Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Trend and forecasting of the COVID-19 outbreak in China

Trend and forecasting of the COVID-19 outbreak in China Trend and forecasting of the COVID-19 outbreak in China 1 2 Qiang Li Wei Feng School of Physical science and Technology, Northwestern Polytechnical University, Xian, 710129, P. R. China. School of Electronic Engineering, Xidian University, Xian, 710071, P. R. China. Abstract: By using the public data from Jan. 20 to Feb. 11, 2020, we perform data-driven analysis and forecasting on the COVID-19 epidemic in mainland China, especially Hubei province. Our results show that the turning points of the daily infections are predicted to be Feb. 6 and Feb. 1, 2020, for Hubei and China other than Hubei, respectively. The epidemic in China is predicted to end up after Mar. 10, 2020, and the number of the total infections are predicted to be 51600. The data trends reveal that quick and active strategies taken by China to reduce human exposure have already had a good impact on the control of the epidemic. liruo@nwpu.edu.cn wfeng@xidian.edu.cn arXiv:2002.05866v1 [q-bio.PE] 14 Feb 2020 Contents 1 Introduction 1 2 Data sources 2 3 Methods 3 4 Results 5 4.1 Predictions of the epidemic in Hubei 5 4.2 Predictions of the epidemic in China other than Hubei 6 5 Summary and discussions 7 1 Introduction China, especially the Hubei province, is ghting the pneumonia epidemic by implementing various prevention and control measures [1]. Pneumonia, caused by the novel coronavirus (2019-nCoV), which may originate from the bat [2], is just named as the COVID-19 by the World Health Organization (WHO). The COVID-19 outbroke from Wuhan, the capital of Hubei province in China in Dec. 2019, and has spread to other provinces of China and even other countries [3]. Strong human-to-human transmission is established [4], and there exists the risk of transportation of COVID-19 from Wuhan to 369 other cities in China [5]. Until Feb. 11, 2020, there have been 44653 cases of COVID-19 infections con rmed in mainland China, including 1113 deaths. To prevent and control the spread of the epidemic, many strategies are needed [6], and China has already taken quick and e ective strategies to reduce population mobility and interpersonal contact rates, and also increase quarantine on migrants. Predicting the trend of the epidemic are quite important to the allocation of medical resources, the arrangement of production activities, and even the domestic economic development all over China. In recent decades, two other new coronaviruses,Severe Acute Respiratory Syndrome coronavirus (SARS-CoV)" and \Middle East respiratory syndrome coronavirus (MERS- CoV)", have been considered to be major epidemics worldwide. However, concerning these viruses, COVID-19 presents a higher degree of uncertainty in the scale and geographical scope of the outbreak within and outside China. SARS-CoV and MERS-CoV-based prop- agation analysis and prediction models may be no longer suitable in the ght against the COVID-19 pneumonia. Therefore, it is very urgent to use the latest data to establish an ecient and highly suitable epidemic analysis and prediction model according to the ac- tual situation, and then to give reliable epidemic predictions. This work could provide an important reference for the government to formulate emergency macroeconomic decisions { 1 { and medical resources allocation. Moreover, the signi cance is also of great reference value for the country's deployment and adjustment of economic activities throughout the year Very recently, the susceptible-exposed-infectious-recovered (SEIR) model is used to forecast the potential domestic and international spread of this COVID-19 outbreak [7] with parameters estimated from other sources. The e ective daily reproduction ratio and epidemics peak are also predicted in a Ref. [8]. The real situation could be much more complicated and changing all the time. Especially, with the implementation of the Chinese government's multiple epidemic control policies, the control of nationwide epidemic has become obvious. However, the medical supplies in Hubei will still a ect the implementation of national policies. In this study, the models driven by data are used to describe the current data of the epidemic, to predict the ongoing trend, and to estimate the outbreak size of the COVID-19 in both Hubei and other areas in mainland China. This data-driven study could present another prediction to the epidemic compared with the susceptible-infectious- recovered (SIR) or SEIR model. We will estimate the daily number and the total number of infections and deaths until the end of this epidemic, and also the corresponding turning points. This work is organized as below. In section II, we specify the data used and also shows the varies of epidemic over time. In section III, the models used to describe the data are introduced. Then in section IV, we present the nal results of the predictions. Finally in section V, we give a brief summary and discussion. 