Impact of epidemic prevention measures on the incidence of HFMD during the COVID-19 epidemic period in Beijing

doi:10.21203/rs.3.rs-3147598/v1

Download PDF

Article

Impact of epidemic prevention measures on the incidence of HFMD during the COVID-19 epidemic period in Beijing

https://doi.org/10.21203/rs.3.rs-3147598/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

The non-pharmacological interventions during the COVID-19 epidemic period may also affected the incidence of other infectious diseases. This study used the autoregressive integrated moving average (ARIMA) model to predict the theoretical incidence of hand-foot-and-mouth disease (HFMD) in Beijing in 2020, then used the Wilcoxon-Mann-Whitney test to compare it with the actual value, in order to explore the impact of public health prevention measure on the incidence of HFMD during the COVID-19 epidemic period. Between January 2013 and December 2019, 210,101 cases of HFMD were reported in Beijing. The number of reported cases in Beijing in 2020 decreased by 92.7% compared to the average annual number of reported cases from 2013 to 2019. The optimal model (1,1,1) (1,1,1) ₁₂ established in this study predicted the monthly incidence of HFMD in 2020, the model-fitting effect was good, the test results showing that there was a statistically significant difference between the predicted and actual number of HFMD cases in 2020. Epidemic prevention measures during the COVID-19 epidemic period had a great impact on the incidence of HFMD in Beijing.

Health sciences/Diseases

Health sciences/Health care

Since the beginning of the COVID-19 pandemic, human life and work have been greatly affected. To control the prevalence of COVID-19, Beijing implemented a series of non-pharmacological interventions (NPIs) such as community closure management, online teaching, and closure of large public places. Health education for citizens included wearing masks, washing hands frequently, and providing more ventilation ^[1][2]. Multiple studies have shown that in the early stages of an epidemic, without vaccines or targeted treatment drugs, joint NPIs can effectively reduce the number of cases ^[3][4]. Studies have shown that these NPIs also affect the incidence of other diseases with similar transmission routes ^[5][6].

Hand-foot-and-mouth disease (HFMD) is a common infectious disease caused by various intestinal viruses and is mainly transmitted through respiratory droplets and/or oral fecal routes. From 2010 to 2019, the number of reported cases of HFMD in Beijing showed a cyclical fluctuation over two years, with a high incidence in even years and a low incidence in odd years^[7]. The transmission mode of COVID-19 is similar to HFMD, mainly transmitted through respiratory droplets, aerosols, close contact, and other means. Previous studies used interrupted time-series segmented regression analysis and a Bayesian structural time series model to analyze the impact of NPIs on the incidence of HFMD in other cities during the COVID-19 epidemic^[8][9]. Due to the cyclical and seasonal characteristics of HFMD, this study applied the seasonal autoregressive integrated moving average (SARIMA) model to predict the theoretical incidence of HFMD in Beijing in 2020 and then compared it with the actual value to explore the impact of NPIs on the incidence of HFMD during the COVID-19 epidemic period.

The epidemic situation of HFMD in Beijing

From January 2013 to December 2019, 210,101 cases of HFMD were reported in Beijing, with two epidemic peaks each year, a major peak from May to August and a minor peak from September to November. In addition, there was a cyclical fluctuation over two years, with high occurrence in even years and low occurrence in odd years. The seasonal distribution patterns were significant. The average number of reported cases from 2017 to 2019 was 30,014 and 2,172 cases were reported in 2020, a decrease of 92.7% compared with the average number of reported cases from 2013 to 2019. Only one epidemic peak occurred in 2020.

ARIMA model construction:

(1) Data pre-processing and model recognition. The original series showed obvious trends and seasonal changes and needed to stabilize. After natural logarithmic transformation and seasonal splitting, the series stabilized (Fig. 1).

Parameter estimation and model diagnosis. According to the times of the splitting transformation, it was preliminarily determined that the form of the seasonal model is (p,1,q)(P,1,Q)₁₂. We drew ACF and PACF graphs for the split sequence (Fig. 2), enumerated all models with p and q values between 0 and 2, and selected the five best models for comparison; the residual errors of the five models were white noise (Table 1). The AIC and BIC values of the five models were compared and the model with the lowest AIC and BIC values was selected as the optimal ARIMA model. Based on this information, the optimal model was determined to be (1,1,1) (1,1,1) ₁₂.

