Articles

Clinical course and risk factors for mortality of adult

inpatients with COVID-19 in Wuhan, China: a retrospective
cohort study
Fei Zhou*, Ting Yu*, Ronghui Du*, Guohui Fan*, Ying Liu*, Zhibo Liu*, Jie Xiang*, Yeming Wang, Bin Song, Xiaoying Gu, Lulu Guan, Yuan Wei,
Hui Li, Xudong Wu, Jiuyang Xu, Shengjin Tu, Yi Zhang, Hua Chen, Bin Cao

Summary
Background Since December, 2019, Wuhan, China, has experienced an outbreak of coronavirus disease 2019 Published Online
(COVID-19), caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Epidemiological and March 9, 2020
https://doi.org/10.1016/
clinical characteristics of patients with COVID-19 have been reported but risk factors for mortality and a detailed
S0140-6736(20)30566-3
clinical course of illness, including viral shedding, have not been well described.
This online publication has
been corrected. The corrected
Methods In this retrospective, multicentre cohort study, we included all adult inpatients (≥18 years old) with laboratory- version first appeared at
confirmed COVID-19 from Jinyintan Hospital and Wuhan Pulmonary Hospital (Wuhan, China) who had been thelancet.com on
March 12, 2020
discharged or had died by Jan 31, 2020. Demographic, clinical, treatment, and laboratory data, including serial
samples for viral RNA detection, were extracted from electronic medical records and compared between survivors *Contributed equally

and non-survivors. We used univariable and multivariable logistic regression methods to explore the risk factors Department of Pulmonary and
Critical Care Medicine, Center
associated with in-hospital death. of Respiratory Medicine,
National Clinical Research
Findings 191 patients (135 from Jinyintan Hospital and 56 from Wuhan Pulmonary Hospital) were included in this Center for Respiratory Diseases,
study, of whom 137 were discharged and 54 died in hospital. 91 (48%) patients had a comorbidity, with hypertension Institute of Respiratory
Medicine, Chinese Academy of
being the most common (58 [30%] patients), followed by diabetes (36 [19%] patients) and coronary heart disease Medical Sciences, Peking Union
(15 [8%] patients). Multivariable regression showed increasing odds of in-hospital death associated with older age Medical College, Beijing, China
(odds ratio 1·10, 95% CI 1·03–1·17, per year increase; p=0·0043), higher Sequential Organ Failure Assessment (SOFA) (F Zhou MD, G Fan MS, Z Liu MD,
score (5·65, 2·61–12·23; p<0·0001), and d-dimer greater than 1 µg/mL (18·42, 2·64–128·55; p=0·0033) on admission. Y Wang MD, X Gu PhD, H Li MD,
Y Zhang MD, Prof B Cao MD);
Median duration of viral shedding was 20·0 days (IQR 17·0–24·0) in survivors, but SARS-CoV-2 was detectable until Department of Tuberculosis
death in non-survivors. The longest observed duration of viral shedding in survivors was 37 days. and Respiratory Disease
(T Yu MD, Y Liu MD, B Song MS,
Interpretation The potential risk factors of older age, high SOFA score, and d-dimer greater than 1 µg/mL could help Y Wei MS, S Tu MD,
Prof H Chen MD) and
clinicians to identify patients with poor prognosis at an early stage. Prolonged viral shedding provides the rationale Department of Clinical
for a strategy of isolation of infected patients and optimal antiviral interventions in the future. Laboratory (J Xiang MS), and
GCP Center (X Wu MS),
Jinyintan Hospital, Wuhan,
Funding Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences; National Science Grant for
China; Department of
Distinguished Young Scholars; National Key Research and Development Program of China; The Beijing Science and Pulmonary and Critical Care
Technology Project; and Major Projects of National Science and Technology on New Drug Creation and Development. Medicine, Wuhan Pulmonary
Hospital, Wuhan, China
(R Du MD, L Guan MD);
Copyright © 2020 Elsevier Ltd. All rights reserved.
Department of Pulmonary and
Critical Care Medicine
Introduction asymptomatic infection, mild upper respiratory tract (Prof B Cao) and Institute of
In December, 2019, Wuhan city, the capital of Hubei illness, and severe viral pneumonia with respiratory Clinical Medical Sciences (G Fan,
X Gu), China-Japan Friendship
province in China, became the centre of an outbreak of failure and even death, with many patients being
Hospital, Beijing, China;
pneumonia of unknown cause. By Jan 7, 2020, Chinese hospitalised with pneumonia in Wuhan.5–7 Although some Department of Respiratory
scientists had isolated a novel coronavirus, severe acute case series have been published, many patients in these Medicine, Capital Medical
respiratory syndrome coronavirus 2 (SARS-CoV-2; previ­ series remained hospitalised at time of publication. To our University, Beijing, China
(Y Wang, Prof B Cao); and
ously known as 2019-nCoV), from these patients with knowledge, no previous studies have been done among Tsinghua University School of
virus-infected pneumonia,1,2 which was later designated patients with definite outcomes. The estimation of risk Medicine, Beijing, China
coronavirus disease 2019 (COVID-19) in February, 2020, factors for severe disease and death in these earlier case (J Xu MD, Prof B Cao)
by WHO.3 series are therefore not very robust. Additionally, details of Correspondence to:
Although the outbreak is likely to have started from a the clinical and virological course of illness have not yet Prof Bin Cao, Department of
Pulmonary and Critical Care
zoonotic transmission event associated with a large been well described.
Medicine, China-Japan
seafood market that also traded in live wild animals, it Here, we present details of all patients admitted Friendship Hospital, Beijing
soon became clear that efficient person-to-person trans­ to the two designated hospitals in Wuhan—Jinyintan 100029, China
mission was also occurring.4 The clinical spectrum of Hospital and Wuhan Pulmonary Hospital—with labo­ [email protected]
SARS-CoV-2 infection appears to be wide, encompassing ratory-confirmed COVID-19 and a definite clinical or

www.thelancet.com Published online March 9, 2020 https://doi.org/10.1016/S0140-6736(20)30566-3 1

 Articles

Prof Hua Chen, Department of

Tuberculosis and Respiratory Research in context
Disease, Jinyintan Hospital,
Dongxihu District, Wuhan Evidence before this study Added value of this study
430023, China We searched PubMed on Feb 23, 2020, for articles that In this retrospective cohort study of adult inpatients in
[email protected] documented the risk factors of mortality and viral shedding two hospitals in Wuhan, China, we found increasing odds of
in patients with coronavirus disease 2019 (COVID-19), in-hospital death associated with older age (odds ratio 1·10,
resulting from infection with severe acute respiratory 95% CI 1·03–1·17; p=0·0043), higher Sequential Organ Failure
syndrome coronavirus 2 (SARS-CoV-2), using the search Assessment (SOFA) score (5·65, 2·61–12·23; p<0·0001),
terms (“novel coronavirus” OR “SARS-CoV-2” OR “COVID-19”) and d-dimer levels greater than 1·0 µg/mL (18·42, 2·64–128·55;
AND (“death” OR “mortality” OR “viral shedding”) with no p=0·0033) on admission. Duration of viral shedding ranged
language or time restrictions. Age, comorbidities, between 8 and 37 days. The median duration of viral shedding
lymphocytopenia and elevated alanine aminotransferase, was 20·0 days (IQR 17·0–24·0) in survivors, but continued until
d-dimer, creatine kinase, high-sensitivity cardiac troponin I, death in fatal cases.
prothrombin time, and disease severity were reported to be
Implications of all the available evidence
associated with intensive care unit admission. However,
Older age, elevated d-dimer levels, and high SOFA score could
no published works were found about the risk factors of
help clinicians to identify at an early stage those patients with
mortality for adult patients with COVID-19. One study
COVID-19 who have poor prognosis. Prolonged viral shedding
compared the sensitivity of SARS-CoV-2 RNA detection in
provides the rationale for a strategy of isolation of infected
throat and nasopharyngeal swab in 17 patients with
patients and optimal antiviral interventions in the future.
COVID-19.