2 Data sources In this data-driven study, the main data used are the numbers of daily con rmed infections, totally con rmed infections, daily deaths, and total deaths. The data used are from Jan. 20, 2020, to Feb. 11, 2020, where both data of Jan. 20 and Feb. 11 are included, reported by the National Health Commission of the Republic of China (NHC)[9], and Health Commission of Hubei Province (HCH)[10]. Jan. 20, 2020, containing all the cases reported from 0 to 24, is the zeroth day in our research, and then others are implied. Notice the epidemic in Hubei is quite di erent from the other areas of China, and we will deal with the relevant data separately. All the original data used are listed in Tab. I, which are also graphically shown in Fig. 1 Fig. 3, in which China is used to denote the mainland China, and \Other" mainland China other than Hubei province. Fig. 1 shows the number of total and daily suspected and con rmed cases in mainland China, Hubei, and China other than Hubei. The accumulated number of suspected cases reaches the maximum on the 19th day (Feb. 8), and then drops rapidly. Fig. 2 displays the number of total and daily deaths caused by this epidemic in main- land China, Hubei, and China other than Hubei. Notice the data of the deaths caused by the epidemic in China (black) almost coincides with that in Hubei (red). Almost all the cases of deaths (1068/1113, 96%, until Feb. 11, 2020) locates in Hubei province. This also motivates us to process the data in Hubei separately with that in China other than Hubei. { 2 { Tab. I: The data of epidemic caused by the COVID-19 pneumonia in the mainland China and Hubei, including (A) the daily infections, (B) daily deaths, (C) total infections, (D) total deaths, (E) daily and (F) total suspected cases. China Hubei Date A B C D E F A B C D 2020/1/20 77 2 291 6 27 54 72 2 270 6 2020/1/21 149 3 440 9 26 37 105 3 375 9 2020/1/22 131 8 571 17 257 393 69 8 444 17 2020/1/23 259 8 830 25 680 1072 105 7 549 24 2020/1/24 444 16 1287 41 1118 1965 180 15 729 39 2020/1/25 688 15 1975 56 1309 2684 323 13 1052 52 2020/1/26 769 24 2744 80 3806 5794 371 24 1423 76 2020/1/27 1771 26 4515 106 2077 6973 1291 24 2714 100 2020/1/28 1459 26 5974 132 3248 9239 840 25 3554 125 2020/1/29 1737 38 7711 170 4148 12167 1032 37 4586 162 2020/1/30 1982 43 9692 213 4812 15238 1220 42 5806 204 2020/1/31 2102 46 11791 259 5019 17988 1347 45 7153 249 2020/2/1 2590 45 14380 304 4562 19544 1921 45 9074 294 2020/2/2 2829 57 17205 361 5173 21558 2103 56 11177 350 2020/2/3 3235 64 20438 425 5072 23214 2345 64 13522 414 2020/2/4 3887 65 24324 490 3971 23260 3156 65 16678 479 2020/2/5 3694 73 28018 563 5328 24702 2987 70 19665 549 2020/2/6 3143 73 31161 636 4833 26359 2447 69 22112 618 2020/2/7 3399 86 34546 722 4214 27657 2841 81 24953 699 2020/2/8 2656 89 37198 811 3916 28942 2147 81 27100 780 2020/2/9 3062 97 40171 908 4008 23589 2618 91 29631 871 2020/2/10 2478 108 42638 1016 3536 21675 2097 103 31728 974 2020/2/11 2015 97 44653 1113 3342 16067 1638 94 33366 1068 Fig. 3(a) shows the total and daily deaths over time in China other than Hubei, and the number of total deaths is 45 until Feb. 11, 2020. The varies of death rate over time is displayed in Fig. 3(b), and the death rate in China other than Hubei is multiplied by a factor of 10. 3 Methods The key point to describe and predict the trend of the epidemic is to nd the proper functions. We use the symmetrical function h(t) to describe the data of daily infections { 3 { Suspected(China) Suspected(China) Infected(China) Infected(China) Infected(Hubei) Infected(Hubei) Infected(Other) Infected(Other) 0 0 0 2 4 6 8 10 12 14 16 18 20 22 0 2 4 6 8 10 12 14 16 18 20 22 ith Day ith Day (a) (b) Fig. 1: The number of (a) total and (b) daily suspected and con rmed cases in mainland China, Hubei, and China other than Hubei. 1200 120 Death(China) Death(China) 1000 Death(Hubei) 100 Death(Hubei) Death(Other) Death(Other) 800 80 600 60 400 40 200 20 0 0 0 2 4 6 8 10 12 14 16 18 20 22 0 2 4 6 8 10 12 14 16 18 20 22 ith Day ith Day (a) (b) Fig. 2: The number of (a) total and (b) daily deaths in mainland China, Hubei, and China other than Hubei. and deaths in Hubei, namely, 1 1 h(t) = A ; (3.1) kx k(x1) 1 + e 1 + e where x = (t + 0:5 t ) with t denoting the day, which starts from 0 (Jan. 20, 2020) in the data, and t representing the turning point; A and k are the tting parameters and are determined by the data together with t . And then the accumulated data of infections or deaths are determined by the integration over h(t), namely, H (t) = h( )d; (3.2) For the epidemic in the other areas of China, the data of infections shows an asymmetric { 4 { Total number Number of total deaths Daily number Number of daily deaths 50 0.06 China Total deaths(Other) 0.05 Hubei Daily deaths(Other) Other×10 0.04 