Rapid communication

Pattern of early human-to-human transmission of

Wuhan 2019 novel coronavirus (2019-nCoV), December
2019 to January 2020
Julien Riou1 , Christian L. Althaus1
1. Institute of Social and Preventive Medicine, University of Bern, Bern, Switzerland
Correspondence: Julien Riou (julien.riou@ispm.unibe.ch)

Citation style for this article:

Riou Julien , Althaus Christian L. . Pattern of early human-to-human transmission of Wuhan 2019 novel coronavirus (2019-nCoV), December 2019 to January 2020.
Euro Surveill. 2020;25(4):pii=2000058. https://doi.org/10.2807/1560-7917.ES.2020.25.4.2000058

Article submitted on 24 Jan 2020 / accepted on 30 Jan 2020 / published on 30 Jan 2020

Since December 2019, China has been experiencing Epidemic parameters

a large outbreak of a novel coronavirus (2019-nCoV) Two key properties will determine further spread
which can cause respiratory disease and severe pneu- of 2019-nCoV. Firstly, the basic reproduction num-
monia. We estimated the basic reproduction num- ber  R0  describes the average number of secondary
ber R0 of 2019-nCoV to be around 2.2 (90% high density cases generated by an infectious index case in a fully
interval: 1.4–3.8), indicating the potential for sus- susceptible population, as was the case during the
tained human-to-human transmission. Transmission early phase of the outbreak. If  R0  is above the critical
characteristics appear to be of similar magnitude to threshold of 1, continuous human-to-human transmis-
severe acute respiratory syndrome-related coronavi- sion with sustained transmission chains will occur.
rus (SARS-CoV) and pandemic influenza, indicating a Secondly, the individual variation in the number of
risk of global spread. secondary cases provides further information about
the expected outbreak dynamics and the potential for
On 31 December 2019, the World Health Organization superspreading events [7-9]. If the dispersion of the
(WHO) was alerted about a cluster of pneumonia of number of secondary cases is high, a small number of
unknown aetiology in the city of Wuhan, China [1,2]. cases may be responsible for a disproportionate num-
Only a few days later, Chinese authorities identified ber of secondary cases, while a large number of cases
and characterised a novel coronavirus (2019-nCoV) will not transmit the pathogen at all. While super-
as the causative agent of the outbreak [3]. The out- spreading always remain a rare event, it can result in
break appears to have started from a single or multiple a large and explosive transmission event and have a
zoonotic transmission events at a wet market in Wuhan lot of impact on the course of an epidemic. Conversely,
where game animals and meat were sold [4] and has low dispersion would lead to a steadier growth of the
resulted in 5,997 confirmed cases in China and 68 con- epidemic, with more homogeneity in the number of
firmed cases in several other countries by 29 January secondary cases per index case. This has important
2020 [5]. Based on the number of exported cases iden- implications for control efforts.
tified in other countries, the actual size of the epidemic
in Wuhan has been estimated to be much larger [6]. At Simulating early outbreak trajectories
this early stage of the outbreak, it is important to gain In a first step, we initialised simulations with one index
understanding of the transmission pattern and the case. For each primary case, we generated second-
potential for sustained human-to-human transmission ary cases according to a negative-binomial offspring
of 2019-nCoV. Information on the transmission char- distribution with mean  R0  and dispersion  k  [7,8]. The
acteristics will help coordinate current screening and dispersion parameter k quantifies the variability in the
containment strategies, support decision making on number of secondary cases, and can be interpreted
whether the outbreak constitutes a public health emer- as a measure of the impact of superspreading events
gency of international concern (PHEIC), and is key for (the lower the value of  k, the higher the impact of
anticipating the risk of pandemic spread of 2019-nCoV. superspreading). The generation time interval  D  was
In order to better understand the early transmission assumed to be gamma-distributed with a shape
pattern of 2019-nCoV, we performed stochastic simula- parameter of 2, and a mean that varied between 7 and
tions of early outbreak trajectories that are consistent 14 days. We explored a wide range of parameter com-
with the epidemiological findings to date. binations (Table) and ran 1,000 stochastic simulations
  for each individual combination. This corresponds to

www.eurosurveillance.org 1
Table Figure 1
Parameter ranges for stochastic simulations of outbreak Values of R0 and k most compatible with the estimated
trajectories, 2019 novel coronavirus outbreak, China, size of the 2019 novel coronavirus epidemic in China, on
2019–2020 18 January 2020