outcome (death or discharge) as of Jan 31, 2020. We aim form, which was a modified version of the WHO/
to explore risk factors of in-hospital death for patients International Severe Acute Respiratory and Emerging
and describe the clinical course of symptoms, viral Infection Consortium case record form for severe
shedding, and tem­poral changes of laboratory findings acute respiratory infections. All data were checked by
during hospitalisation. two physicians (FZ and ZL) and a third researcher (GF)
adjudicated any difference in interpretation between the
Methods two primary reviewers.
Study design and participants
This retrospective cohort study included two cohorts of Laboratory procedures
adult inpatients (≥18 years old) from Jinyintan Hospital and Methods for laboratory confirmation of SARS-CoV-2
Wuhan Pulmonary Hospital (Wuhan, China). All adult infection have been described elsewhere.5 Briefly, four
patients who were diagnosed with COVID-19 according to institutions—the Chinese Center for Disease Control
WHO interim guidance were screened, and those who died and Prevention, the Chinese Academy of Medical
or were discharged between Dec 29, 2019 (ie, when the first Science, the Academy of Military Medical Sciences, and
patients were admitted), and Jan 31, 2020, were included in the Wuhan Institute of Virology, Chinese Academy of
our study. Since these two hospitals were the only desig­ Sciences—were responsible for SARS-CoV-2 detection
nated hospitals for transfer of patients with COVID-19 in respiratory specimens by next-generation sequencing
from other hospitals in Wuhan until Feb 1, 2020, our study or real-time RT-PCR methods. From Jan 11, 2020,
enrolled all adult inpatients who were hospitalised for SARS-CoV-2 RNA were detected by local Centers for
COVID-19 and had a definite outcome (dead or discharged) Disease Control and Prevention, local health institutions,
at the early stage of the outbreak. and Jingyintan Hospital and Wuhan Pulmonary Hospital.
Before Jan 11, 2020, SARS-CoV-2 RNA detection results Throat-swab specimens were obtained for SARS-CoV-2
were not available in the electronic medical records, from PCR re-examination every other day after clinical
which data for this study were obtained retrospectively; remission of symptoms, including fever, cough, and
therefore, this study includes 29 of the 41 patients dyspnoea, but only qualitative data were available. The
originally reported on.5 criteria for discharge were absence of fever for at least
The study was approved by the Research Ethics 3 days, substantial improvement in both lungs in chest
Commission of Jinyintan Hospital (KY-2020–01.01) and CT, clinical remission of respiratory symptoms, and two
the requirement for informed consent was waived by the throat-swab samples negative for SARS-CoV-2 RNA
Ethics Commission as described previously.5 obtained at least 24 h apart.
Routine blood examinations were complete blood
Data collection count, coagulation profile, serum biochemical tests
Epidemiological, demographic, clinical, laboratory, treat­ (including renal and liver function, creatine kinase,
ment, and outcome data were extracted from electronic lactate dehydrogenase, and electrolytes), myocardial
medical records using a standardised data collection enzymes, interleukin-6 (IL-6), serum ferritin, and

2 www.thelancet.com Published online March 9, 2020 https://doi.org/10.1016/S0140-6736(20)30566-3

 Articles

procalcitonin. Chest radiographs or CT scan were also

done for all inpatients. Frequency of examinations was
determined by the treating physician.
Total Non-survivor Survivor p value
(n=191) (n=54) (n=137)
Definitions
Demographics and clinical characteristics
Fever was defined as axillary temperature of at least
Age, years 56·0 (46·0–67·0) 69·0 (63·0–76·0) 52·0 (45·0–58·0) <0·0001
37·3°C. Sepsis and septic shock were defined according
Sex ·· ·· ·· 0·15
to the 2016 Third International Consensus Definition
for Sepsis and Septic Shock.5 Secondary infection was Female 72 (38%) 16 (30%) 56 (41%) ··

diagnosed when patients showed clinical symptoms or Male 119 (62%) 38 (70%) 81 (59%) ··

signs of pneumonia or bacteraemia and a positive culture Exposure history 73 (38%) 14 (26%) 59 (43%) 0·028
of a new pathogen was obtained from lower respiratory Current smoker 11 (6%) 5 (9%) 6 (4%) 0·21
tract specimens (qualified sputum, endotracheal aspirate, Comorbidity 91 (48%) 36 (67%) 55 (40%) 0·0010
or bronchoalveolar lavage fluid) or blood samples Hypertension 58 (30%) 26 (48%) 32 (23%) 0·0008
after admission.5 Ventilator-associated pneumonia was Diabetes 36 (19%) 17 (31%) 19 (14%) 0·0051
diagnosed according to the guidelines for treatment of Coronary heart disease 15 (8%) 13 (24%) 2 (1%) <0·0001
hospital-acquired and ventilator-associated pneumonia.8 Chronic obstructive lung 6 (3%) 4 (7%) 2 (1%) 0·047
disease
Acute kidney injury was diagnosed according to the
KDIGO clinical practice guidelines9 and acute respiratory Carcinoma 2 (1%) 0 2 (1%) 0·37

distress syndrome (ARDS) was diagnosed according to Chronic kidney disease 2 (1%) 2 (4%) 0 0·024