0.03 0.02 0.01 0 0 0 2 4 6 8 10 12 14 16 18 20 22 0 2 4 6 8 10 12 14 16 18 20 22 ith Day ith Day (a) (b) Fig. 3: (b) The number of total and daily deaths over time in China other than Hubei; (a) the death rate varies over time for mainland China, Hubei, and China other than Hubei; character, and then will be described by x=k s(t) = B exp[(x + k e )]; (3.3) where x = t t ; the parameters B, k , and k together with t , are then determined by T 1 2 T tting to the data. The is function behaves similar with the Gauss function, but has a long tail in the right hand side of the maximum-value location. The corresponding accumulated data can be then obtained similarly with that in Hubei. Though the two models listed here are quite simple, the obtained results in next section show that they would give quite well descriptions to the data of the epidemic. 4 Results 4.1 Predictions of the epidemic in Hubei With considered data, namely, data from Jan. 20 to Feb. 11, the average error is bout 166 for this model to describe the daily infections, and 190 for the accumulated total infections. The average relative error is about 8:6% for the number of daily infections during this 23 days, and 1:6% for the number of the accumulated infections. Our estimates of the number of total and daily infections are showed in Fig. 4. The extracted turning point of the infection number in Hubei is the 17th day, namely, Feb. 6, 2020. The epidemic in Hubei is predicted to end after the 50th day, namely, after Mar. 10, 2020. The daily infection number is predicted to be less than 1000, 100, and 10 on the 25th (Feb. 14), the 33th (Feb. 22), and the 41th (Mar. 1) day, respectively. We estimated that the epidemic is to end up with a total of 39, 000 infections in Hubei, China. With the considered data, the average errors are bout 4 and 22 for this model to describe the daily and accumulated death numbers, respectively, and the corresponding relative errors are about 8:6% and 6:2%, respectively. { 5 { Death number in China other than Hubei Death rate 3500 0 0 - 5 0 5 10 15 20 25 30 35 40 45 50 - 5 0 5 10 15 20 25 30 35 40 45 50 ith Day ith Day (a) (b) Fig. 4: Data and t of the infections in Hubei of China, with (a) the total infection number, and (b) the daily infection number; the black circle denotes the data, and the green dotted line is the predicted trend; the turning point for infection number in Hubei is calculated to be 17th day, namely, Feb. 6, 2020. 0 0 0 10 20 30 40 50 0 10 20 30 40 50 ith Day ith Day (a) (b) Fig. 5: Data and t of the deaths in Hubei of China, with (a) the total death number, and (b) the daily death number; the black circle denotes the data, and the green dotted line is the predicted trend; the turning point for death number in Hubei is calculated to be the 23th day, namely, Feb. 12, 2020. Fig. 5 shows the estimations of the total and daily death number in Hubei, China. The predicted turning point is the 23st day (Feb. 12, 2020) with the maximum daily death to be less than 100 individuals. The daily death number is estimated to be less than 80, 40, and 10 on the 28th (Feb. 17), 35th (Feb. 24), and 44th (Mar. 4) day, respectively. Notice the distribution of the daily deaths is delayed about 5 6 days compared with the that of the daily infections. 4.2 Predictions of the epidemic in China other than Hubei With the data in the considered period, the average errors are bout 41 and 58 for this model to describe the daily and total accumulated infections, and the corresponding relative errors are about 8:4% and 1:2%, respectively. The numbers of the daily and total infections in China other than Hubei are showed in { 6 { Number of total infections Number of total deaths Number of daily deaths Number of daily infections 1000 8000 600 0 0 - 5 0 5 10 15 20 25 30 35 40 45 50 - 5 0 5 10 15 20 25 30 35 40 45 50 ith Day ith Day (a) (b) Fig. 6: Data (black circles) and t (dotted line) of the infections in China other than Hubei, with (a) the total infection number, and (b) the daily infection number; the turning point for infection number is calculated to be the12th day, namely, Feb. 1, 2020. Fig. 6(a) and Fig. 6(b), respectively. The extracted turning point is the 12th day (Feb. 1, 2020). The epidemic in China other than Hubei is expected to end on the 45th day, namely, on Mar. 5, 2020. The daily infection number is predicted to be less than 300, 80, and 10 on the 23th (Feb. 12), the 28th (Feb. 17), and the 36th (Feb. 25) day, respectively. The estimated number of accumulated infections is about 12600 in China other than Hubei. Due to the minority of the statistical data in death number of China other than Hubei, 45 deaths in the last 20 days, we did not parameterize this data, and hence did not give a description or trend prediction. 