Reanalyze the Number and Mortality of COVID-19 Patients in Wuhan from Current Data



Author: Sisyphus Opinions (Caixin Blog)

Translator: COVID-19 Readings

There are many places worth analyzing and researching about the currently reported COVID-19 data.

Data Sources:

The analysis in this article is based on the figures as of January 27, 8:30 a.m.

According to data from National Health Commission, Hubei Health Commission and various provincial health commissions

As of 2:00pm on January 26, a total of 2,744 confirmed cases and 80 deaths were reported nationwide

Among them, Hubei cumulatively confirmed 1,423 cases, 221 severe cases, 69 critical cases and 76 deaths.

Within Hubei, 698 cases in Wuhan, 141 severe cases, 51 critical cases, 63 deaths

Hubei (outside Wuhan) 725 cases, 80 severe cases, 18 critically cases, 13 deaths

Quick calculations found 1,321 confirmed cases nationwide (outside Hubei) and 4 deaths


Based on this data, calculating the mortality rate in confirmed cases, we get completely unreasonable figures.

Death / confirmed cases: National average 80/2744 = 2.9%

Wuhan 63/698 = 9%

Hubei (outside Wuhan) 13/725 = 1.8%

National (outside Hubei) 4/1321 = 0.3%

This is obviously problematic. The mortality rate in Wuhan is 30 times that of other provinces across the country and 5 times that of other regions in Hubei. This data is definitely wrong.


Let’s analyze the problem with the above data.

First, I made three assumptions based on common sense.

  1. For confirmed cases, we have reason to believe that statistics outside Hubei are more accurate. Right now other provinces are on extremely high alert for people from Wuhan. Whenever a positive case is found, all contacts are isolated.

  2. Statistics on deaths and confirmed cases will only be underreported and not overreported. It is possible that people who died from COVID-19 were reported to have died from other causes due to a lack of confirmed diagnosis. Especially given the current situation of extreme shortage of medical resources in Wuhan.

  3. The statistics of deaths should be more accurate than the confirmed cases. After all, the death of a patient is a major event, and the cause of death must be given. But for mild patients, they may recover on their own before diagnosis.

Based on these assumptions, we put forward some possible reasons for the above-mentioned problem of Wuhan mortality rate being five times that of other regions in Hubei and 30 times that of the whole country.

Possible Reasons

  1. The number of confirmed cases in Wuhan is much smaller than the actual number of cases.

  2. Medical institutions in Wuhan are overwhelmed, leading to inadequate treatment of severely ill patients and causing high mortality rate.

  3. Outside cases are at an early stage and have not yet reached death.

  4. Most patients in Wuhan are early in the chain of transmission, and the virus is more toxic (the mortality rate in the first batch of 41 cases was as high as 15%).

My estimation is that all four reasons are there, but the latter three reasons cannot explain the 30-fold difference between the mortality rates in Wuhan and other provinces.


Based on currently available data, let’s do a very rough analysis on the actual number of cases in Wuhan / Hubei, and a reasonable estimate of the mortality rate of COVID-19.

Analysis data source: National epidemic data, speech by the mayor of Wuhan (total population 14 million, staying in Wuhan 9 million, outflow 5 million) Baidu Wisdom migration data: of all Wuhan’s outflow population, 65% are within the province and 35% outside the province.

wuhan migration data
During the Spring Festival travel period, 60-70% of people who left Wuhan each day went to other cities in Hubei

Start of analysis:

Of the 1,321 patients nationwide (outside Hubei), the vast majority are imported cases, with a rough estimate of 90% being imported cases

1321 * 90% = 1189 cases

In Hubei outside Wuhan, the vast majority of the 725 patients are also imported cases. Assuming 95% to be imported cases,

725 * 95% = 689 cases

1.75 million Wuhanese outside of Wuhan, of which 1189 are patients, the prevalence rate is 6.79 per 10,000

689 cases in Hubei outside Wuhan, prevalence rate 2.12 per 10,000

698 cases in Wuhan, prevalence rate 0.76 per 10,000

This data is obviously unreasonable. The prevalence rate of Wuhanese outside Hubei is 9 times that of Wuhan locals. In addition, the population of Wuhanese outside Hubei should not increase again (due to the quarantine of the city), whereas the number of patients outside Hubei will definitely increase, so the proportion of patients outside Hubei will continue to rise.

At the same time, people who stay in Wuhan are more likely to be cross-infected, and the isolation conditions in hospitals are not as good as those in other provinces. The prevalence rate for people who stay in Wuhan should be higher than that of Wuhanese who have left Hubei.

The most reasonable explanation is that due to the overload of local medical resources and test kit shortage in Wuhan, the number of diagnosed patients is much smaller than the actual number of patients. Many of these patients have recovered on their own and not been diagnosed at all. Similar conditions exist in other cities in Hubei, but to a lesser degree than in Wuhan.