the Berlin Definition.10 Acute cardiac injury was diag­ Other 22 (12%) 11 (20%) 11 (8%) 0·016
nosed if serum levels of cardiac biomarkers (eg, high- Respiratory rate 56 (29%) 34 (63%) 22 (16%) <0·0001
>24 breaths per min
sensitivity cardiac troponin I) were above the 99th
Pulse ≥125 beats per min 2 (1%) 2 (4%) 0 0·024
percentile upper reference limit, or if new abnormalities
Systolic blood pressure 1 (1%) 0 1 (1%) 0·53
were shown in electrocardiography and echocardiography.5 <90 mm Hg
The illness severity of COVID-19 was defined according Fever (temperature ≥37·3°C) 180 (94%) 51 (94%) 129 (94%) 0·94
to the Chinese management guideline for COVID-19
Cough 151 (79%) 39 (72%) 112 (82%) 0·15
(version 6.0).11 Coagulopathy was defined as a 3-second
Sputum 44 (23%) 14 (26%) 30 (22%) 0·55
extension of pro­thrombin time or a 5-second extension of
Myalgia 29 (15%) 8 (15%) 21 (15%) 0·93
activated partial thromboplastin time. Hypoproteinaemia
Fatigue 44 (23%) 15 (28%) 29 (21%) 0·33
was defined as blood albumin of less than 25 g/L.
Diarrhoea 9 (5%) 2 (4%) 7 (5%) 0·67
Exposure history was defined as exposure to people with
Nausea or vomiting 7 (4%) 3 (6%) 4 (3%) 0·40
confirmed SARS-CoV-2 infection or to the Wuhan seafood
SOFA score 2·0 (1·0–4·0) 4·5 (4·0–6·0) 1·0 (1·0–2·0) <0·0001
market.
qSOFA score 1·0 (0·0–1·0) 1·0 (1·0–1·0) 0·0 (0·0–1·0) <0·0001
CURB-65 score 0·0 (0·0–2·0) 2·0 (1·0–3·0) 0·0 (0·0–1·0) <0·0001
Statistical analysis
0–1 141/188 (75%) 16 (30%) 125/134 (93%) <0·0001*
Continuous and categorical variables were presented as
2 32/188 (17%) 23 (43%) 9/134 (7%) ··
median (IQR) and n (%), respectively. We used the Mann-
Whitney U test, χ² test, or Fisher’s exact test to compare 3–5 15/188 (8%) 15 (28%) 0/134 ··

differences between survivors and non-survivors where Disease severity status ·· ·· ·· <0·0001
appropriate. To explore the risk factors associated with General 72 (38%) 0 72 (53%) ··
in-hospital death, univariable and multivariable logistic Severe 66 (35%) 12 (22%) 54 (39%) ··
regression models were used. Considering the total Critical 53 (28%) 42 (78%) 11 (8%) ··
number of deaths (n=54) in our study and to avoid Time from illness onset to 11·0 (8·0–14·0) 11·0 (8·0–15·0) 11·0 (8·0–13·0) 0·53
hospital admission, days
overfitting in the model, five variables were chosen for
Laboratory findings
multivariable analysis on the basis of previous findings and
White blood cell count, 6·2 (4·5–9·5) 9·8 (6·9–13·9) 5·2 (4·3–7·7) <0·0001
clinical constraints. Previous studies have shown blood
× 10⁹ per L
levels of d-dimer and Sequential Organ Failure Assessment
<4 32 (17%) 5 (9%) 27 (20%) <0·0001*
(SOFA) scores to be higher in critically ill or fatal cases,
4–10 119 (62%) 24 (44%) 95 (69%) ··
whereas lymphopenia and cardiovascular disease have
>10 40 (21%) 25 (46%) 15 (11%) ··
been less commonly observed in non-critical or surviving
Lymphocyte count, 1·0 (0·6–1·3) 0·6 (0·5–0·8) 1·1 (0·8–1·5) <0·0001
patients with SARS-COV-2 infection.5,6,12 Similar risk × 10⁹ per L
factors, including older age, have been reported associated <0·8 77 (40%) 41 (76%) 36 (26%) <0·0001
with adverse clinical outcomes in adults with SARS and Haemoglobin, g/L 128·0 126·0 128·0 0·30
Middle East respiratory syndrome (MERS).3,13 Some (119·0–140·0) (115·0–138·0) (120·0–140·0)
laboratory find­ings, including alanine amino­ transferase (Table 1 continues on next page)
(ALT), lactate dehydrogenase, high-sensitivity cardiac

www.thelancet.com Published online March 9, 2020 https://doi.org/10.1016/S0140-6736(20)30566-3 3