5 Summary and discussions The data of the daily and total infections and deaths in Hubei and China other than Hubei are studied and parameterized with proper models. The turning points of the daily infections are predicted to be Feb. 6 and Feb. 1, 2020, for Hubei and China other than Hubei, respectively. The epidemic caused by the COVID-19 in China is predicted to end up after Mar. 10, and the number of the total infections are predicted to be 51600. The data trends show that the quick and active strategies to reduce human exposure taken in China have already had a good impact on control of the epidemic. The techniques used here are data-driven and quite di erent from the work in Ref. [7]. The methodology and forecast results here could give some support for the prevention and control of the outbreaks. The model in this work is parameterized with the latest data until Feb. 11, 2020, reported by the NHC and HCH. A major limitation of this work is that we describe the epidemic data in Hubei with the symmetrical even function. Usually, the epidemic would recede with a long tail, and we might underestimate the total infections and deaths in Hubei, and also the end date. Another limitation is that although we process separately the data of Hubei and China other than Hubei, this is still quite rough for the considerable di erences among cities within Hubei province. Also, as a data-driven forecast study, the seasonality of COVID-19 transmission, the e ects of the quarantine, { 7 { Number of total infections Number of daily infections the limitations on population transportation, and other speci c reasons that may a ect the spread of the epidemic, are not explicitly included in the study. At last, we expect this outbreak would end as soon as possible. Acknowledgments We thank Jing Li and Hao-Nan Wang for the helpful discusses and suggestions. This work is supported by the Open Research Fund of Key Laboratory of Digital Earth Science (2019LDE005), and by the Fundamental Research Funds for the Central Universities under Grant No. 310201911QD054. Con icts of Interest: The authors declare no con ict of interest. Note added. After this work has been written, the Hubei reported 14840 con rmed infections (including 13332 clinically diagnosed cases) on Feb. 12, 2020, which is almost 9 times greater than the data of the previous day. This huge uctuation of the infections is due to the changing of diagnostic criteria in Hubei. And we believe this clinical criteria taken by Hubei would play an active and important role in controlling the outbreak. In order to continue this forecasting research, we consider these 13332 clinically diag- nosed infections follow the same statistical distribution as these cases con rmed by nucleic acid detection. Then the updated prediction of nal accumulated infections would be 1.4 times larger than before, namely, totally there would be 54600 infections in Hubei. How- ever, this would overestimate the real COVID-19 infections, since part of the clinically diagnosed cases would be excluded by the nucleic acid detection later. References [1] Zhu N, Zhang D, Wang W, et al. A novel coronavirus from patients with pneumonia in China, 2019. N. Engl. J. Med. DOI: 10.1056/NEJMoa2001017. [2] Xiang Li, Yuhe Song, Gary Wong, Jie Cui. (2020). Bat origin of a new human coronavirus: there and back again. Science China Life Sciences 13. [3] Rothe, Camilla et al, Transmission of 2019-nCoV Infection from an Asymptomatic Contact in Germany, N. Engl. J. Med., DOI:10.1056/NEJMc2001468 [4] Q. Li, X. Guan, P. Wu, et al. Early transmission dynamics in Wuhan, China, of novel coronavirus-infected pneumonia. N. Engl. J. Med. 2020; published online Jan. 29. DOI:10.1056/NEJMoa2001316 [5] Du Z, Wang L, Chauchemez S, Xu X, Wang X, Cowling BJ, et al. Risk for transportation of 2019 novel coronavirus disease from Wuhan to other cities in China. Emerg. Infect. Dis. Original Publication Date: 2/13/2020. DOI: 10.3201/eid2605.200146 [6] Fu-Sheng Wang, Chao Zhang. (2020). What to do next to control the 2019-nCoV epidemic?. The Lancet 395:10222, 391-393. [7] Joseph T Wu, et al. Nowcasting and forecasting the potential domestic and international spread of the 2019-nCoV outbreak originating in Wuhan, China: a modelling study, The Lancet. DOI:10.1016/S0140-6736(20)30260-9. Online published (Jan. 31, 2020). { 8 { [8] Biao Tang, Nicola Luigi Bragazzi, Qian Li, Sanyi Tang, Yanni Xiao, Jianhong Wu. (2020) An updated estimation of the risk of transmission of the novel coronavirus (2019-nCov). Infectious Disease Modelling. [9] National Health Commission of People's Republic of China, Outbreak noti cation (in Chinese), Jan. 21, 2020-Feb. 13, 2020. [10] Health Commission of Hubei Province, Prevention and control of novel coronavirus pneumonia, Information release (in Chinese), Jan. 21, 2020-Feb. 13, 2020. { 9 { http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Quantitative Biology arXiv (Cornell University)

Trend and forecasting of the COVID-19 outbreak in China

Quantitative Biology , Volume 2020 (2002) – Feb 14, 2020

Loading next page...