Number of values
Parameter Description Range explored within 0.8 0.8
the range
Basic 0.6 0.6
R0 reproduction 0.8–5.0 22 (equidistant)

Density

Density
number 0.4 0.4
Dispersion 20 (equidistant on
k 0.01–10*
parameter log10 scale)
0.2 0.2
Generation time
D 7–14* 8 (equidistant)
interval (days)
0.0 0.0
Initial number of 1 2 3 4 5
n 1–50 6 (equidistant) 0.01 0.10 1.00 10.00
index cases
R0 Dispersion parameter k
Date of zoonotic 20 Nov–4 Dec Randomised for
T
transmission 2019 each index case The basic reproduction number R0 quantifies human-to-human
transmission. The dispersion parameter k quantifies the risk of
a superspreading event (lower values of k are linked to a higher
probability of superspreading). Note that the probability density of
k implies a log10 transformation.

a total of 3.52 million one-index-case simulations that

were run on UBELIX (http://www.id.unibe.ch/hpc), the
high performance computing cluster at the University density interval: 1.4–3.8) (Figure 1). The observed data
of Bern, Switzerland. at this point are compatible with a large range of val-
ues for the dispersion parameter k (median: 0.54, 90%
In a second step, we accounted for the uncertainty high density interval: 0.014–6.95). However, our simu-
regarding the number of index cases n and the date T of lations suggest that very low values of k are less likely.
the initial zoonotic animal-to-human transmissions at These estimates incorporate the uncertainty about the
the wet market in Wuhan. An epidemic with several total epidemic size on 18 January 2020 and about the
index cases can be considered as the aggregation of date and scale of the initial zoonotic event (Figure 2).
several independent epidemics with one index case
each. We sampled (with replacement)  n  of the one- Comparison with past emergences of
index-case epidemics, sampled a date of onset for respiratory viruses
each index case and aggregated the epidemic curves Comparison with other emerging coronaviruses in
together. The sampling of the date of onset was done the past allows to put into perspective the avail-
uniformly from a 2-week interval around 27 November able information regarding the transmission pat-
2019, in coherence with early phylogenetic analyses terns of 2019-nCoV.  Figure 3  shows the combinations
of 11 2019-nCoV genomes [10]. This step was repeated of  R0  and  k  that are most likely at this stage of the
100 times for each combination of R0 (22 points), k (20 epidemic. Our estimates of R0 and k are more similar to
points),  D  (8 points) and  n  (6 points) for a total of previous estimates focusing on early human-to-human
2,112,000 full epidemics simulated that included the transmission of SARS-CoV in Beijing and Singapore [7]
uncertainty on D, n and T. Finally, we calculated the pro- than of Middle East respiratory syndrome-related coro-
portion of stochastic simulations that reached a total navirus (MERS-CoV) [9]. The spread of MERS-CoV was
number of infected cases within the interval between characterised by small clusters of transmission fol-
1,000 and 9,700 by 18 January 2020, as estimated lowing repeated instances of animal-to-human trans-
by Imai et al. [6]. In a process related to approximate mission events, mainly driven by the occurrence of
Bayesian computation (ABC), the parameter value com- superspreading events in hospital settings. MERS-CoV
binations that led to simulations within that interval could however not sustain human-to-human transmis-
were treated as approximations to the posterior distri- sion beyond a few generations [12]. Conversely, the
butions of the parameters with uniform prior distribu- international spread of SARS-CoV lasted for 9 months
tions. Model simulations and analyses were performed and was driven by sustained human-to-human trans-
in the R software for statistical computing [11]. Code mission, with occasional superspreading events. It
files are available on https://github.com/jriou/wcov. led to more than 8,000 cases around the world and
required extensive efforts by public health authorities
Transmission characteristics of the 2019 to be contained [13]. Our assessment of the early trans-
novel coronavirus mission of 2019-nCoV suggests that 2019-nCoV might
In order to reach between 1,000 and 9,700 infected follow a similar path. 
cases by 18 January 2020, the early human-to-human
transmission of 2019-nCoV was characterised by val- Our estimates for 2019-nCoV are also compatible with
ues of  R0  around 2.2 (median value, with 90% high those of 1918 pandemic influenza, for which  k  was

