperature on food poisoning: a time-series analysis of salmonellosis in ten European countries. Epidemiol Infect. 2004;132(3):443–53. pmid:15188714
- View Article
- PubMed/NCBI
- Google Scholar
30. Wang LX, Li XJ, Fang LQ, Wang DC, Cao WC, Kan B. Association between the incidence of typhoid and paratyphoid fever and meteorological variables in Guizhou, China. Chin Med J (Engl). 2012;125(3):455–60. pmid:22490402
- View Article
- PubMed/NCBI
- Google Scholar
31. DeRoeck D, Ochiai RL, Yang J, Anh DD, Alag V, Clemens JD. Typhoid vaccination: the Asian experience. Expert Rev Vaccines. 2008;7(5):547–60. pmid:18564010
- View Article
- PubMed/NCBI
- Google Scholar
32. Wang LX, Yang B, Yan MY, Tang YQ, Liu ZC, Wang RQ, et al. [Spatial and temporal clustering characteristics of typhoid and paratyphoid fever and its change pattern in 3 provinces in southwestern China, 2001–2012]. Zhonghua Liu Xing Bing Xue Za Zhi. 2017;38(11):1504–8. pmid:29141338
- View Article
- PubMed/NCBI
- Google Scholar
33. WHO. Newsroom.Typhoid. 2018 Jan 31 [cited 05 July 2020]. In: World Health Organization [Internet]. [about 2 screens]. Available from: https://www.who.int/news-room/fact-sheets/detail/typhoid#.
34. Gibani MM, Britto C, Pollard AJ. Typhoid and paratyphoid fever: a call to action. Current opinion in infectious diseases. 2018;31(5):440–8. pmid:30138141
- View Article
- PubMed/NCBI
- Google Scholar
35. Wang JF, Wang Y, Zhang J, Christakos G, Sun JL, Liu X, et al. Spatiotemporal transmission and determinants of typhoid and paratyphoid fever in Hongta District, Yunnan Province, China. PLoS Negl Trop Dis. 2013;7(3):e2112. pmid:23516653
- View Article
- PubMed/NCBI
- Google Scholar
36. Mahmod WE, Watanabe K. Modified Grey Model and its application to groundwater flow analysis with limited hydrogeological data: a case study of the Nubian Sandstone, Kharga Oasis, Egypt. Environ Monit Assess. 2014;186(2):1063–81. pmid:24092253
- View Article
- PubMed/NCBI
- Google Scholar
37. Lee CC, Wan TJ, Kuo CY, Chung CY. Modified grey model for estimating traffic tunnel air quality. Environ Monit Assess. 2007;132(1–3):351–64. pmid:17342440
- View Article
- PubMed/NCBI
- Google Scholar
38. Liu L, Luan RS, Yin F, Zhu XP, Lu Q. Predicting the incidence of hand, foot and mouth disease in Sichuan province, China using the ARIMA model. Epidemiol Infect. 2016;144(1):144–51. pmid:26027606
- View Article
- PubMed/NCBI
- Google Scholar
39. Earnest A, Evans SM, Sampurno F, Millar J. Forecasting annual incidence and mortality rate for prostate cancer in Australia until 2022 using autoregressive integrated moving average (ARIMA) models. BMJ Open. 2019;9(8):e031331. pmid:31431447
- View Article
- PubMed/NCBI
- Google Scholar

[ref1] 1. Obaro SK, Iroh Tam PY, Mintz ED. The unrecognized burden of typhoid fever. Expert Rev Vaccines. 2017;16(3):249–60. pmid:27797598
View Article
PubMed/NCBI
Google Scholar

[2] View Article

[3] PubMed/NCBI

[4] Google Scholar

[ref2] 2. Mathur R, Oh H, Zhang D, Park SG, Seo J, Koblansky A, et al. A mouse model of Salmonella typhi infection. Cell. 2012;151(3):590–602. pmid:23101627
View Article
PubMed/NCBI
Google Scholar