If we apply the nationwide prevalence rate to Wuhan and Hubei, the reasonable number of patients will be:

Wuhan = 900 * 6.79 = 6000 or more people, about 8-9 times the currently reported cases

Hubei outside Wuhan = 325 * 6.79 = around 2200 people, about 2.6 times the currently reported cases

I think these numbers are reasonable. Wuhan is currently constructing two “Wuhan Xiaotangshan” hospitals for a total of 2,000 beds. At the same time, existing hospitals are also constantly requisitioning beds. According to the mayor of Wuhan, 10,000 beds have been requisitioned. If there are only 2,000 potentially confirmed cases (currently 698 diagnosed +1000 suspected cases), such efforts are not required. If about half of the suspected cases will be diagnosed, then about 6,000 patients correspond to 12,000 suspected cases, and more than 10,000 beds will be required for treatment. It also seems that the Wuhan government is currently making preparations for at least 5,000 cases.

If I use the cases I speculated to calculate the current mortality rate.

Wuhan 63 / 6000 = 1.05%

Hubei outside Wuhan 13 / 1800 = 0.7%

Nationwide outside Hubei 4 / 1040 = 0.4%

The death rate in Wuhan is still higher than in other parts of the country, but this data is much more reasonable. The last three reasons beyond the number of confirmed cases may be the main reasons for the high mortality rate in Wuhan.

In addition, I personally think that 6000 cases may still have underestimated the actual number of patients in Wuhan. Because common sense indicates that people who are still staying in Wuhan should have a higher probability of getting sick than those who have left Wuhan. At the same time, the number of patients in other provinces outside Hubei is still increasing rapidly. It is inevitable that the total number of imported cases from Wuhan in other provinces will reach 1500-2000. If one presumes the case prevalence rate of Wuhanese in other provinces to be one in 1,000, and the prevalence rate for people staying in Wuhan to be higher, I will not be surprised that the actual number of patients in Wuhan exceeds 10,000.

However, there is no need to panic for this. Because this novel coronavirus pneumonia is itself characterized by high infectivity, high number of patients, low pathogenicity, and low mortality. People should not regard the novel coronavirus pneumonia as another SARS, but should regard it as a super strong flu.


The novel coronavirus is in between influenza and SARS, its infectivity being higher than SARS, but its pathogenicity and mortality rate being much lower than SARS.

Nature is fair. Among the three coronaviruses that trigger public safety incidents, MERS is less infectious than SARS, while its mortality rate is higher than SARS. The novel coronavirus is more infectious than SARS, while its mortality rate is lower than SARS.

SARS infected young and middle-aged people, being in good health didn’t make any difference. This time most of the deaths caused by the novel coronavirus were among elderly people who were not in good health. From the aspect that most of those affected and deaths are among elderly people who are not in good health, the novel coronavirus is closer to the flu than SARS.

In the United States, the total influenza infection rate in 2018-2019 was 10% (over 30 million infections with a population of over 300 million people in the US), the severe (hospitalization) rate was 1.4%, the excess mortality rate of the total population caused by influenza was 1 in 10,000 (30,000 more people died, with a population of over 300 million). The mortality rate of the infected people was 0.1%.

At present, my prediction for the novel coronavirus is: for the epicenter Wuhan, the infection rate of the total population will be above 0.1%, and the excess mortality rate of the total population will definitely be lower than 5 per 100,000 (I estimate it should be below 3 per 100,000). The patients’ mortality rate will definitely be below 5% (my own estimate is below 2%, around 1%).

Compared with the 2018-2019 US flu, even in the epidemic epicenter Wuhan, the total population infection rate is 1% of the 2018-2019 US flu, even according to the worst forecast, the total excess mortality rate from the epicenter Wuhan will not reach half the rate of the 2018-2019 US flu.

As for nationwide statistics, the total population infection rate is about one ten-thousandth of the 2018-2019 US flu. Even according to the worst forecast, the excess mortality rate of the total population is at most five thousandths of the 2018-2019 US flu. (The excess influenza mortality rate in the United States is 1 in 10,000, whereas the nationwide excess mortality rate for the novel coronavirus in China is at most 1/10,000 multiplied by 5/1000).

So even the students in Wuhan really don’t need to panic. Even if you do get infected by the novel coronavirus, as long as you are generally in good health, it should all turn out okay.

We need to relax and don’t pay too much attention to the number of confirmed cases each day. The total number of SARS diagnoses in China was over 5,000 with 327 deaths. Although I think the total number of patients from the novel coronavirus will be higher than SARS, the number of deaths I hope, and I believe, will be lower than SARS. In addition, young and middle-aged people accounted for a large proportion of SARS deaths, and most of the novel coronavirus deaths were elderly people with poor health and complications. Even with common pneumonia, many of them might not have been able to survive.

Please don’t worry about the rising number of confirmed cases every day. I am most concerned about two pieces of data right now. The first is whether there is large-scale local transmission outside Hubei, and the second is the number of deaths outside Hubei. In both cases, as of the morning of the 27th, both were okay.

The number of patients in Wuhan is definitely more than expected, but with the strength of the current measures, I feel that the number of newly added cases is already trending down. My optimistic estimate is that the novel coronavirus should basically be under control within February.

It’s no big deal, we can definitely defeat the virus.

Related Articles:

Those Outside of The Statistics: They Died from Common Pneumonia? 2020-02-02