 Articles

We compared patient characteristics between the

Total Non-survivor Survivor p value
(n=191) (n=54) (n=137) two hospitals and used a generalised linear model to
adjust for possible differences in patients’ characteristics
(Continued from previous page)
and treatment between the two study centres.
Anaemia 29 (15%) 14 (26%) 15 (11%) 0·0094
A two-sided α of less than 0·05 was considered statis­
Platelet count, × 10⁹ per L 206·0 165·5 220·0 <0·0001
tically significant. Statistical analyses were done using
(155·0–262·0) (107·0–229·0) (168·0–271·0)
the SAS software (version 9.4), unless otherwise
<100 13 (7%) 11 (20%) 2 (1%) <0·0001
indicated.
Albumin, g/L 32·3 (29·1–35·8) 29·1 (26·5–31·3) 33·6 (30·6–36·4) <0·0001
ALT, U/L 30·0 (17·0–46·0) 40·0 (24·0–51·0) 27·0 (15·0–40·0) 0·0050
Role of the funding source
>40 59/189 (31%) 26 (48%) 33/135 (24%) 0·0015
The funder of the study had no role in study design, data
Creatinine >133 μmol/L 8/186 (4%) 5 (9%) 3/132 (2%) 0·045
collection, data analysis, data interpretation, or writing
Lactate dehydrogenase, U/L 300·0 521·0 253·5 <0·0001
(234·0–407·0) (363·0–669·0) (219·0–318·0)
of the report. The corresponding authors (BC and HC)
>245 123/184 (67%) 53 (98%) 70/130(54%) <0·0001
had full access to all the data in the study and had final
responsibility for the decision to submit for publication.
Creatine kinase, U/L 21·5 (13·0–72·4) 39·0 (19·5–151·0) 18·0 (12·5–52·1) 0·0010
>185 22/168 (13%) 11/52 (21%) 11/116 (9%) 0·038
High-sensitivity cardiac 4·1 (2·0–14·1) 22·2 (5·6–83·1) 3·0 (1·1–5·5) <0·0001
Results
troponin I, pg/mL 813 adult patients were hospitalised in Jinyintan Hospital
>28 24/145 (17%) 23/50 (46%) 1/95 (1%) <0·0001 or Wuhan Pulmonary Hospital with COVID-19 before
Prothrombin time, s 11·6 (10·6–13·0) 12·1 (11·2–13·7) 11·4 (10·4–12·6) 0·0004 Jan 31, 2020. After excluding 613 patients that were
<16 171/182 (94%) 47 (87%) 124/128 (97%) 0·016*
still hospitalised or not confirmed by SARS-CoV-2 RNA
≥16 11/182 (6%) 7 (13%) 4/128 (3%) ··
detection as of Jan 31, 2020, and nine inpatients without
D-dimer, μg/mL 0·8 (0·4–3·2) 5·2 (1·5–21·1) 0·6 (0·3–1·0) <0·0001
available key information in their medical records, we
≤0·5 55/172 (32%) 4 (7%) 51/118 (43%) <0·0001*
included 191 inpatients (135 from Jinyintan Hospital and
56 from Wuhan Pulmonary Hospital) in the final
>0·5 to ≤1 45/172 (26%) 6 (11%) 39/118 (33%) ··
analysis. 54 patients died during hospitalisation and
>1 72/172 (42%) 44 (81%) 28/118 (24%) ··
137 were discharged. The median age of the 191 patients
Serum ferritin, μg/L 722·0 1435·3 503·2 <0·0001
(377·2–1435·3) (728·9–2000·0) (264·0–921·5) was 56·0 years (IQR 46·0–67·0), ranging from 18 years
>300 102/128 (80%) 44/46 (96%) 58/82 (71%) 0·0008 to 87 years, and most patients were male (table 1).
IL-6, pg/mL 7·4 (5·3–10·8) 11·0 (7·5–14·4) 6·3 (5·0–7·9) <0·0001 Co­morbidities were present in nearly half of patients,
Procalcitonin, ng/mL 0·1 (0·1–0·1) 0·1 (0·1–0·5) 0·1 (0·1–0·1) <0·0001
with hypertension being the most common comorbidity,
<0·1 114/164 (70%) 19/51 (37%) 95/113 (84%) <0·0001*
followed by diabetes and coronary heart disease (table 1).
≥0·1 to <0·25 30/164 (18%) 16/51 (31%) 14/113 (12%) ··
The most common symptoms on admission were
≥0·25 to <0·5 6/164 (4%) 3/51 (6%) 3/113 (3%) ··
fever and cough, followed by sputum production and
fatigue (table 1). Lymphocytopenia occurred in 77 (40%)
≥0·5 14/164 (9%) 13/51 (25%) 1/113 (1%) ··
patients. 181 (95%) patients received antibiotics and
Imaging features
41 (21%) received antivirals (lopinavir/ritonavir; table 2).
Consolidation 112 (59%) 40 (74%) 72 (53%) 0·0065
Systematic corticosteroid and intravenous immuno­
Ground-glass opacity 136 (71%) 44 (81%) 92 (67%) 0·049
globulin use differed significantly between non-survivors
Bilateral pulmonary 143 (75%) 45 (83%) 98 (72%) 0·090
infiltration and survivors (table 2). The comparison of charac­
teristics, treatment, and outcomes of patients from the
Data are median (IQR), n (%), or n/N (%). p values were calculated by Mann-Whitney U test, χ² test, or Fisher’s exact
test, as appropriate. SOFA=Sequential Organ Failure Assessment. qSOFA=Quick SOFA. ALT=alanine aminotransferase.
two hospitals are shown in the appendix (pp 2–4).
IL-6=interleukin-6. *χ² test comparing all subcategories. The median time from illness onset (ie, before admission)
to discharge was 22·0 days (IQR 18·0–25·0), whereas the
Table 1: Demographic, clinical, laboratory, and radiographic findings of patients on admission
median time to death was 18·5 days (15·0–22·0; table 2).
32 patients required invasive mechanical ventilation, of
See Online for appendix troponin I, creatine kinase, d-dimer, serum ferritin, and whom 31 (97%) died. The median time from illness onset to
IL-6, might be unavailable in emergency circumstances. invasive mechanical ventilation was 14·5 days (12·0–19·0).
Therefore, we chose lymphocyte count, d-dimer, SOFA Extracorporeal membrane oxygenation was used in three
score, coronary heart disease, and age as the five variables patients, none of whom survived. Sepsis was the most
for our multivariable logistic regression model. frequently observed complication, followed by respiratory
We excluded variables from the univariable analysis if failure, ARDS, heart failure, and septic shock (table 2). Half
their between-group differences were not significant, if of non-survivors experienced a secondary infection, and
their accuracy was unconfirmed (eg, exposure, which ventilator-associated pneumonia occurred in ten (31%) of
was self-reported), if the number of events was too small 32 patients requiring invasive mechanical ventilation. The
to calculate odds ratios, and if they had colinearity with frequency of complications were higher in non-survivors
the SOFA score. than survivors (table 2).

4 www.thelancet.com Published online March 9, 2020 https://doi.org/10.1016/S0140-6736(20)30566-3

 Articles

In univariable analysis, odds of in-hospital death was

Total Non-survivor Survivor p value
higher in patients with diabetes or coronary heart disease (n=191) (n=54) (n=137)
(table 3). Age, lymphopenia, leucocytosis, and elevated
Treatments*
ALT, lactate dehydrogenase, high-sensitivity cardiac
Antibiotics 181 (95%) 53 (98%) 128 (93%) 0·15
troponin I, creatine kinase, d-dimer, serum ferritin, IL-6,
Antiviral treatment 41 (21%) 12 (22%) 29 (21%) 0·87
prothrombin time, creatinine, and procalcitonin were
Corticosteroids 57 (30%) 26 (48%) 31 (23%) 0·0005
also associated with death (table 3).
Intravenous immunoglobin 46 (24%) 36 (67%) 10 (7%) <0·0001
We included 171 patients with complete data for all
High-flow nasal cannula 41 (21%) 33 (61%) 8 (6%) <0·0001
variables (53 non-survivors and 118 survivors) in the
oxygen therapy
multivariable logistic regression model. We found that
Non-invasive mechanical 26 (14%) 24 (44%) 2 (1%) <0·0001
older age, higher SOFA score, and d-dimer greater than ventilation
1 μg/mL at admission were associated with increased Invasive mechanical ventilation 32 (17%) 31 (57%) 1 (1%) <0·0001
odds of death (table 3). When adjusting for study centre, ECMO 3 (2%) 3 (6%) 0 0·0054
our generalised linear model showed similar results Renal replacement therapy 10 (5%) 10 (19%) 0 <0·0001
(appendix p 5). Outcomes
For survivors, the median duration of viral shedding Sepsis 112 (59%) 54 (100%) 58 (42%) <0·0001
was 20·0 days (IQR 17·0–24·0) from illness onset, but the Respiratory failure 103 (54%) 53 (98%) 50 (36%) <0·0001
virus was continuously detectable until death in non- ARDS 59 (31%) 50 (93%) 9 (7%) <0·0001
survivors (table 2; figure 1). The shortest observed duration
Heart failure 44 (23%) 28 (52%) 16 (12%) <0·0001
of viral shedding among survivors was 8 days, whereas the
Septic shock 38 (20%) 38 (70%) 0 <0·0001
longest was 37 days. Among 29 patients who received
Coagulopathy 37 (19%) 27 (50%) 10 (7%) <0·0001
lopinavir/ritonavir and were discharged, the median time
Acute cardiac injury 33 (17%) 32 (59%) 1 (1%) <0·0001
from illness onset to initiation of antiviral treatment was
Acute kidney injury 28 (15%) 27 (50%) 1 (1%) <0·0001
14·0 days (IQR 10·0–17·0) and the median duration of
Secondary infection 28 (15%) 27 (50%) 1 (1%) <0·0001
viral shedding was 22·0 days (18·0–24·0). The median
Hypoproteinaemia 22 (12%) 20 (37%) 2 (1%) <0·0001
duration of viral shedding was 19·0 days (17·0–22·0) in
Acidosis 17 (9%) 16 (30%) 1 (1%) <0·0001
patients with severe disease status and 24·0 days
ICU admission 50 (26%) 39 (72%) 11 (8%) <0·0001
(22·0–30·0) in patients with critical disease status.
ICU length of stay, days 8·0 (4·0–12·0) 8·0 (4·0–12·0) 7·0 (2·0–9·0) 0·41
Major laboratory markers were tracked from illness
onset (figure 2). Baseline lymphocyte count was signifi­ Hospital length of stay, days 11·0 (7·0–14·0) 7·5 (5·0–11·0) 12·0 (9·0–15·0) <0·0001