[6] View Article

[7] PubMed/NCBI

[8] Google Scholar

[ref3] 3. Crump JA, Mintz ED. Global trends in typhoid and paratyphoid Fever. Clin Infect Dis. 2010;50(2):241–6. pmid:20014951
View Article
PubMed/NCBI
Google Scholar

[10] View Article

[11] PubMed/NCBI

[12] Google Scholar

[ref4] 4. Crump JA. Progress in Typhoid Fever Epidemiology. Clin Infect Dis. 2019;68(Suppl 1):S4–S9. pmid:30767000
View Article
PubMed/NCBI
Google Scholar

[14] View Article

[15] PubMed/NCBI

[16] Google Scholar

[ref5] 5. Buckle GC, Walker CL, Black RE. Typhoid fever and paratyphoid fever: Systematic review to estimate global morbidity and mortality for 2010. J Glob Health. 2012;2(1):010401. pmid:23198130
View Article
PubMed/NCBI
Google Scholar

[18] View Article

[19] PubMed/NCBI

[20] Google Scholar

[ref6] 6. Stanaway JD, Reiner RC, Blacker BF. The global burden of typhoid and paratyphoid fevers: a systematic analysis for the Global Burden of Disease Study 2017. The Lancet Infectious Diseases. 2019;19(4):369–81. pmid:30792131
View Article
PubMed/NCBI
Google Scholar

[22] View Article

[23] PubMed/NCBI

[24] Google Scholar

[ref7] 7. Arya SC, Agarwal N. Comment on: The diagnosis of typhoid fever in the Democratic Republic of the Congo. Trans R Soc Trop Med Hyg. 2012;106(12):776; author reply 7. pmid:23062607
View Article
PubMed/NCBI
Google Scholar

[26] View Article

[27] PubMed/NCBI

[28] Google Scholar

[ref8] 8. Lunguya O, Phoba MF, Mundeke SA, Bonebe E, Mukadi P, Muyembe JJ, et al. The diagnosis of typhoid fever in the Democratic Republic of the Congo. Trans R Soc Trop Med Hyg. 2012;106(6):348–55. pmid:22551639
View Article
PubMed/NCBI
Google Scholar

[30] View Article

[31] PubMed/NCBI

[32] Google Scholar

[ref9] 9. Liu Z, Lao J, Zhang Y, Liu Y, Zhang J, Wang H, et al. Association between floods and typhoid fever in Yongzhou, China: Effects and vulnerable groups. Environ Res. 2018;167:718–24. pmid:30241731
View Article
PubMed/NCBI
Google Scholar

[34] View Article

[35] PubMed/NCBI

[36] Google Scholar

[ref10] 10. Naheed A, Ram PK, Brooks WA, Hossain MA, Parsons MB, Talukder KA, et al. Burden of typhoid and paratyphoid fever in a densely populated urban community, Dhaka, Bangladesh. Int J Infect Dis. 2010;14 Suppl 3:e93–9. pmid:20236850
View Article
PubMed/NCBI
Google Scholar

[38] View Article

[39] PubMed/NCBI

[40] Google Scholar

[ref11] 11. McMichael AJ, Woodruff RE, Hales S. Climate change and human health: present and future risks. Lancet. 2006;367(9513):859–69. pmid:16530580
View Article
PubMed/NCBI
Google Scholar

[42] View Article

[43] PubMed/NCBI

[44] Google Scholar

[ref12] 12. McIver LJ, Imai C, Buettner PG, Gager P, Chan VS, Hashizume M, et al. Diarrheal Diseases and Climate Change in Cambodia. Asia Pac J Public Health. 2016;28(7):576–85. pmid:27485898
View Article
PubMed/NCBI
Google Scholar

[46] View Article

[47] PubMed/NCBI

[48] Google Scholar

[ref13] 13. Liu HX, Zhang J. [Analysis of reported infectious diarrhea (other than cholera, dysentery, typhoid and paratyphoid) in China in 2011]. Zhonghua Yu Fang Yi Xue Za Zhi. 2013;47(4):328–32. pmid:23928638
View Article
PubMed/NCBI
Google Scholar

