Early outbreak of viral pneumonia of unknown aetiology as at 9 January
On 31 December 2019, the Wuhan Municipal Health Commission announced a cluster of cases of viral pneumonia of unexplained aetiology. The Southern China Seafood Wholesale Market in Wuhan was suspected to be related to the first 27 pneumonia cases without identified pathogenic agents that were reported in late December 2019. Most of the early cases were reportedly either shop owners, largely in the West District of the Southern China Seafood Wholesale Market, or people who visited the market before symptom onset. This market is a large open complex of 50,000 square metres including sections selling seafood, fresh meat, produce, other perishable goods, and a very wide variety of live wild animals for consumption. While wet markets selling such perishable food products are common in China, they usually do not sell such a wide variety of wild animals. Environmental disinfection of the Southern China Seafood Wholesale Market was initiated on 30 December 2019 and the market was closed on 1 January 2020. Contact tracing was initiated, and by 5 January, more than 160 close contacts of these cases were under medical surveillance, and none of them had an infection.
Early information on 2019-nCoV infections as at 12 January
On 9 January 2020, a novel coronavirus, 2019-nCoV, was officially identified as the cause of the outbreak of pneumonia in Wuhan. Following the official announcement of the genetic sequence of the virus, on 11 January, 41 laboratory-confirmed cases of 2019-nCov infection with pneumonia were reported in Wuhan. While the case definition for laboratory-confirmed cases has not been officially published, our understanding is that the initial case definition required (i) fever, (ii) x-ray evidence of pneumonia, (iii) white blood cell count normal or low or low lymphocyte count, (iv) antibiotic treatment for 3 days without improvement, in addition to (v) one or more recent visits to Wuhan or direct or indirect exposure to a wet market in Wuhan, and (vi) a respiratory specimen positive for 2019-nCoV and confirmed as 2019-nCoV by whole genome sequencing. The earliest known case had illness onset on 8 December 2019.
Between 10 and 12 January when there were no new cases reported, the Wuhan Municipal Health Commission announced that no new cases of 2019-nCoV infection were identified with illness onset after 3 January. As at 11 January, more than 700 close contacts were under medical surveillance, more than half being healthcare workers (HCWs), with no infections being identified. Among the first 41 confirmed cases, approximately 70% reported exposure to the Southern China Seafood Wholesale Market. Using the method described by Cauchemez et al., assuming that the market was the only source of zoonotic infections, we estimated R
0 to be 0.3 (95% confidence interval (95% CI): 0.17–0.44).
Information on exported cases identified outside mainland China as at 17 January
On 13 January 2020, Thai health authorities reported an imported case in a person in their 60s who had travelled from Wuhan. This person had onset of an upper respiratory illness on 5 January and landed in Bangkok on 8 January, and was screened out by the thermal scanners at the airport. They did not visit the Southern China Seafood Wholesale Market, but reported visiting another wet market in Wuhan. On 16 January, Japanese health authorities reported a confirmed imported case in a person in their 30s who had travelled from Wuhan and landed in Kanagawa on 6 January. Illness onset was on 3 January and the person has since recovered. The person had not visited any wet market in Wuhan, but had visited a close relative who was in hospital in Wuhan with pneumonia. A third case was reported by Thai health authorities on 17 January. This person was in their 70s and had landed in Bangkok on 13 January, reporting an illness onset on 6 January and unclear history of exposure to the market in Wuhan. On 20 January, the fourth exported case from Wuhan, a person in their 30s visiting South Korea, was identified in Seoul. This person reported no previous visit to wet markets in Wuhan nor contact with any other case within 2 weeks before illness onset. The lack of exposure history to wet markets in Wuhan in two of four generally-mild exported cases indicated that there might be a larger number of undetected infections in Wuhan.
Two family clusters involving five people were reported from 15 to 16 January. One cluster of three close relatives all with illness onset on the same day, were thought to have occurred through a common exposure since they all lived together and worked in the same stall in the Southern China Seafood Wholesale Market. The other cluster involved a couple and may have occurred through human-to-human transmission with a serial interval of 5 days.
The basic reproduction number R
0 is defined as the expected number of secondary cases produced by a typical single infection in a completely susceptible population. If there had been only one case infected by human-to-human transmission among the first 41 identified cases by that date, it implies R
0 was 0.02 (i.e. 1/41).