cantly higher in survivors than non-survivors; in survivors, Time from illness onset to 1·0 (1·0–1·0) 1·0 (1·0–1·0) 1·0 (1·0–1·0) 0·16
fever, days
lymphocyte count was lowest on day 7 after illness onset
Time from illness onset to 1·0 (1·0–3·0) 1·0 (1·0–1·0) 1·0 (1·0–4·0) 0·30
and improved during hospitalisation, whereas severe cough, days
lymphopenia was observed until death in non-survivors. Time from illness onset to 7·0 (4·0–9·0) 7·0 (4·0–10·0) 7·0 (4·0–9·0) 0·51
Levels of d-dimer, high-sensitivity cardiac troponin I, dyspnoea, days
serum ferritin, lactate dehydrogenase, and IL-6 were Time from illness onset to 9·0 (7·0–13·0) 10·0 (7·0–14·0) 9·0 (7·0–12·0) 0·22
clearly elevated in non-survivors compared with survivors sepsis, days
throughout the clinical course, and increased with illness Time from illness onset to 12·0 (8·0–15·0) 12·0 (8·0–15·0) 10·0 (8·0–13·0) 0·65
ARDS, days
deterioration (figure 2). In non-survivors, high-sensitivity
cardiac troponin I increased rapidly from day 16 after Time from illness onset to ICU 12·0 (8·0–15·0) 12·0 (8·0–15·0) 11·5 (8·0–14·0) 0·88
admission, days
disease onset, whereas lactate dehydrogenase increased
Time from illness onset to 12·0 (10·0–16·0) 13·0 (10·0–17·0) 12·0 (10·0–15·0) 0·55
for both survivors and non-survivors in the early stage of corticosteroids treatment, days
illness, but decreased from day 13 for survivors. Time from illness onset to 21·0 (17·0–25·0) 18·5 (15·0–22·0) 22·0 (18·0–25·0) 0·0003
Median time from illness onset to dyspnoea was similar death or discharge, days
in survivors and non-survivors, with a median duration of Duration of viral shedding after 20·0 (16·0–23·0) 18·5 (15·0–22·0)† 20·0 (17·0–24·0) 0·024
dyspnoea of 13·0 days (9·0–16·5) for survivors (table 2; COVID-19 onset, days
figure 1). In survivors, the median duration of fever was Data are median (IQR) or n (%). p values were calculated by Mann-Whitney U test, χ² test, or Fisher’s exact test,
12·0 days (8·0–13·0) and cough persisted for 19·0 days as appropriate. ECMO=extracorporeal membrane oxygenation. ARDS=acute respiratory distress syndrome.
ICU=intensive care unit. COVID-19=coronavirus disease 2019. *Ordered by escalating scale of respiratory support.
(IQR 12·0–23·0; figure 1). 62 (45%) survivors still had
†Detectable until death.
cough on discharge and 39 (72%) non-survivors still had
cough at the time of death. The dynamic profiles of fever, Table 2: Treatments and outcomes
cough, and dyspnoea are shown in the appendix (p 6).
Sepsis developed at a median of 9·0 days (7·0–13·0) after initiation time and duration of systematic corticosteroid
illness onset among all patients, followed by ARDS use was also similar between the two groups. Among
(12·0 days [8·0–15·0]), acute cardiac injury (15·0 days non-survivors, the median time from illness onset was
[10·0–17·0]), acute kidney injury (15·0 days [13·0–19·5]), 10·0 days (7·0–14·0) to sepsis, 12·0 days (8·0–15·0) to
and secondary infection (17·0 days [13·0–19·0]). The ARDS, 14·5 days (9·5–17·0) to acute cardiac injury, and

www.thelancet.com Published online March 9, 2020 https://doi.org/10.1016/S0140-6736(20)30566-3 5

 Articles

Univariable OR p value Multivariable p value Univariable OR p value Multivariable p value

(95% CI) OR (95% CI) (95% CI) OR (95% CI)
Demographics and clinical characteristics (Continued from previous column)
Age, years* 1·14 <0·0001 1·10 0·0043 Creatinine, μmol/L
(1·09–1·18) (1·03–1·17)
≤133 1 (ref) ·· ·· ··
Female sex 0·61 0·15 ·· ··
>133 4·39 0·048 ·· ··
(vs male) (0·31–1·20)
(1·01–19·06)
Current smoker 2·23 0·20 ·· ··
Lactate dehydrogenase, U/L
(vs non- (0·65–7·63)
smoker) ≤245 1 (ref) ·· ·· ··
Comorbidity present (vs not present) >245 45·43 0·0002 ·· ··
(6·10–338·44)
Chronic 5·40 0·056 ·· ··
obstructive (0·96–30·40) Creatine kinase, U/L
lung disease ≤185 1 (ref) ·· ·· ··
Coronary 21·40 <0·0001 2·14 0·48 >185 2·56 0·043 ·· ··
heart disease (4·64–98·76) (0·26–17·79) (1·03–6·36)
Diabetes 2·85 0·0062 ·· ·· High-sensitivity cardiac troponin I, pg/mL
(1·35–6·05)
≤28 1 (ref) ·· ·· ··
Hypertension 3·05 0·0010 ·· ··
>28 80·07 <0·0001 ·· ··
(1·57–5·92)
(10·34–620·36)
Respiratory rate, breaths per min
D-dimer, μg/mL
≤24 1 (ref) ·· ·· ··
≤0·5 1 (ref) ·· 1 (ref) ··
>24 8·89 <0·0001 ·· ··
> 0·5 1·96 0·32 2·14 0·52
(4·34–18·19)
(0·52–7·43) (0·21–21·39)
SOFA score 6·14 <0·0001 5·65 <0·0001
>1 20·04 <0·0001 18·42 0·0033
(3·48–10·85) (2·61–12·23)
(6·52–61·56) (2·64–128·55)
qSOFA score 12·00 <0·0001 ·· ··
Prothrombin time, s
(5·06–28·43)
<16 1 (ref) ·· ·· ··
Laboratory findings
≥16 4·62 0·019 ·· ··
White blood cell count, × 10⁹ per L
(1·29–16·50)
<4 0·73 0·56 ·· ··
Serum ferritin, μg/L
(0·26–2·10)
≤300 1 (ref) ·· ·· ··
4–10 1 (ref) ·· ·· ··
>300 9·10 0·0038 ·· ··
>10 6·60 <0·0001 ·· ··
(2·04–40·58)
(3·02–14·41)
IL-6, pg/mL* 1·12 0·0080 ·· ··
Lymphocyte 0·02 <0·0001 0·19 0·13
(1·03–1·23)
count, (0·01–0·08) (0·02–1·62)
× 10⁹ per L* Procalcitonin, 13·75 0·011 ·· ··
ng/mL* (1·81–104·40)
ALT, U/L
≤40 1 (ref) ·· ·· ·· OR=odds ratio. S­OFA=Sequential Organ Failure Assessment. qSOFA=Quick SOFA.
ALT=alanine aminotransferase. IL-6=interleukin-6. *Per 1 unit increase.
>40 2·87 0·0018 ·· ··
(1·48–5·57) Table 3: Risk factors associated with in-hospital death
(Table 3 continues in next column)