[50] View Article

[51] PubMed/NCBI

[52] Google Scholar

[ref14] 14. Yang X, Zou J, Kong D, Jiang G. The analysis of GM (1, 1) grey model to predict the incidence trend of typhoid and paratyphoid fevers in Wuhan City, China. Medicine (Baltimore). 2018;97(34):e11787. pmid:30142765
View Article
PubMed/NCBI
Google Scholar

[54] View Article

[55] PubMed/NCBI

[56] Google Scholar

[ref15] 15. Safaeian M, Gao YT, Sakoda LC, Quraishi SM, Rashid A, Wang BS, et al. Chronic typhoid infection and the risk of biliary tract cancer and stones in Shanghai, China. Infect Agent Cancer. 2011;6:6. pmid:21535882
View Article
PubMed/NCBI
Google Scholar

[58] View Article

[59] PubMed/NCBI

[60] Google Scholar

[ref16] 16. Gu H, Yan C, Jiang Z, Li X, Chen E, Jiang J, et al. Epidemiological Trend of Typhoid and Paratyphoid Fevers in Zhejiang Province, China from 1953 to 2014. Int J Environ Res Public Health. 2018;15(11). pmid:30388758
View Article
PubMed/NCBI
Google Scholar

[62] View Article

[63] PubMed/NCBI

[64] Google Scholar

[ref17] 17. Gu H, Fan W, Liu K, Qin S, Li X, Jiang J, et al. Spatio-temporal variations of typhoid and paratyphoid fevers in Zhejiang Province, China from 2005 to 2015. Sci Rep. 2017;7(1):5780. pmid:28720886
View Article
PubMed/NCBI
Google Scholar

[66] View Article

[67] PubMed/NCBI

[68] Google Scholar

[ref18] 18. Zhang X, Liu Y, Yang M, Zhang T, Young AA, Li X. Comparative study of four time series methods in forecasting typhoid fever incidence in China. PLoS One. 2013;8(5):e63116. pmid:23650546
View Article
PubMed/NCBI
Google Scholar

[70] View Article

[71] PubMed/NCBI

[72] Google Scholar

[ref19] 19. Olsson GE, Hjertqvist M, Lundkvist A, Hornfeldt B. Predicting high risk for human hantavirus infections, Sweden. Emerg Infect Dis. 2009;15(1):104–6. pmid:19116065
View Article
PubMed/NCBI
Google Scholar

[74] View Article

[75] PubMed/NCBI

[76] Google Scholar

[ref20] 20. Zhang L, Wang L, Zheng Y, Wang K, Zhang X, Zheng Y. Time Prediction Models for Echinococcosis Based on Gray System Theory and Epidemic Dynamics. Int J Environ Res Public Health. 2017;14(3). pmid:28273856
View Article
PubMed/NCBI
Google Scholar

[78] View Article

[79] PubMed/NCBI

[80] Google Scholar

[ref21] 21. Anwar MY, Lewnard JA, Parikh S, Pitzer VE. Time series analysis of malaria in Afghanistan: using ARIMA models to predict future trends in incidence. Malar J. 2016;15(1):566. pmid:27876041
View Article
PubMed/NCBI
Google Scholar

[82] View Article

[83] PubMed/NCBI

[84] Google Scholar

[ref22] 22. Hui-Yu H, Hua-Qin S, Shun-Xian Z, Lin AI, Yan LU, Yu-Chun C, et al. [Application of ARIMA model to predict number of malaria cases in China]. Zhongguo Xue Xi Chong Bing Fang Zhi Za Zhi. 2017;29(4):436–40. pmid:29508575
View Article
PubMed/NCBI
Google Scholar