2 www.eurosurveillance.org
Figure 2
Illustration of the simulation strategy, 2019 novel coronavirus outbreak, China, 2019–2020

12,500

100
Daily incidence
75

50
10,000
25

Nov 20 Nov 27 Dec 4 Dec 11

Uncertainty on epidemic
size on Jan. 18 January
7,500
Cumulative incidence

5,000
Uncertainty on
initial seed

Uncertainty on starting date

2,500

1 Dec 15 Dec 1 Jan 15 Jan

Time

The lines represent the cumulative incidence of 480 simulations with R0 = 1.8 and k = 1.13. The other parameters are left to vary according to
the Table. Among these simulated epidemics, 54.3% led to a cumulative incidence between 1,000 and 9,700 on 18 January 2020 (in red).

estimated [14]. Human-to-human transmission of influ- not yet allow for reliable forecasts of epidemic spread.
enza viruses is characterised by R0 values between 1.5 Case counts provided by local authorities in the early
and 2 and a larger value of  k, implying a more steady stage of an emerging epidemic are notoriously unreli-
transmission without superspreading. The emergence able as reporting rates are unstable and vary with time.
of new strains of influenza, for which human popula- This is due to many factors such as the initial lack of
tions carried little to no immunity contrary to seasonal proper diagnosis tools, the focus on the more severe
influenza, led to pandemics with different sever- cases or the overcrowding of hospitals. We avoided
ity such as the ones in1918, 1957 1968 and 2009. It this surveillance bias by relying on an indirect estimate
is notable that coronaviruses differ from influenza of epidemic size on 18 January, based on cases identi-
viruses in many aspects, and evidence for the 2019- fied in foreign countries before quarantine measures
nCoV with respect to case fatality rate, transmissibility were implemented on 23 January. This estimated range
from asymptomatic individuals and speed of transmis- of epidemic size relies itself on several assumptions,
sion is still limited. Without speculating about possible including that all infected individuals who travelled
consequences, the values of R0 and k found here during from Wuhan to other countries have been detected [6].
the early stage of 2019-nCoV emergence and the lack of This caveat may lead to an underestimation of trans-
immunity to 2019-nCoV in the human population leave missibility, especially considering the recent reports
open the possibility for pandemic circulation of this about asymptomatic cases [4]. Conversely, our results
new virus. do not depend on any assumption about the existence
of asymptomatic transmission, and only reflect the
Strengths and limitations possible combinations of transmission events that lead
The scarcity of available data, especially on case counts to the situation on 18 January.
by date of disease onset as well as contact tracing,
greatly limits the precision of our estimates and does

www.eurosurveillance.org 3
Figure 3
Proportion of simulated epidemics that lead to a cumulative incidence between 1,000 and 9,700 of the 2019 novel
coronavirus outbreak, China, on 18 January 2020

4 0.5

0.4
SARS−CoV − Singapore

0.3
R0

SARS−CoV − Beijing

2 0.2

1918 Influenza 0.1

0.0

MERS−CoV

0.01 0.10 1.00 10.00

Dispersion parameter k

MERS: Middle East respiratory syndrome-related coronavirus; SARS: severe acute respiratory syndrome-related coronavirus.