hospitalised with COVID-19. In particular, older age,

17·0 days (13·0–19·0) to secondary infection (figure 1; d-dimer levels greater than 1 μg/mL, and higher SOFA
table 2). Among survivors, secondary infection, acute score on admission were associated with higher odds of
kidney injury, and acute cardiac injury were observed in in-hospital death. Additionally, elevated levels of blood
one patient each, occurring 9 days (acute kidney injury), IL-6, high-sensitivity cardiac troponin I, and lactate
14 days (secondary infection), and 21 days (acute cardiac dehydrogenase and lymphopenia were more commonly
injury) after illness onset. seen in severe COVID-19 illness. Sustained viral
The median time from dyspnoea to intubation was detection in throat samples was observed in both
10·0 days (IQR 5·0–12·5) for patients who received survivors and non-survivors.
invasive mechanical ventilation and the time from inva­ Previously, older age has been reported as an impor­
sive mechanical ventilation to occurrence of ventilator- tant independent predictor of mortality in SARS and
associated pneumonia was 8·0 days (2·0–9·0; figure 1). MERS.14,15 The current study confirmed that increased
age was associated with death in patients with COVID-19.
Discussion Previous studies in macaques inoculated with SARS-CoV
This retrospective cohort study identified several risk found that older macaques had stronger host innate
factors for death in adults in Wuhan who were responses to virus infection than younger adults, with an

6 www.thelancet.com Published online March 9, 2020 https://doi.org/10.1016/S0140-6736(20)30566-3

 Articles

Survivors

Fever
Cough
Dyspnoea
ICU admission
Systematic corticosteroid
SARS-CoV-2 RNA positive
Days after illness onset Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7 Day 8 Day 9 Day 10 Day 11 Day 12 Day 13 Day 14 Day 15 Day 16 Day 17 Day 18 Day 19 Day 20 Day 21 Day 22

Sepsis ARDS Discharged

Non-survivors

Fever
Cough
Dyspnoea
ICU admission
Invasive ventilation
Systematic corticosteroid
SARS-CoV-2 RNA positive
Days after illness onset Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7 Day 8 Day 9 Day 10 Day 11 Day 12 Day 13 Day 14 Day 15 Day 16 Day 17 Day 18 Day 19 Day 20

Sepsis ARDS Death

Secondary infection
Acute kidney injury
Acute cardiac injury

Figure 1: Clinical courses of major symptoms and outcomes and duration of viral shedding from illness onset in patients hospitalised with COVID-19
Figure shows median duration of symptoms and onset of complications and outcomes. ICU=intensive care unit. SARS-CoV-2=severe acute respiratory syndrome coronavirus 2. ARDS=acute respiratory
distress syndrome. COVID-19=coronavirus disease 2019.

increase in differential expression of genes associated pneumonia include older age, pre-existing cardiovascular
with inflammation, whereas expression of type I inter­ diseases, and greater severity of pneumonia at presen­
feron beta was reduced.16 The age-dependent defects in tation.22 Coronary heart disease has also been found to be
T-cell and B-cell function and the excess production of associated with acute cardiac events and poor outcomes
type 2 cytokines could lead to a deficiency in control of in influenza and other respiratory viral infections.22–24 In
viral replication and more prolonged proinflammatory this study, increased high-sensitivity cardiac troponin I
responses, potentially leading to poor outcome.17 during hospitalisation was found in more than half of
SOFA score is a good diagnostic marker for sepsis those who died. The first autopsy of a 53-year-old woman
and septic shock, and reflects the state and degree of with chronic renal failure in Jinyintan Hospital showed
multi-organ dysfunction.18,19 Although bacterial infec­ acute myocardial infarction (data not published; personal
tions are usually regarded as a leading cause of sepsis, communication with a pathologist from the Chinese
viral infection can also cause sepsis syndrome. Previ­ Academy of Science). About 90% of inpatients with
ously, we determined that sepsis occurred in nearly pneumonia had increased coagulation activity, marked by
40% of adults with community-acquired pneumonia due increased d-dimer concentrations.25 In this study, we
to viral infection.20 In the current study, we found that found d-dimer greater than 1 µg/mL is associated with
more than half of patients developed sepsis. Additionally, fatal outcome of COVID-19. High levels of d-dimer have a
we found that more than 70% of patients had white reported asso­ ciation with 28-day mortality in patients
blood cell count below 10·0 × 10⁹ per L or procalcitonin with infection or sepsis identified in the emergency
below 0·25 ng/mL, and no bacterial pathogens were department.26 Contri­butory mechanisms include systemic
detected in these patients on admission. Sepsis was a pro-inflammatory cytokine responses that are mediators
common complication, which might be directly caused of atherosclerosis directly contributing to plaque rupture
by SARS-CoV-2 infection, but further research is needed through local inflammation, induction of procoagulant
to investigate the pathogenesis of sepsis in COVID-19 factors, and haemodynamic changes, which predispose to
illness. ischaemia and thrombosis.27–29 In addition, angiotensin
Cardiac complications, including new or worsening converting enzyme 2, the receptor for SARS-CoV-2, is
heart failure, new or worsening arrhythmia, or myocardial expressed on myocytes and vascular endothelial cells,30,31
infarction are common in patients with pneumonia. so there is at least theoretical potential possibility of direct
Cardiac arrest occurs in about 3% of inpatients with cardiac involvement by the virus. Of note, interstitial
pneumonia.21 Risk factors of cardiac events after mononuclear inflammatory infiltrates in heart tissue has

www.thelancet.com Published online March 9, 2020 https://doi.org/10.1016/S0140-6736(20)30566-3 7

 Articles

In the current study, we found that the detectable

A B
SARS-CoV-2 RNA persisted for a median of 20 days in
48 Survivors 2·0
Non-survivors 42·2 1·8 survivors and that it was sustained until death in non-

Lymphocyte count (×109 per L)

42
1·6
1·41 1·42
survivors. This has important implications for both patient
36
1·4 1·20
isolation decision making and guidance around the length
D-dimer (µg/mL)

35·6 1·43
30 1·18
1·2 1·08 0·97
24 1·0 0·91 of antiviral treatment. In severe influenza virus infection,
23·8
18 14·4 0·8 prolonged viral shedding was associated with fatal outcome
16·7 0·67
12
0·6 0·60
0·54 0·49 0·54
0·42
and delayed antiviral treatment was an independent risk
2·6 0·4 0·52
6 1·5
0·44 factor for prolonged virus detection.36 Similarly, effective
0·3 0·6 0·6 0·7
0·2
0
0·5 1·0 0·5
0·0 antiviral treatment might improve outcomes in COVID-19,
4 7 10 13 16 19 22 4 7 10 13 16 19 22 25 although we did not observe shortening of viral shedding
duration after lopinavir/ritonavir treatment in the cur­
C D
rent study. Ran­ domised clinical trials for lopinavir/
30 2500
26·4 ritonavir (ChiCTR2000029308) and intravenous remdesivir
2000 2000
24 2000 (NCT04257656, NCT04252664) in treatment of COVID-19
Serum ferritin (µg/L)

1698 are currently in progress.