[86] View Article

[87] PubMed/NCBI

[88] Google Scholar

[ref23] 23. Fu T, Chen T, Dong ZB, Luo SY, Miao Z, Song XP, et al. Development and comparison of forecast models of hand-foot-mouth disease with meteorological factors. Sci Rep. 2019;9(1):15691. pmid:31666565
View Article
PubMed/NCBI
Google Scholar

[90] View Article

[91] PubMed/NCBI

[92] Google Scholar

[ref24] 24. Li K, Liu M, Feng Y, Ning C, Ou W, Sun J, et al. Using Baidu Search Engine to Monitor AIDS Epidemics Inform for Targeted intervention of HIV/AIDS in China. Sci Rep. 2019;9(1):320. pmid:30674890
View Article
PubMed/NCBI
Google Scholar

[94] View Article

[95] PubMed/NCBI

[96] Google Scholar

[ref25] 25. Yan CQ, Wang RB, Liu HC, Jiang Y, Li MC, Yin SP, et al. [Application of ARIMA model in predicting the incidence of tuberculosis in China from 2018 to 2019]. Zhonghua Liu Xing Bing Xue Za Zhi. 2019;40(6):633–7. pmid:31238610
View Article
PubMed/NCBI
Google Scholar

[98] View Article

[99] PubMed/NCBI

[100] Google Scholar

[ref26] 26. Wang YW, Shen ZZ, Jiang Y. Comparison of ARIMA and GM(1,1) models for prediction of hepatitis B in China. PLoS One. 2018;13(9):e0201987. pmid:30180159
View Article
PubMed/NCBI
Google Scholar

[102] View Article

[103] PubMed/NCBI

[104] Google Scholar

[ref27] 27. Liu Q, Li Z, Ji Y, Martinez L, Zia UH, Javaid A, et al. Forecasting the seasonality and trend of pulmonary tuberculosis in Jiangsu Province of China using advanced statistical time-series analyses. Infect Drug Resist. 2019;12:2311–22. pmid:31440067
View Article
PubMed/NCBI
Google Scholar

[106] View Article

[107] PubMed/NCBI

[108] Google Scholar

[ref28] 28. Kelly-Hope LA, Alonso WJ, Thiem VD, Anh DD, Canh DG, Lee H, et al. Geographical distribution and risk factors associated with enteric diseases in Vietnam. Am J Trop Med Hyg. 2007;76(4):706–12. pmid:17426175
View Article
PubMed/NCBI
Google Scholar

[110] View Article

[111] PubMed/NCBI

[112] Google Scholar

[ref29] 29. Kovats RS, Edwards SJ, Hajat S, Armstrong BG, Ebi KL, Menne B. The effect of temperature on food poisoning: a time-series analysis of salmonellosis in ten European countries. Epidemiol Infect. 2004;132(3):443–53. pmid:15188714
View Article
PubMed/NCBI
Google Scholar

[114] View Article

[115] PubMed/NCBI

[116] Google Scholar

[ref30] 30. Wang LX, Li XJ, Fang LQ, Wang DC, Cao WC, Kan B. Association between the incidence of typhoid and paratyphoid fever and meteorological variables in Guizhou, China. Chin Med J (Engl). 2012;125(3):455–60. pmid:22490402
View Article
PubMed/NCBI
Google Scholar

[118] View Article

[119] PubMed/NCBI

[120] Google Scholar

[ref31] 31. DeRoeck D, Ochiai RL, Yang J, Anh DD, Alag V, Clemens JD. Typhoid vaccination: the Asian experience. Expert Rev Vaccines. 2008;7(5):547–60. pmid:18564010
View Article
PubMed/NCBI
Google Scholar

[122] View Article

[123] PubMed/NCBI

[124] Google Scholar

[ref32] 32. Wang LX, Yang B, Yan MY, Tang YQ, Liu ZC, Wang RQ, et al. [Spatial and temporal clustering characteristics of typhoid and paratyphoid fever and its change pattern in 3 provinces in southwestern China, 2001–2012]. Zhonghua Liu Xing Bing Xue Za Zhi. 2017;38(11):1504–8. pmid:29141338
View Article
PubMed/NCBI
Google Scholar

[126] View Article

[127] PubMed/NCBI

[128] Google Scholar

[ref33] 33. WHO. Newsroom.Typhoid. 2018 Jan 31 [cited 05 July 2020]. In: World Health Organization [Internet]. [about 2 screens]. Available from: https://www.who.int/news-room/fact-sheets/detail/typhoid#.

