新冠肺炎深圳市数据分析 —— 2月16日 Data Analysis of COVID-19 (Coronavirus) Cases in Shenzhen - 16 Feb
- （2月25日更新）年龄较大的患者康复所需时间更长; 80岁患者在25天内治愈出院的概率只有20岁患者的一半
The following patterns have been observed for COVID-19 cases in Shenzhen:
- Similar to Beijing, there is no evidence to suggest either males or females are more susceptible to the virus
- The average age of cases has been decreasing with more children being infected
- The time between the onset of symptoms and admission has been reducing, thanks to the increasing awareness of the disease and effort in prevention and control
- Most of the cases found in Shenzhen are people who have travel history in Hubei or other provinces or/and contact history of other infected cases; the risk of being infected in public places in Shenzhen remains low
- Nanshan, Futian, Longgang and Baoan have the most number of cases. However, the growth rate has been dropping
- (Updated on 25 Feb) Older patients take longer time to recover; the chance of recovery within 25 days for a 80-year-old patient is only half of that of a 20-year-old
Shenzhen government has been publishing places where the new cases were found. I have gathered the data and made a map. This will be updated daily. Hope you all stay safe and healthy!
Of the first 406 cases, 52.0% is female. The difference is not statistically significant.
Below chart shows the age distribution of the first 406 cases in Shenzhen. It tentatively shows two peaks, one at age 30-40 and one at age 60-65. This is similar to what we have observed for the first 175 cases in Beijing. The peak consists of younger people could be linked to their mobility while the older group might be more susceptible to infection. With a bigger sample size, the noise can be reduced and we can take another look to confirm this.
Over time, we saw more young people being infected (particularly those younger than 10 years old), dragging down the average age. Those young people could be the close contact of the imported cases. But it is also likely to be the case that young people have longer incubation period, delaying the onset of the symptoms. Understanding the relation between cases would help us to test these hypotheses.
发病到入院时间 (Time Between First Symptoms and Hospital Admission)
Shorter time between the onset of symptoms and hospital admission leads to lower risk of spreading the disease. Until now, the average time between first symptoms and admission is 4 days. This was reduced from 9 days at the beginning of the outbreak to 1 day now, thanks to the increasing awareness of the virus and effort of prevention and control. However, there were a couple of days when many cases spent more than a week with symptoms before admitting into hospitals, leading to a high risk of spreading. Cases who do not have infected relatives spent a longer period (4.7 days) with symptoms before being hospitalised than those with infected relatives (3.1 days). The latter group were usually quarantined when their close contacts were diagnosed and this speeds up the admission process. While the former group tend to delay seeking medical advice when developing symptoms.
接触史 (Contact History)
Unlike Beijing, Shenzhen COVID-19 cases tend to be people with travel history in Hubei other other provinces and the close contacts of infected people. However, there has been an increasing number of cases who have neither known contact nor travel history (blue bars in the chart) since 13 Jan. This suggests they may have been infected in public places in Shenzhen. Tracking down their activities is vital to prevent further outbreaks of the disease.
康复时间 (Recovery Time)
Recovery Time model was last updated on 25 Feb. Shenzhen government had stopped published detail information on discharged patients from 26 Feb.
截止到2月25日，共有262位病人顺利出院（样本共417例）。已出院的病人住院时间的中位数为17天。通过Cox Proportional Hazards模型建模分析发现以下关系：
262 COVID-19 patients have been discharged from hospitals by 25 Feb (out of the total 417 cases in the sample). The median number of days they spent in hospital is 17. Using Cox Proportional Hazards model, I was able to recover the following patterns:
- Older patients need longer time to recover. A 20-year-old patient has a 86% chance of recovering within 25 days while an 80-year-old patient only has 42% of chance.
- Building models using the existing data is useful to predict the demand for medical resources. The accuracy of this model will improve as more patients being discharged.
各城区情况 (Cases by District)
Nanshan, Futian, Longgang and Baoan have the most number of cases in Shenzhen. However, the cases have been growing at a slower pace now. There is no significant growth in new cases in the other districts.
发病例活动小区或场所 (Known Places of the New Cases)
The map below is constructed using data from Shenzhen Municipal Health Commission. They have been publishing known places of the new cases since 31 jan. Confirmed cases all have been hospitalised but necessary protection and caution is still required for residents nearby. I will be updating the map daily, please click here for the latest map.
数据说明 (About the Data)
Data for details of the COVID-19 cases, number of cases in each districts and known places of the cases were gathered from Shenzhen Municipal Health Commission. The data is mainly in the form of text or image. I have manually (or with the help of scripts) collected them into databases, which I have published in this GitHub repo. Please feel free to use it under the Creative Common License. Special thanks to Lchiffon, the author of REmap, which utilises Baidu Map to make the visualisation possible.