Table 1

Parameter estimation and testing of alternative models
Model	AIC	BIC	L-B residual P-value
(1,1,0)(0,1,1)₁₂	73.757	78.282	0.6049
(0,1,1)(0,1,1)₁₂	69.744	74.269	0.6576
(2,1,1)(1,1,1)₁₂	72.383	81.434	0.3583
(0,1,1)(1,1,1)₁₂	70.773	77.561	0.5649
(1,1,1)(0,1,1)₁₂	71.430	78.218	0.4701
(1,1,1)(1,1,1)₁₂	72.055	81.106	0.3323

(3) Evaluation of model fitting effect and prediction.

Using the optimal model for prediction analysis, the actual incidence of HFMD in Beijing from 2013 to 2019 was used as the original sequence and the predicted incidence of HFMD was used as the prediction sequence for model fitting. The results are shown in Fig. 3, and indicate a high degree of agreement between the predicted and actual values of the model. The model was used to predict the incidence of HFMD in Beijing in 2020. All fitting values were within the 95% confidence interval and the model-fitting effect was good.

(4) Difference test

The predicted number of HFMD cases in 2020 based on the optimal model was 23,276 (the Lower limit of confidence interval (LCL) and Upper limit of confidence interval (UCL) predicted values were 7,752 and 70,055, respectively). The actual number was 2,172. Because the actual value group data did not conform to the normal distribution (W = 0.765, P = 0.039), the nonparametric Wilcoxon rank sum test was used to compare the difference between the predictive and actual values of HFMD monthly incidence in 2020. The results showed that the difference between the two groups was statistically significant (Z = 3.666, P = 0.0002).

The time-series model is based on the past epidemic situation of a disease to predict the future. This study predicted the incidence of HFMD in Beijing in 2020 without COVID-19 and NPIs, all of which fell within the 95% confidence interval, indicating a good fit and high prediction accuracy. However, in January 2020, the COVID-19 epidemic occurred in Beijing and a series of NPIs were implemented^[11], resulting in an extremely low number of HFMD cases by 2020, with only 9.3% of the predicted number of cases throughout the year.

In 2020, COVID-19 broke out nationwide, and different cities faced various prevention and control problems and difficulties. In the early stages of an epidemic that lacks targeted preventive vaccines and drugs, NPIs are mainly used throughout the country to suppress the spread of the epidemic^[12]; specifically, they include traditional infectious disease prevention and control measures, such as controlling the source of infection and cutting off the transmission routes. The implementation of these measures also affects the incidence rate of other infectious diseases. In 2020, COVID-19 in Beijing experienced three distinct stages: a domestic epidemic stage, in which cases were mainly imported from Wuhan; outbound import epidemic stage; and a local outbreak stage^[13]. In the first half of 2020, offline teaching was suspended in Beijing, and primary and secondary schools and kindergartens opened successively in September^[14]. Immediately thereafter, it ushered in only a small peak in HFMD incidence throughout the year. In a study by Geng^[15], the national epidemic of HFMD was interrupted between February and August 2020 and rebounded after schools and kindergartens reopened in September, which is consistent with the results of this study.

HFMD and COVID-19 can be transmitted through close contact, respiratory tract, and other means. During epidemic prevention, China focused on strengthening intervention measures such as wearing masks, washing hands frequently, more ventilation, and less gathering, among others, while gradually cultivating good hygiene practices to inhibit the spread of HFMD. In June 2020, Beijing entered the stage of local cluster epidemic prevention and control, focusing mainly on the Beijing Xinfadi Market ^[13]. In previous years, June was the peak season for HFMD outbreaks in Beijing. At this time, due to the cluster epidemic in the local area, an emergency response to major public health emergencies was initiated, and prevention and control measures were tightened. Measures such as school and park closures, closed public places, and travel restrictions were implemented which greatly reduced opportunities for everyone to have contact and reduced the incidence of HFMD, while also avoiding the occurrence of clustered outbreaks of HFMD in schools and communities. The NPIs used to prevent COVID-19 can also significantly reduce the incidence of HFMD. In the future, good personal hygiene practices should be advocated in schools, kindergartens, and other key units.