[ref34] 34. Gibani MM, Britto C, Pollard AJ. Typhoid and paratyphoid fever: a call to action. Current opinion in infectious diseases. 2018;31(5):440–8. pmid:30138141
View Article
PubMed/NCBI
Google Scholar

[131] View Article

[132] PubMed/NCBI

[133] Google Scholar

[ref35] 35. Wang JF, Wang Y, Zhang J, Christakos G, Sun JL, Liu X, et al. Spatiotemporal transmission and determinants of typhoid and paratyphoid fever in Hongta District, Yunnan Province, China. PLoS Negl Trop Dis. 2013;7(3):e2112. pmid:23516653
View Article
PubMed/NCBI
Google Scholar

[135] View Article

[136] PubMed/NCBI

[137] Google Scholar

[ref36] 36. Mahmod WE, Watanabe K. Modified Grey Model and its application to groundwater flow analysis with limited hydrogeological data: a case study of the Nubian Sandstone, Kharga Oasis, Egypt. Environ Monit Assess. 2014;186(2):1063–81. pmid:24092253
View Article
PubMed/NCBI
Google Scholar

[139] View Article

[140] PubMed/NCBI

[141] Google Scholar

[ref37] 37. Lee CC, Wan TJ, Kuo CY, Chung CY. Modified grey model for estimating traffic tunnel air quality. Environ Monit Assess. 2007;132(1–3):351–64. pmid:17342440
View Article
PubMed/NCBI
Google Scholar

[143] View Article

[144] PubMed/NCBI

[145] Google Scholar

[ref38] 38. Liu L, Luan RS, Yin F, Zhu XP, Lu Q. Predicting the incidence of hand, foot and mouth disease in Sichuan province, China using the ARIMA model. Epidemiol Infect. 2016;144(1):144–51. pmid:26027606
View Article
PubMed/NCBI
Google Scholar

[147] View Article

[148] PubMed/NCBI

[149] Google Scholar

[ref39] 39. Earnest A, Evans SM, Sampurno F, Millar J. Forecasting annual incidence and mortality rate for prostate cancer in Australia until 2022 using autoregressive integrated moving average (ARIMA) models. BMJ Open. 2019;9(8):e031331. pmid:31431447
View Article
PubMed/NCBI
Google Scholar

[151] View Article

[152] PubMed/NCBI

[153] Google Scholar

Figures

Abstract

Introduction

Methods

Data sources

Time series models

Data analysis

Results

Typhoid and paratyphoid fevers in China

Age distribution of typhoid and paratyphoid fevers

Geographical cluster of typhoid and paratyphoid fevers

Prediction of annual incidence by GM (1, 1)

Prediction of monthly incidence by SARIMA

Discussion

Supporting information

S1 Fig. Cumulative incidences of typhoid and paratyphoid from 2004 to 2016 in 31 provincial regions.

S2 Fig. Geographical map of China.

S3 Fig. Cumulative incidences of typhoid and paratyphoid in four segments of China.

S1 Table. Comparison of actual and predicted incidence of typhoid and paratyphoid fevers by GM (1,1).

S2 Table. Comparison of actual and predicted incidence of typhoid and paratyphoid fevers in 2016 by SARIMA (0,1,7) × (1,0,1) 12.

Acknowledgments

References

S2 Table. Comparison of actual and predicted incidence of typhoid and paratyphoid fevers in 2016 by SARIMA (0,1,7) × (1,0,1) ₁₂.