IL-6 (pg/mL)

18 1500 1646 1645

17·2 Our study has some limitations. First, due to the
12·0
12 9·5 10·7
11·7 1000 1025 retrospective study design, not all laboratory tests were
7·0 635 done in all patients, including lactate dehydrogenase,
6 500
5·5
6·8 6·6 6·1 6·3
393
531 546 447 432
IL-6, and serum ferritin. Therefore, their role might be
0 0 underestimated in predicting in-hospital death. Second,
4 7 10 13 16 19 4 7 10 13 16 19 patients were sometimes transferred late in their illness to
the two included hospitals. Lack of effective antivirals,
E F
inadequate adherence to standard supportive therapy, and
350 600 590
high-dose corticosteroid use might have also contributed
High-sensitivity cardiac troponin I (pg/mL)

550 528
300 290·6
500 to the poor clinical outcomes in some patients. Third,
Lactate dehydrogenase (U/L)

250 450 the estimated duration of viral shedding is limited by

413
400 390
397 388 the frequency of respiratory specimen collection, lack
200
350 332 of quantitative viral RNA detection, and relatively low
301
150 300 323 302 positive rate of SARS-CoV-2 RNA detection in throat-
134·5
250
243 swabs.37 Fourth, by excluding patients still in hospital as of
100 240
55·7 200 236
217 198 200
Jan 31, 2020, and thus relatively more severe disease at an
22·0 57·6
50 24·7 150 earlier stage, the case fatality ratio in our study cannot
8·8 4·4
0
2·5 3·3 4·1 2·5 2·5 3·8
0
reflect the true mortality of COVID-19. Last but not least,
4 7 10 13 16 19 22 4 7 10 13 16 19 22 25 interpretation of our findings might be limited by the
Days from illness onset Days from illness onset sample size. However, by including all adult patients in
the two designated hospitals for COVID-19, we believe our
Figure 2: Temporal changes in laboratory markers from illness onset in patients hospitalised with COVID-19
Figure shows temporal changes in d-dimer (A), lymphocytes (B), IL-6 (C), serum ferritin (D), high-sensitivity study population is representative of cases diagnosed and
cardiac troponin I (E), and lactate dehydrogenase (F). Differences between survivors and non-survivors were treated in Wuhan.
significant for all timepoints shown, except for day 4 after illness onset for d-dimer, IL-6, and high-sensitivity To the best of our knowledge, this is the largest retro­
cardiac troponin I. For serum ferritin (D), the median values after day 16 exceeded the upper limit of detection,
spective cohort study among patients with COVID-19
as indicated by the dashed line. COVID-19=coronavirus disease 2019. IL-6=interleukin-6.
who have experienced a definite outcome. We found that
older age, higher SOFA score, and elevated d-dimer at
been documented in fatal cases of COVID-19, although admission were risk factors for death of adult patients
viral detection studies were not reported.32 with COVID-19. The prolonged viral shedding provides
The level and duration of infectious virus replication are the rationale for testing novel coronavirus antiviral inter­
important factors in assessing the risk of trans­mission and ventions in efforts to improve outcomes.
guiding decisions regarding isolation of patients. Because Contributors
coronavirus RNA detection is more sensitive than virus BC and HC had the idea for and designed the study and had full access to
isolation, most studies have used qualitative or quantitative all of the data in the study and take responsibility for the integrity of the
data and the accuracy of the data analysis. BC, FZ, GF, ZL, JXi, and YL
viral RNA tests as a potential marker for infectious drafted the paper. BC, FZ, RD, GF, ZL, XG, HL, and YWa did the analysis,
coronavirus. For SARS-CoV, viral RNA was detected in and all authors critically revised the manuscript for important intellectual
respiratory specimens from about a third of patients as long content and gave final approval for the version to be published. FZ, RD,
as 4 weeks after disease onset.33 Similarly, the dura­tion of GF, ZL, YL, BS, LG, YWe, XW, JXu, ST, and YZ collected the data.
All authors agree to be accountable for all aspects of the work in ensuring
MERS-CoV RNA detection in lower respiratory specimans that questions related to the accuracy or integrity of any part of the work are
persisted for at least 3 weeks,34,35 whereas the duration of appropriately investigated and resolved.
SARS-CoV-2 RNA detection has not been well characterised.

8 www.thelancet.com Published online March 9, 2020 https://doi.org/10.1016/S0140-6736(20)30566-3