 
/lp/arxiv-cornell-university/trend-and-forecasting-of-the-covid-19-outbreak-in-china-JoXQlOblO8
ISSN
0163-4453
eISSN
ARCH-3345
DOI
10.1016/j.jinf.2020.02.014
Publisher site
See Article on Publisher Site

Abstract

Trend and forecasting of the COVID-19 outbreak in China 1 2 Qiang Li Wei Feng School of Physical science and Technology, Northwestern Polytechnical University, Xian, 710129, P. R. China. School of Electronic Engineering, Xidian University, Xian, 710071, P. R. China. Abstract: By using the public data from Jan. 20 to Feb. 11, 2020, we perform data-driven analysis and forecasting on the COVID-19 epidemic in mainland China, especially Hubei province. Our results show that the turning points of the daily infections are predicted to be Feb. 6 and Feb. 1, 2020, for Hubei and China other than Hubei, respectively. The epidemic in China is predicted to end up after Mar. 10, 2020, and the number of the total infections are predicted to be 51600. The data trends reveal that quick and active strategies taken by China to reduce human exposure have already had a good impact on the control of the epidemic. liruo@nwpu.edu.cn wfeng@xidian.edu.cn arXiv:2002.05866v1 [q-bio.PE] 14 Feb 2020 Contents 1 Introduction 1 2 Data sources 2 3 Methods 3 4 Results 5 4.1 Predictions of the epidemic in Hubei 5 4.2 Predictions of the epidemic in China other than Hubei 6 5 Summary and discussions 7 1 Introduction China, especially the Hubei province, is ghting the pneumonia epidemic by implementing various prevention and control measures [1]. Pneumonia, caused by the novel coronavirus (2019-nCoV), which may originate from the bat [2], is just named as the COVID-19 by the World Health Organization (WHO). The COVID-19 outbroke from Wuhan, the capital of Hubei province in China in Dec. 2019, and has spread to other provinces of China and even other countries [3]. Strong human-to-human transmission is established [4], and there exists the risk of transportation of COVID-19 from Wuhan to 369 other cities in China [5]. Until Feb. 11, 2020, there have been 44653 cases of COVID-19 infections con rmed in mainland China, including 1113 deaths. To prevent and control the spread of the epidemic, many strategies are needed [6], and China has already taken quick and e ective strategies to reduce population mobility and interpersonal contact rates, and also increase quarantine on migrants. Predicting the trend of the epidemic are quite important to the allocation of medical resources, the arrangement of production activities, and even the domestic economic development all over China. In recent decades, two other new coronaviruses,Severe Acute Respiratory Syndrome coronavirus (SARS-CoV)" and \Middle East respiratory syndrome coronavirus (MERS- CoV)", have been considered to be major epidemics worldwide. However, concerning these viruses, COVID-19 presents a higher degree of uncertainty in the scale and geographical scope of the outbreak within and outside China. SARS-CoV and MERS-CoV-based prop- agation analysis and prediction models may be no longer suitable in the ght against the COVID-19 pneumonia. Therefore, it is very urgent to use the latest data to establish an ecient and highly suitable epidemic analysis and prediction model according to the ac- tual situation, and then to give reliable epidemic predictions. This work could provide an important reference for the government to formulate emergency macroeconomic decisions { 1 { and medical resources allocation. Moreover, the signi cance is also of great reference value for the country's deployment and adjustment of economic activities throughout the year Very recently, the susceptible-exposed-infectious-recovered (SEIR) model is used to forecast the potential domestic and international spread of this COVID-19 outbreak [7] with parameters estimated from other sources. The e ective daily reproduction ratio and epidemics peak are also predicted in a Ref. [8]. The real situation could be much more complicated and changing all the time. Especially, with the implementation of the Chinese government's multiple epidemic control policies, the control of nationwide epidemic has become obvious. However, the medical supplies in Hubei will still a ect the implementation of national policies. In this study, the models driven by data are used to describe the current data of the epidemic, to predict the ongoing trend, and to estimate the outbreak size of the COVID-19 in both Hubei and other areas in mainland China. This data-driven study could present another prediction to the epidemic compared with the susceptible-infectious- recovered (SIR) or SEIR model. We will estimate the daily number and the total number of infections and deaths until the end of this epidemic, and also the corresponding turning points. This work is organized as below. In section II, we specify the data used and also shows the varies of epidemic over time. In section III, the models used to describe the data are introduced. Then in section IV, we present the nal results of the predictions. Finally in section V, we give a brief summary and discussion. 