Chinese study confirms increased age, other risk factors in COVID-19 deaths

  • Zhou F, et al.
  • The Lancet
  • 9 Mar 2020

  • curated by Liz Scherer
  • Clinical Essentials
Access to the full content of this site is available only to registered healthcare professionals. Access to the full content of this site is available only to registered healthcare professionals.

Takeaway

  • Chinese researchers have definitively linked older age, sepsis, elevated d-dimer to mortality in COVID-19 illness. 
  • Prolonged viral shedding confirms need to isolate infected patients. 

Why this matters

  • Clinical suspicion is warranted in older patients with higher Sequential Organ Failure Assessment (SOFA) score, elevated d-dimer (>1 µg/L) at admission.
  • Consider elevated blood IL-6, high-sensitivity cardiac troponin I, lactate dehydrogenase, lymphopenia as prognostic for severity. 

Key results

  • 191 hospitalized patients, median age 56.0 (interquartile range [IQR], 46.0-67.0) years, 62% male. 
  • 54 died in-hospital, 137 discharged.
    • Median time for illness onset to discharge: 22.0 (IQR, 18-25) days.
    • Median time to death: 18.5 (IQR, 15-22) days.
  • 48% (91) had comorbidities, most commonly hypertension (30%), diabetes (19%), and coronary heart disease (CHD; 8%).
  • Multivariate, higher mortality odds (OR) linked to:
    • Older age: 1.10 (P=.0043);
    • CHD: 2.14 (P=.48);
    • SOFA score: 5.65 (P<.0001>
    • Lymphocyte count, 10.0×109/L: 0.19 (P=.13); and 
    • d-dimer >1 µg/L: 18.42 (P=.0033).
  • Median duration of viral shedding among survivors was 20.0 (IQR, 17.0-24.0) days; shedding was detectable until death among nonsurvivors.  
  • Longest viral shedding duration was 37 days.

Study design

  • Retrospective, multicenter cohort analysis exploring mortality, clinical course, viral shedding among 2 adult cohorts.
  • Funding: Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences.

Limitations

  • Retrospective.
  • Late hospital transfer in some.
  • Viral shedding duration estimated.
  • Case mortality ratio not reflective of true mortality rates.