This study had some limitations. Because HFMD is a self-limiting disease, patients are concerned about travel and medical treatment during the COVID-19 pandemic, causing fewer HFMD cases to be discovered, may resulting in a lower reported number than the actual number of infections.

Ethics statement

This study was based on official HFMD surveillance data in Beijing, all data were collected from the China Information System for Disease Control and Prevention. Analyses were conducted at aggregate level, all data employed in this study were deidentified prior to analysis and anonymized confidentiality, no personal information was involved. The study was approved by the ethics committee of Beijing Center for Disease Control and Prevention.

Data collection

Monitoring data for HFMD in Beijing from January 2013 to December 2020 were obtained from the Infectious Disease Report Information Management System of the National Notifiable Disease Reporting System (NNDRS). The monthly incidence data of HFMD from January 2013 to December 2019 were used to fit the autoregressive integrated moving average (ARIMA) model, and data from January to December 2020 were used for difference testing and analysis.

Time series analysis

Time series are numerical sequences formed at time points, and time-series predictions predict future values by observing historical data. SARIMA is a seasonal autoregressive integrated moving average model that combines seasonal differences with the ARIMA model and is used for the data modeling of time series with periodic characteristics. SARIMA (P, D, Q) (P, D, Q)_s is divided into two parts: a non-seasonal model with parameters P, D, Q, and a seasonal model with parameters P, D, and Q, where s is the period length, and the formula for SARIMA(p, d, q)(P, D, Q)s is as follows^[10] :

ϕ(B)Ф(B^s)(1-B)^d(1-B)^Dsyt = θ(B)θ(B^s) ε_t

Here, B signifies the backward shift operator, and ε_t denotes the residual. Other variables are as follows: ϕ(B) = 1-ϕ₁B-…-ϕpB^p; θ(B) = 1-θ₁B-…θqB^q; Ф(B^s) = 1-Ф₁B^s-…-ФpB^Ps; θ(B^s) = 1-θ₁B^s-…-θ_QB^Qs

SAS9.4 software was used to model, analyze, and predict data. A P < 0.05 was considered as statistically significant. Establishing an ARIMA model requires time series stationarity. If the original sequence was not stationary, data transformation or differences were used to render the data stationary. Draw autocorrelation function (ACF) and partial autocorrelation function (PACF) graphs for the sequences that meet the stationarity requirements and estimate the p,q values. Estimation of the parameters P and Q values can be attempted step-by-step from low to high order and generally does not exceed 2. The selected parameters were introduced into the SARIMA model for estimation and the residual sequence of the model was tested using white noise. If the statistics show that the difference is not statistically significant and the residual is white noise, the best SARIMA model is selected by combining the Akaike information criterion (AIC) and Bayesian information criterion (BIC) values. We fitted the data from January 2013 to December 2019 using the optimal model and predicted the monthly incidence of HFMD in Beijing in 2020.

Funding: Supported by High level public health technical talent training plan: 2022-3-019; and Beijing Natural Science Foundation(7202073); and Key research projects of Beijing Natural Science Foundation-Haidian District Joint Fund(L192012); and National Key R&D Program of China (2021ZD0114103); and Capital's Funds for Health Improvement and Research (2022-1G-3014).

Data availability

The data are available within the paper.

Author information

Authors and Affiliations

Institute for Infectious Disease and Endemic Disease Control, Beijing Center for Disease Prevention and Control, Beijing, China

Hao Zhao, Shuaibing Dong, Peng Yang, Daitao Zhang, Quanyi Wang, Lei Jia & Da Huo

School of Public Health, Capital Medical University, Beijing, China

Quanyi Wang & Da Huo

Beijing Municipal Key Laboratory of Clinical Epidemiology, Beijing, China

Quanyi Wang & Da Huo

Beijing Office of Center for Global Health, Beijing Center for Disease Prevention and Control, Beijing, China

Xiaoli Wang & Peng Yang

Contributions

H.Z. wrote the first draft of the manuscript and prepared the figures and table. H.Z., SB.D. contributed to the acquisition of data and analyzed the data. H.Z., XL.W., P.Y. and D.H. designed the study, XL.W., P.Y., QY.W., DT.Z., L.J. and D.H.contributed to the editing of the manuscript. All authors read and contributed to the final version of the manuscript.