2 Data sources In this data-driven study, the main data used are the numbers of daily con rmed infections, totally con rmed infections, daily deaths, and total deaths. The data used are from Jan. 20, 2020, to Feb. 11, 2020, where both data of Jan. 20 and Feb. 11 are included, reported by the National Health Commission of the Republic of China (NHC)[9], and Health Commission of Hubei Province (HCH)[10]. Jan. 20, 2020, containing all the cases reported from 0 to 24, is the zeroth day in our research, and then others are implied. Notice the epidemic in Hubei is quite di erent from the other areas of China, and we will deal with the relevant data separately. All the original data used are listed in Tab. I, which are also graphically shown in Fig. 1 Fig. 3, in which China is used to denote the mainland China, and \Other" mainland China other than Hubei province. Fig. 1 shows the number of total and daily suspected and con rmed cases in mainland China, Hubei, and China other than Hubei. The accumulated number of suspected cases reaches the maximum on the 19th day (Feb. 8), and then drops rapidly. Fig. 2 displays the number of total and daily deaths caused by this epidemic in main- land China, Hubei, and China other than Hubei. Notice the data of the deaths caused by the epidemic in China (black) almost coincides with that in Hubei (red). Almost all the cases of deaths (1068/1113, 96%, until Feb. 11, 2020) locates in Hubei province. This also motivates us to process the data in Hubei separately with that in China other than Hubei. { 2 { Tab. I: The data of epidemic caused by the COVID-19 pneumonia in the mainland China and Hubei, including (A) the daily infections, (B) daily deaths, (C) total infections, (D) total deaths, (E) daily and (F) total suspected cases. China Hubei Date A B C D E F A B C D 2020/1/20 77 2 291 6 27 54 72 2 270 6 2020/1/21 149 3 440 9 26 37 105 3 375 9 2020/1/22 131 8 571 17 257 393 69 8 444 17 2020/1/23 259 8 830 25 680 1072 105 7 549 24 2020/1/24 444 16 1287 41 1118 1965 180 15 729 39 2020/1/25 688 15 1975 56 1309 2684 323 13 1052 52 2020/1/26 769 24 2744 80 3806 5794 371 24 1423 76 2020/1/27 1771 26 4515 106 2077 6973 1291 24 2714 100 2020/1/28 1459 26 5974 132 3248 9239 840 25 3554 125 2020/1/29 1737 38 7711 170 4148 12167 1032 37 4586 162 2020/1/30 1982 43 9692 213 4812 15238 1220 42 5806 204 2020/1/31 2102 46 11791 259 5019 17988 1347 45 7153 249 2020/2/1 2590 45 14380 304 4562 19544 1921 45 9074 294 2020/2/2 2829 57 17205 361 5173 21558 2103 56 11177 350 2020/2/3 3235 64 20438 425 5072 23214 2345 64 13522 414 2020/2/4 3887 65 24324 490 3971 23260 3156 65 16678 479 2020/2/5 3694 73 28018 563 5328 24702 2987 70 19665 549 2020/2/6 3143 73 31161 636 4833 26359 2447 69 22112 618 2020/2/7 3399 86 34546 722 4214 27657 2841 81 24953 699 2020/2/8 2656 89 37198 811 3916 28942 2147 81 27100 780 2020/2/9 3062 97 40171 908 4008 23589 2618 91 29631 871 2020/2/10 2478 108 42638 1016 3536 21675 2097 103 31728 974 2020/2/11 2015 97 44653 1113 3342 16067 1638 94 33366 1068 Fig. 3(a) shows the total and daily deaths over time in China other than Hubei, and the number of total deaths is 45 until Feb. 11, 2020. The varies of death rate over time is displayed in Fig. 3(b), and the death rate in China other than Hubei is multiplied by a factor of 10. 3 Methods The key point to describe and predict the trend of the epidemic is to nd the proper functions. We use the symmetrical function h(t) to describe the data of daily infections { 3 { Suspected(China) Suspected(China) Infected(China) Infected(China) Infected(Hubei) Infected(Hubei) Infected(Other) Infected(Other) 0 0 0 2 4 6 8 10 12 14 16 18 20 22 0 2 4 6 8 10 12 14 16 18 20 22 ith Day ith Day (a) (b) Fig. 1: The number of (a) total and (b) daily suspected and con rmed cases in mainland China, Hubei, and China other than Hubei. 1200 120 Death(China) Death(China) 1000 Death(Hubei) 100 Death(Hubei) Death(Other) Death(Other) 800 80 600 60 400 40 200 20 0 0 0 2 4 6 8 10 12 14 16 18 20 22 0 2 4 6 8 10 12 14 16 18 20 22 ith Day ith Day (a) (b) Fig. 2: The number of (a) total and (b) daily deaths in mainland China, Hubei, and China other than Hubei. and deaths in Hubei, namely, 1 1 h(t) = A ; (3.1) kx k(x1) 1 + e 1 + e where x = (t + 0:5 t ) with t denoting the day, which starts from 0 (Jan. 20, 2020) in the data, and t representing the turning point; A and k are the tting parameters and are determined by the data together with t . And then the accumulated data of infections or deaths are determined by the integration over h(t), namely, H (t) = h( )d; (3.2) For the epidemic in the other areas of China, the data of infections shows an asymmetric { 4 { Total number Number of total deaths Daily number Number