This can be interpreted as the combinations of R0 and k values most compatible with the estimation of epidemic size before quarantine
measures were put in place. As a comparison, we show the estimates of R0 and k for the early human-to-human transmission of SARS-CoV in
Singapore and Beijing and of 1918 pandemic influenza [7,9,14].

Our analysis, while limited because of the scarcity of pattern. However, values of  k  in the range of 0.1–0.2
data, has two important strengths. Firstly, it is based are still compatible with a small risk of occurrence of
on the simulation of a wide range of possibilities large superspreading events, especially impactful in
regarding epidemic parameters and allows for the full hospital settings [15,16].
propagation on the final estimates of the many remain-
ing uncertainties regarding 2019-nCoV and the situa- Conclusions
tion in Wuhan: on the actual size of the epidemic, on Our analysis suggests that the early pattern of human-
the size of the initial zoonotic event at the wet market, to-human transmission of 2019-nCoV is reminiscent of
on the date(s) of the initial animal-to-human transmis- SARS-CoV emergence in 2002. International collabora-
sion event(s) and on the generation time interval. As tion and coordination will be crucial in order to con-
it accounts for all these uncertainties, our analysis tain the spread of 2019-nCoV. At this stage, particular
provides a summary of the current state of knowledge attention should be given to the prevention of possible
about the human-to-human transmissibility of 2019- rare but explosive superspreading events, while the
nCoV. Secondly, its focus on the possibility of super- establishment of sustained transmission chains from
spreading events by using negative-binomial offspring single cases cannot be ruled out. The previous experi-
distributions appears relevant in the context of emerg- ence with SARS-CoV has shown that established prac-
ing coronaviruses [7,8]. While our estimate of k remains tices of infection control, such as early detection and
imprecise, the simulations suggest that very low values isolation, contact tracing and the use of personal pro-
of  k < 0.1 are less likely than higher values < 0.1 that tective equipment, can stop such an epidemic. Given
correspond to a more homogeneous transmission the existing uncertainty around the case fatality rate

4 www.eurosurveillance.org
 preliminary-phylogenetic-analysis-of-11-ncov2019-
and transmission, our findings confirm the importance genomes-2020-01-19/329
of screening, surveillance and control efforts, particu- 11. R Core Team. R: A language and environment for statistical
larly at airports and other transportation hubs, in order computing. Vienna: R Foundation for Statistical Computing;
2018. Available from: http://www.R-project.org/
to prevent further international spread of 2019-nCoV. 12. Breban R, Riou J, Fontanet A. Interhuman transmissibility of
Middle East respiratory syndrome coronavirus: estimation of
pandemic risk. Lancet. 2013;382(9893):694-9. https://doi.
org/10.1016/S0140-6736(13)61492-0 PMID: 23831141
*Authors’ correction 13. Cowling BJ, Ho LM, Leung GM. Effectiveness of control
measures during the SARS epidemic in Beijing: a comparison
On request of the authors, the ranges for the generation time of the Rt curve and the epidemic curve. Epidemiol
and the dispersion parameter in the Table were corrected on Infect. 2008;136(4):562-6. https://doi.org/10.1017/
17 February 2020. S0950268807008722 PMID: 17568476
14. Fraser C, Cummings DAT, Klinkenberg D, Burke DS, Ferguson
NM. Influenza transmission in households during the 1918
pandemic. Am J Epidemiol. 2011;174(5):505-14. https://doi.
Acknowledgements org/10.1093/aje/kwr122 PMID: 21749971
15. Oh MD, Choe PG, Oh HS, Park WB, Lee S-M, Park J,
JR is funded by the Swiss National Science Foundation (grant et al. Middle EAST respiratory syndrome coronavirus
174281). superspreading event involving 81 persons, Korea 2015. J
Korean Med Sci. 2015;30(11):1701-5. https://doi.org/10.3346/
jkms.2015.30.11.1701 PMID: 26539018
16. Assiri A, McGeer A, Perl TM, Price CS, Al Rabeeah AA,
Conflict of interest Cummings DA, et al. KSA MERS-CoV Investigation Team.
Hospital outbreak of Middle East respiratory syndrome
None declared. coronavirus. N Engl J Med. 2013;369(5):407-16. https://doi.
org/10.1056/NEJMoa1306742 PMID: 23782161