 Articles

Declaration of interests 17 Opal SM, Girard TD, Ely EW. The immunopathogenesis of sepsis
We declare no competing interests. in elderly patients. Clin Infect Dis 2005; 41 (suppl 7): S504–12.
18 Singer M, Deutschman CS, Seymour CW, et al. The Third
Acknowledgments International Consensus Definitions for Sepsis and Septic Shock
This study received the following funding: Chinese Academy of Medical (Sepsis-3). JAMA 2016; 315: 801–10.
Sciences Innovation Fund for Medical Sciences (CIFMS 19 Ferreira FL, Bota DP, Bross A, Mélot C, Vincent JL. Serial evaluation
2018-I2M-1–003); National Science Grant for Distinguished Young of the SOFA score to predict outcome in critically ill patients. JAMA
Scholars (81425001/H0104); National Key Research and Development 2001; 286: 1754–58.
Program of China (2018YFC1200100, 2018YFC1200102); The Beijing 20 Zhou F, Wang Y, Liu Y, et al. Disease severity and clinical outcomes
Science and Technology Project (Z19110700660000); Major Projects of of community-acquired pneumonia caused by non-influenza
National Science and Technology on New Drug Creation and respiratory viruses in adults: a multicentre prospective registry
Development (2020ZX09201001). We acknowledge all health-care study from the CAP-China Network. Eur Respir J 2019; 54: 1802406.
workers involved in the diagnosis and treatment of patients in Wuhan; 21 Marrie TJ, Shariatzadeh MR. Community-acquired pneumonia
and we thank Chen Wang (Peking Union Medical College, Beijing, requiring admission to an intensive care unit: a descriptive study.
China), Peter W Horby (University of Oxford, Oxford, UK), and Medicine (Baltimore) 2007; 86: 103–11.
Frederick G Hayden (University of Virginia School of Medicine, 22 Corrales-Medina VF, Musher DM, Shachkina S, Chirinos JA.
Charlottesville, VA, USA) for guidance in study design and Acute pneumonia and the cardiovascular system. Lancet 2013;
interpretation of results, and review of the manuscript. 381: 496–505.
23 Udell JA, Zawi R, Bhatt DL, et al. Association between influenza
References vaccination and cardiovascular outcomes in high-risk patients:
1 Phelan AL, Katz R, Gostin LO. The novel coronavirus originating in a meta-analysis. JAMA 2013; 310: 1711–20.
Wuhan, China: challenges for global health governance. JAMA
24 Blackburn R, Zhao H, Pebody R, Hayward A, Warren-Gash C.
2020; published online Jan 30. DOI:10.1001/jama.2020.1097.
Laboratory-confirmed respiratory infections as predictors of hospital
2 Gorbalenya AE, Baker SC, Baric RS, et al. Severe acute respiratory admission for myocardial infarction and stroke: time-series analysis
syndrome-related coronavirus: the species and its viruses—a of English data for 2004–2015. Clin Infect Dis 2018; 67: 8–17.
statement of the Coronavirus Study Group. bioRxiv 2020; published
25 Milbrandt EB, Reade MC, Lee M, et al. Prevalence and significance
online Feb 11. DOI:10.1101/2020.02.07.937862 (preprint).
of coagulation abnormalities in community-acquired pneumonia.
3 Chan JWM, Ng CK, Chan YH, et al. Short term outcome and risk Mol Med 2009; 15: 438–45.
factors for adverse clinical outcomes in adults with severe acute
26 Rodelo JR, De la Rosa G, Valencia ML, et al. D-dimer is a significant
respiratory syndrome (SARS). Thorax 2003; 58: 686–89.
prognostic factor in patients with suspected infection and sepsis.
4 Li Q, Guan X, Wu P, et al. Early transmission dynamics in Wuhan, Am J Emerg Med 2012; 30: 1991–99.
China, of novel coronavirus-infected pneumonia. N Engl J Med
27 Smeeth L, Thomas SL, Hall AJ, Hubbard R, Farrington P,
2020; published online Jan 29. DOI:10.1056/NEJMoa2001316.
Vallance P. Risk of myocardial infarction and stroke after acute
5 Huang C, Wang Y, Li X, et al. Clinical features of patients infected infection or vaccination. N Engl J Med 2004; 351: 2611–18.
with 2019 novel coronavirus in Wuhan, China. Lancet 2020;
28 Corrales-Medina VF, Musher DM, Wells GA, Chirinos JA, Chen L,
395: 497–506.
Fine MJ. Cardiac complications in patients with community-
6 Wang D, Hu B, Hu C, et al. Clinical characteristics of acquired pneumonia: incidence, timing, risk factors, and association
138 hospitalized patients with 2019 novel coronavirus-infected with short-term mortality. Circulation 2012; 125: 773–81.
pneumonia in Wuhan, China. JAMA 2020; published online Feb 7.
29 Davidson JA, Warren-Gash C. Cardiovascular complications of
DOI:10.1001/jama.2020.1585.
acute respiratory infections: current research and future directions.
7 Chen N, Zhou M, Dong X, et al. Epidemiological and clinical Expert Rev Anti Infect Ther 2019; 17: 939–42.
characteristics of 99 cases of 2019 novel coronavirus pneumonia in
30 Gallagher PE, Ferrario CM, Tallant EA. Regulation of ACE2 in
Wuhan, China: a descriptive study. Lancet 2020; 395: 507–13.
cardiac myocytes and fibroblasts. Am J Physiol Heart Circ Physiol
8 Kalil AC, Metersky ML, Klompas M, et al. Management of adults with 2008; 295: H2373–79.
hospital-acquired and ventilator-associated pneumonia: 2016 clinical
31 Mendoza-Torres E, Oyarzún A, Mondaca-Ruff D, et al. ACE2 and
practice guidelines by the Infectious Diseases Society of America and
vasoactive peptides: novel players in cardiovascular/renal
the American Thoracic Society. Clin Infect Dis 2016; 63: e61–111.
remodeling and hypertension. Ther Adv Cardiovasc Dis 2015;
9 Khwaja A. KDIGO clinical practice guidelines for acute kidney 9: 217–37.
injury. Nephron Clin Pract 2012; 120: c179–84.
32 Xu Z, Shi L, Wang Y, et al. Pathological findings of COVID-19
10 Ranieri VM, Rubenfeld GD, Thompson BT, et al. Acute respiratory associated with acute respiratory distress syndrome. Lancet Respir Med
distress syndrome: the Berlin Definition. JAMA 2012; 307: 2526–33. 2020; published online Feb 18. https://doi.org/10.1016/
11 National Health Commission of the People’s Republic of China. S2213-2600(20)30076-X.
Chinese management guideline for COVID-19 (version 6.0). 33 Xu D, Zhang Z, Jin L, et al. Persistent shedding of viable SARS-CoV
Feb 19, 2020. http://www.nhc.gov.cn/yzygj/s7653p/202002/8334a832 in urine and stool of SARS patients during the convalescent phase.
6dd94d329df351d7da8aefc2/files/b218cfeb1bc54639af227f922bf6b817. Eur J Clin Microbiol Infect Dis 2005; 24: 165–71.
pdf (accessed Feb 19, 2020; in Chinese).
34 Corman VM, Albarrak AM, Omrani AS, et al. Viral shedding and
12 Yang X, Yu Y, Xu J, et al. Clinical course and outcomes of critically antibody response in 37 patients with Middle East Respiratory
ill patients with SARS-CoV-2 pneumonia in Wuhan, China: Syndrome coronavirus infection. Clin Infect Dis 2016; 62: 477–83.
a single-centered, retrospective, observational study.
35 Oh MD, Park WB, Choe PG, et al. Viral load kinetics of MERS
Lancet Respir Med 2020; published online Feb 24. https://doi.org/
coronavirus infection. N Engl J Med 2016; 375: 1303–05.
10.1016/S2213-2600(20)30079-5.
36 Wang Y, Guo Q, Yan Z, et al. Factors associated with prolonged viral
13 Arabi YM, Balkhy HH, Hayden FG, et al. Middle East respiratory
shedding in patients with avian influenza A(H7N9) virus infection.
syndrome. N Engl J Med 2017; 376: 584–94.
J Infect Dis 2018; 217: 1708–17.
14 Choi KW, Chau TN, Tsang O, et al. Outcomes and prognostic
37 Zou L, Ruan F, Huang M, et al. SARS-CoV-2 viral load in upper
factors in 267 patients with severe acute respiratory syndrome in
respiratory specimens of infected patients. N Engl J Med 2020;
Hong Kong. Ann Intern Med 2003; 139: 715–23.
published online Feb 19. DOI:10.1056/NEJMc2001737.
15 Hong K-H, Choi J-P, Hong S-H, et al. Predictors of mortality in
Middle East respiratory syndrome (MERS). Thorax 2018;
73: 286–89.
16 Smits SL, de Lang A, van den Brand JMA, et al. Exacerbated innate
host response to SARS-CoV in aged non-human primates.
PLoS Pathog 2010; 6: e1000756-e.

www.thelancet.com Published online March 9, 2020 https://doi.org/10.1016/S0140-6736(20)30566-3 9