of daily deaths 50 0.06 China Total deaths(Other) 0.05 Hubei Daily deaths(Other) Other×10 0.04 0.03 0.02 0.01 0 0 0 2 4 6 8 10 12 14 16 18 20 22 0 2 4 6 8 10 12 14 16 18 20 22 ith Day ith Day (a) (b) Fig. 3: (b) The number of total and daily deaths over time in China other than Hubei; (a) the death rate varies over time for mainland China, Hubei, and China other than Hubei; character, and then will be described by x=k s(t) = B exp[(x + k e )]; (3.3) where x = t t ; the parameters B, k , and k together with t , are then determined by T 1 2 T tting to the data. The is function behaves similar with the Gauss function, but has a long tail in the right hand side of the maximum-value location. The corresponding accumulated data can be then obtained similarly with that in Hubei. Though the two models listed here are quite simple, the obtained results in next section show that they would give quite well descriptions to the data of the epidemic. 4 Results 4.1 Predictions of the epidemic in Hubei With considered data, namely, data from Jan. 20 to Feb. 11, the average error is bout 166 for this model to describe the daily infections, and 190 for the accumulated total infections. The average relative error is about 8:6% for the number of daily infections during this 23 days, and 1:6% for the number of the accumulated infections. Our estimates of the number of total and daily infections are showed in Fig. 4. The extracted turning point of the infection number in Hubei is the 17th day, namely, Feb. 6, 2020. The epidemic in Hubei is predicted to end after the 50th day, namely, after Mar. 10, 2020. The daily infection number is predicted to be less than 1000, 100, and 10 on the 25th (Feb. 14), the 33th (Feb. 22), and the 41th (Mar. 1) day, respectively. We estimated that the epidemic is to end up with a total of 39, 000 infections in Hubei, China. With the considered data, the average errors are bout 4 and 22 for this model to describe the daily and accumulated death numbers, respectively, and the corresponding relative errors are about 8:6% and 6:2%, respectively. { 5 { Death number in China other than Hubei Death rate 3500 0 0 - 5 0 5 10 15 20 25 30 35 40 45 50 - 5 0 5 10 15 20 25 30 35 40 45 50 ith Day ith Day (a) (b) Fig. 4: Data and t of the infections in Hubei of China, with (a) the total infection number, and (b) the daily infection number; the black circle denotes the data, and the green dotted line is the predicted trend; the turning point for infection number in Hubei is calculated to be 17th day, namely, Feb. 6, 2020. 0 0 0 10 20 30 40 50 0 10 20 30 40 50 ith Day ith Day (a) (b) Fig. 5: Data and t of the deaths in Hubei of China, with (a) the total death number, and (b) the daily death number; the black circle denotes the data, and the green dotted line is the predicted trend; the turning point for death number in Hubei is calculated to be the 23th day, namely, Feb. 12, 2020. Fig. 5 shows the estimations of the total and daily death number in Hubei, China. The predicted turning point is the 23st day (Feb. 12, 2020) with the maximum daily death to be less than 100 individuals. The daily death number is estimated to be less than 80, 40, and 10 on the 28th (Feb. 17), 35th (Feb. 24), and 44th (Mar. 4) day, respectively. Notice the distribution of the daily deaths is delayed about 5 6 days compared with the that of the daily infections. 4.2 Predictions of the epidemic in China other than Hubei With the data in the considered period, the average errors are bout 41 and 58 for this model to describe the daily and total accumulated infections, and the corresponding relative errors are about 8:4% and 1:2%, respectively. The numbers of the daily and total infections in China other than Hubei are showed in { 6 { Number of total infections Number of total deaths Number of daily deaths Number of daily infections 1000 8000 600 0 0 - 5 0 5 10 15 20 25 30 35 40 45 50 - 5 0 5 10 15 20 25 30 35 40 45 50 ith Day ith Day (a) (b) Fig. 6: Data (black circles) and t (dotted line) of the infections in China other than Hubei, with (a) the total infection number, and (b) the daily infection number; the turning point for infection number is calculated to be the12th day, namely, Feb. 1, 2020. Fig. 6(a) and Fig. 6(b), respectively. The extracted turning point is the 12th day (Feb. 1, 2020). The epidemic in China other than Hubei is expected to end on the 45th day, namely, on Mar. 5, 2020. The daily infection number is predicted to be less than 300, 80, and 10 on the 23th (Feb. 12), the 28th (Feb. 17), and the 36th (Feb. 25) day, respectively. The estimated number of accumulated infections is about 12600 in China other than Hubei. Due to the minority of the statistical data in death number of China other than Hubei, 45 deaths in the last 20 days, we did not parameterize this data, and hence did not give a description or trend prediction. 