Corresponding author

Correspondence to Lei Jia & Da Huo.

Ethics declarations

Competing interests

The authors declare no competing interests.

Beijing Municipal Commission of Health. Work Plan for Community (Village) Prevention and Control of Pneumonia Infected by novel coronavirus in Beijing (Trial).http://wjw.beijing.gov.cn/wjwh/ztzl/xxgzbd/gzbdzcfg/202002/t20200201_1622571.html(2020).
Beijing Municipal Education Commission. Notice on Further Improving the Work of Delaying the Opening of Primary and Secondary Schools in the Spring Semester of 2020. http://jw.beijing.gov.cn/jyzx/ztzl/yq_2020/yq_sdjy/202004/t20200415_1804035.html(2020).
Lai S. et al. Effect of non-pharmaceutical interventions to contain COVID-19 in China. Nature. 585(7825):410-413. https://doi.org/10.1038/s41586-020-2293-x(2020).
Kraemer MUG. et al. The effect of human mobility and control measures on the COVID-19 epidemic in China. medRxiv [Preprint]. 368(6490):493-497. https://doi.org/10.1101/2020.03.02.20026708(2020).
Huang QS. et al. Impact of the COVID-19 nonpharmaceutical interventions on influenza and other respiratory viral infections in New Zealand. Nature communications. 12(1):1001. https://doi.org/10.1038/s41467-021-21157-9(2021).
Ndeh NT, Tesfaldet YT, Budnard J, Chuaicharoen P. The secondary outcome of public health measures amidst the COVID-19 pandemic in the spread of other respiratory infectious diseases in Thailand. Travel Med Infect Dis. 48:102348. https://doi.org/10.1016/j.tmaid.2022.102348(2022).
Zhao H. et al. Epidemiological characteristics and spatiotemporal clustering analysis of HFMD in Beijing from 2017 to 2019. International Journal of virology.29 (06): 461-466(2022).
Wu K. et al. Effects of different levels of non-pharmaceutical interventions on hand, foot and mouth disease in Guangzhou, China. BMC Public Health. 22(1):2398. https://doi.org/10.1186/s12889-022-14850-x(2022).
Shen L. et al. The impact of anti-COVID-19 nonpharmaceutical interventions on hand, foot, and mouth disease-A spatiotemporal perspective in Xi'an, northwestern China. J Med Virol. 94(7):3121-3132. https://doi.org/10.1002/jmv.27715(2022).
Yu, G., Feng, H., Feng, S., Zhao, J., & Xu, J. Forecasting hand-foot-and-mouth disease cases using wavelet-based SARIMA-NNAR hybrid model. PLoS One. 16(2), e0246673. https://doi.org/10.1371/journal.pone.0246673(2021).
Beijing Municipal Health Commission. A letter from the Beijing Center for Disease Control and Prevention to the general public. http://wjw.beijing.gov.cn/wjwh/ztzl/xxgzbd/gzbdfkgz/202002/t20200223_1666522.html(2020).
The Central People's Government of the People's Republic of China. Local governments take measures to prevent and control pneumonia caused by novel coronavirus. https://www.gov.cn/xinwen/2020-01/26/content_5472338.htm#3(2020).
Ma CN. et al. Analysis of epidemic characteristics of novel coronavirus pneumonia at different stages in Beijing from January to July 2020. International Journal of virology. 27(06):448-454(2020).
Beijing Municipal Commission of Education. Announcement of the New Semester Epidemic Prevention Work Arrangements for Schools at All Levels and Types in Beijing. https://www.beijing.gov.cn/fuwu/bmfw/jyfw/ggts/202008/t20200814_1983941.html(2020).
Geng MJ, Zhang HY, Yu LJ, et al. Changes in notifiable infectious disease incidence in China during the COVID-19 pandemic. Nat Commun. 12(1):6923. https://doi.org/10.1038/s41467-021-27292-7(2021).

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Impact of epidemic prevention measures on the incidence of HFMD during the COVID-19 epidemic period in Beijing

Status:

Version 1

Abstract

Figures

Introduction

Results

The epidemic situation of HFMD in Beijing

ARIMA model construction:

Discussion

Methods

Ethics statement

Data collection

Time series analysis

Declarations

References

Additional Declarations

Status:

Version 1