Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study

Zhou F, Yu T, Du R, et al. Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study. The Lancet. 2020;395(10229):1054-1062. doi:10.1016/s0140-6736(20)30566-3

Review written by:

Harsha Senapathi 

Lillian Zerihun

Green - Retrospective, multicenter cohort, and published in The Lancet

This was a retrospective study with 2 cohorts at Jinyintan and Wuhan Hospitals. Selected patients died or were discharged between Dec 29, 2019, and Jan 31, 2020. SARS-CoV-2 was detected in respiratory specimens by next-generation sequencing or real-time PCR. Upon clinical remission, PCR was subsequently conducted every other day. Discharge criteria was defined as the absence of fever for at least 3 days, substantial lung improvement on CT, and two negative PCR tests taken at least 24 hours apart. All factors studied were collected at admission.

191 patients were included in the study; 54 patients died, and 137 patients were discharged. Mean age was 56 years old (range of 18 – 87), and 62% were male. The median time from illness onset (pre-admission) to discharge among survivors was 22 days.

The most common symptoms on admission were:

  • fever [94%]

  • cough [79%]

  • sputum production [23%]

  • fatigue [23%]

Univariate analysis showed the following variables were associated with death (p-value < 0.05):

  • increased age per year

  • coronary heart disease

  • hypertension  

  • respiratory rate >24 

  • SOFA 

  • q-SOFA

  • lymphopenia 

  • leukocytosis 

  • ALT > 40 U/L

  • creatinine > 133 μmol/L

  • LDH > 245 U/L

  • creatine kinase > 185 U/L

  • troponin 1 >28 pg/mL

  • d-dimer > 1 μg/mL

  • prothrombin time > 16 s

  • serum ferritin > 300 μg/L 

  • IL-6 per unit pg/mL 

  • procalcitonin per unit ng/mL

171/191 patients and 5 factors (coronary heart disease, age, SOFA, d-dimer, and lymphopenia) were included in the multivariable logistic regression and found these factors to be associated with death:

  • older age (OR 1.10 per year (1.03 – 1.17), p = 0.0043),

  • higher SOFA score (OR 5.65 (2.61 – 12.23), p <0.0001)

  • d-dimer greater than 1 μg/mL (OR 18.42 (2.64 – 128.55), p = 0.0033)

The median duration of viral shedding among survivors was 20.0 days (range: 8 to 37 days) and until death for non-survivors. Among non-survivors, the median time from illness onset to sepsis was 10.0 days, 12.0 days to ARDS, 14.5 days to both intubation and acute cardiac injury, 17.0 days to secondary infection, and 18.5 days to death. Non-survivors included 31 of 32 subjects who were intubated.

Review Notes


This is currently the largest retrospective cohort study among COVID-19 patients with a definitive outcome. The factors found to be associated with death may be helpful in predicting outcomes and guiding management.


  1. Due to the retrospective nature of study, not all laboratory tests were done in every patient, including lactate dehydrogenase, IL-6, and serum ferritin. Not all factors were included in the multivariable analysis and their role might be underestimated in predicting in-hospital death.

  2. In addition, the two cohorts had significant differences in several underlying comorbidities. Late admissions and lack of antivirals could have led to poorer outcomes in some patients.

  3. The estimated duration of viral shedding is limited by the frequency of respiratory specimen collection, lack of quantitative viral RNA detection, and relatively low positive rate of SARS-CoV-2 RNA detection in throat-swabs.

Because patients without an outcome (death or discharge) by January 31st weren’t included in the study, the case-fatality rate doesn’t reflect true mortality of COVID-19.

  • Twitter
  • Instagram