5 Summary and discussions The data of the daily and total infections and deaths in Hubei and China other than Hubei are studied and parameterized with proper models. The turning points of the daily infections are predicted to be Feb. 6 and Feb. 1, 2020, for Hubei and China other than Hubei, respectively. The epidemic caused by the COVID-19 in China is predicted to end up after Mar. 10, and the number of the total infections are predicted to be 51600. The data trends show that the quick and active strategies to reduce human exposure taken in China have already had a good impact on control of the epidemic. The techniques used here are data-driven and quite di erent from the work in Ref. [7]. The methodology and forecast results here could give some support for the prevention and control of the outbreaks. The model in this work is parameterized with the latest data until Feb. 11, 2020, reported by the NHC and HCH. A major limitation of this work is that we describe the epidemic data in Hubei with the symmetrical even function. Usually, the epidemic would recede with a long tail, and we might underestimate the total infections and deaths in Hubei, and also the end date. Another limitation is that although we process separately the data of Hubei and China other than Hubei, this is still quite rough for the considerable di erences among cities within Hubei province. Also, as a data-driven forecast study, the seasonality of COVID-19 transmission, the e ects of the quarantine, { 7 { Number of total infections Number of daily infections the limitations on population transportation, and other speci c reasons that may a ect the spread of the epidemic, are not explicitly included in the study. At last, we expect this outbreak would end as soon as possible. Acknowledgments We thank Jing Li and Hao-Nan Wang for the helpful discusses and suggestions. This work is supported by the Open Research Fund of Key Laboratory of Digital Earth Science (2019LDE005), and by the Fundamental Research Funds for the Central Universities under Grant No. 310201911QD054. Con icts of Interest: The authors declare no con ict of interest. Note added. After this work has been written, the Hubei reported 14840 con rmed infections (including 13332 clinically diagnosed cases) on Feb. 12, 2020, which is almost 9 times greater than the data of the previous day. This huge uctuation of the infections is due to the changing of diagnostic criteria in Hubei. And we believe this clinical criteria taken by Hubei would play an active and important role in controlling the outbreak. In order to continue this forecasting research, we consider these 13332 clinically diag- nosed infections follow the same statistical distribution as these cases con rmed by nucleic acid detection. Then the updated prediction of nal accumulated infections would be 1.4 times larger than before, namely, totally there would be 54600 infections in Hubei. How- ever, this would overestimate the real COVID-19 infections, since part of the clinically diagnosed cases would be excluded by the nucleic acid detection later. References [1] Zhu N, Zhang D, Wang W, et al. A novel coronavirus from patients with pneumonia in China, 2019. N. Engl. J. Med. DOI: 10.1056/NEJMoa2001017. [2] Xiang Li, Yuhe Song, Gary Wong, Jie Cui. (2020). Bat origin of a new human coronavirus: there and back again. Science China Life Sciences 13. [3] Rothe, Camilla et al, Transmission of 2019-nCoV Infection from an Asymptomatic Contact in Germany, N. Engl. J. Med., DOI:10.1056/NEJMc2001468 [4] Q. Li, X. Guan, P. Wu, et al. Early transmission dynamics in Wuhan, China, of novel coronavirus-infected pneumonia. N. Engl. J. Med. 2020; published online Jan. 29. DOI:10.1056/NEJMoa2001316 [5] Du Z, Wang L, Chauchemez S, Xu X, Wang X, Cowling BJ, et al. Risk for transportation of 2019 novel coronavirus disease from Wuhan to other cities in China. Emerg. Infect. Dis. Original Publication Date: 2/13/2020. DOI: 10.3201/eid2605.200146 [6] Fu-Sheng Wang, Chao Zhang. (2020). What to do next to control the 2019-nCoV epidemic?. The Lancet 395:10222, 391-393. [7] Joseph T Wu, et al. Nowcasting and forecasting the potential domestic and international spread of the 2019-nCoV outbreak originating in Wuhan, China: a modelling study, The Lancet. DOI:10.1016/S0140-6736(20)30260-9. Online published (Jan. 31, 2020). { 8 { [8] Biao Tang, Nicola Luigi Bragazzi, Qian Li, Sanyi Tang, Yanni Xiao, Jianhong Wu. (2020) An updated estimation of the risk of transmission of the novel coronavirus (2019-nCov). Infectious Disease Modelling. [9] National Health Commission of People's Republic of China, Outbreak noti cation (in Chinese), Jan. 21, 2020-Feb. 13, 2020. [10] Health Commission of Hubei Province, Prevention and control of novel coronavirus pneumonia, Information release (in Chinese), Jan. 21, 2020-Feb. 13, 2020. { 9 {

Journal

Quantitative BiologyarXiv (Cornell University)

Published: Feb 14, 2020

References