Authors’ contributions
License, supplementary material and copyright
JR and CLA designed the study, JR performed model simula-
tions, JR and CLA analysed and interpreted the results and This is an open-access article distributed under the terms of
wrote the manuscript. the Creative Commons Attribution (CC BY 4.0) Licence. You
may share and adapt the material, but must give appropriate
credit to the source, provide a link to the licence and indicate
if changes were made.
References
1. World Health Organization (WHO). Pneumonia Any supplementary material referenced in the article can be
of unknown cause – China. Geneva: WHO; 2020. found in the online version.
Available from: https://www.who.int/csr/don/05-
january-2020-pneumonia-of-unkown-cause-china/
en/ This article is copyright of the authors or their affiliated in-
2. Huang C, Wang Y, Li X, Ren L, Zhao J, Hu Y, et al. Clinical stitutions, 2020.
features of patients infected with 2019 novel coronavirus in
Wuhan, China.. Lancet. 2020;S0140-6736(20)30183-5. https://
doi.org/10.1016/S0140-6736(20)30154-9 PMID: 31986261
3. Zhou P, Yang X-L, Wang X-G, Hu B, Zhang L, Zhang W, et
al. Discovery of a novel coronavirus associated with the
recent pneumonia outbreak in humans and its potential
bat origin. bioRxiv. 2020;10.1101/2020.01.22.914952.
Available from: https://www.biorxiv.org/content/
early/2020/01/23/2020.01.22.914952
4. Chan JF, Yuan S, Kok KH, To KK, Chu H, Yang J, et al. A
familial cluster of pneumonia associated with the 2019 novel
coronavirus indicating person-to-person transmission: a
study of a family cluster. Lancet. 2020;S0140-6736(20)30154-
9. https://doi.org/10.1016/S0140-6736(20)30154-9 PMID:
31986261
5. World Health Organization (WHO). Novel Coronavirus
(2019-nCoV) situation report 9. Geneva: WHO; 2020.
Available from: https://www.who.int/docs/default-source/
coronaviruse/situation-reports/20200129-sitrep-9-ncov-v2.
pdf?sfvrsn=e2c8915_2
6. Imai N, Dorigatti I, Cori A, Donnelly C, Riley S, Ferguson
NM. Report 2: Estimating the potential total number of
novel Coronavirus cases in Wuhan City, China. London:
Imperial College; 2020. Available from: https://www.
imperial.ac.uk/mrc-global-infectious-disease-analysis/
news--wuhan-coronavirus/
7. Lloyd-Smith JO, Schreiber SJ, Kopp PE, Getz WM.
Superspreading and the effect of individual variation on
disease emergence. Nature. 2005;438(7066):355-9. https://
doi.org/10.1038/nature04153 PMID: 16292310
8. Althaus CL. Ebola superspreading. Lancet Infect Dis.
2015;15(5):507-8. https://doi.org/10.1016/S1473-
3099(15)70135-0 PMID: 25932579
9. Kucharski AJ, Althaus CL. The role of superspreading in
Middle East respiratory syndrome coronavirus (MERS-CoV)
transmission. Euro Surveill. 2015;20(25):14-8. https://doi.
org/10.2807/1560-7917.ES2015.20.25.21167 PMID: 26132768
10. Rambaut A. Preliminary phylogenetic analysis of 11
nCoV2019 genomes, 2020-01-19. ARTIC Network;
2020. Available from: http://virological.org/t/

www.eurosurveillance.org 5