Information on cases after creation of the National Health Commission task force on 18 January
On 19 January 2020, a task force under the National Health Commission, created on 18 January, visited Wuhan. On 20 January, 136 new infections in Wuhan, including 100 mild cases, 33 severe cases and three critically ill cases, were reported by the Wuhan Municipal Health Commission. Three traveller cases from Wuhan were also identified in Beijing and Shenzhen, possibly suggesting a large number of infections in Wuhan itself. However, further information on the most recently confirmed 2019-nCoV infections is required to determine the transmissibility of the virus and the severity profile of infections.
On 20 January, the task force reported three possible episodes of human-to-human transmission. This included, in two different families, family members who had not recently visited Wuhan being laboratory-confirmed with 2019-nCoV infection after other family members came back from Wuhan and were confirmed cases. A possible super-spreading event was also speculated by the task force given that there had been 15 HCWs infected with the virus in Wuhan until 20 January. On 21 January, additional exported cases were reported in Taiwan and the United States, with illness onset on 11 and 19 January respectively.
Risk of fatality among hospitalised cases
We estimated the hospital fatality risk, i.e. the risk of fatality among hospitalised cases using the formula (fatal cases)/(fatal cases + recovered cases), which provides a more accurate early estimate of the hospital fatality risk compared with (fatal cases)/(all cases). We estimated the associated 95% CI for the hospital fatality risk using the binomial distribution. According to the update on 21 January 2020 when information on deaths and recoveries were reported, four cases had died while 25 had recovered, and our estimate of the hospital fatality risk is therefore 14% (95% CI: 3.9–32%). The estimate of the hospital fatality risk remained fairly stable over the 10 day period since the first death was announced on 11 January (Figure 2). If deaths continue to be reported without any corresponding increase in reported recoveries, the formula will overestimate the risk of fatality.
Comparison of two hypothetical scenarios: one with large zoonotic spillover and one with small zoonotic spillover
To date, as at 22 January, official information has not been sufficient to determine the source of infection and risk of human-to-human transmission for the newly emerged 2019-nCoV. Two competing hypotheses are therefore proposed and compared as potential explanations for the characteristics of the current outbreak of 2019-nCoV infections within and outside mainland China. Scenario 1 comprises a large zoonotic spillover event starting in early December 2019, perhaps over a number of days or weeks, and very limited human-to-human transmission subsequently. Scenario 2 comprises a small zoonotic spillover event in early December 2019 followed by efficient human-to-human transmission. Evidence consistent or inconsistent with each of these two scenarios is outlined in the Table.
In this article, we describe a preliminary assessment of the outbreak of infections with the newly identified 2019-nCoV. This assessment is based on the cases of infection reported over time by health authorities in Wuhan and then at the national level, as well as the media in China and other countries. One of the most urgent priorities is to determine the degree of human-to-human transmissibility of the novel pathogen, and accordingly, this is where information is most urgently needed. We outline two possible scenarios in the Table and find that the early evidence was most consistent with limited human-to-human transmissibility, however more recent data seem to be increasingly more compatible with scenario 2 in which sustained human-to-human transmission has been occurring. Determining the exposure profile among the recently confirmed cases would directly contribute to this assessment. Additional information on approaches to case identification and laboratory testing protocols in Wuhan and in other cities in China would also be informative. A separate priority is to identify the animal reservoir of this novel pathogen and any intermediary hosts, including potential supply chains of wild or game meat.
It is challenging to judge severity from the information available to date. We estimated the risk of death among hospitalised cases of around 14% (Figure 2). For both SARS-CoV and MERS-CoV infections, the risk of severe disease increases substantially with age and with the presence of underlying conditions. One other caveat with estimating severity is that there can be long delays between hospitalisation and death for infections that are ultimately fatal. For SARS in Hong Kong, the average time from illness to death for fatal cases was 24 days. This means that early estimates of the case fatality risk that ignore the potential outcomes of cases still in hospital are typically underestimates of the final severity profile. We accounted for that by only including cases that either died or recovered in our estimate of the hospital fatality risk. Given that the cases reported outside Wuhan have mostly not been severe, it would be reasonable to infer that there might be a large number of undetected relatively mild infections in Wuhan and that the infection fatality risk is below 1% or even below 0.1%.
There are a number of limitations to our analyses. Most importantly, they are only based on data in the public domain to date. Detailed information has not yet been released by authoritative sources on the most recently reported cases. In our analysis in the Table, support for scenario 2 might increase if contact tracing or medical surveillance was incomplete, if there was incomplete ascertainment of clusters or if there was an early focus in testing cases linked to the Southern China Seafood Wholesale Market that led to a selection bias in the prevalence of market exposures among the early cases.