COVID-19: US critical illness rates are higher among racial, ethnic minorities

  • Cummings MG, & et al.
  • Lancet
  • 19 May 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

  • Critical COVID-19 illness was more common among racial and ethnic minorities in this study of patients in New York City.
  • These authors identified IL-6 and D-dimers as independent predictors of increased mortality risk.

Why this matters

  • Related editorial notes "relatively high" proportion of patients age accompanying comorbidities that might be factors.

Key results

  • 257 patients were critically ill.
    • Median age, 62 years (interquartile range [IQR], 51-72).
    • 67% were men.
    • 62% were Hispanic/Latino.
    • 82% had >1 comorbidity (hypertension, diabetes most common). 
  • 39% died.
  • 79% (203/257) received invasive mechanical ventilation (median, 18 days; IQR, 9-28 days). 
    • 41% (84/203) of these patients died.
  • 37% (94/257) remained hospitalized at publication (median duration, 33 days; IQR, 29-36 days).
  • Medications: 66% vasopressors, 72% hydroxychloroquine, 9% remdesivir, 89% antibacterials, 26% corticosteroids, 17% IL-6 receptor antagonist.
  • On multivariate analysis, risk factors (adjusted HRs; 95% CIs) independently associated with in-hospital mortality included:
    • Older age: 1.31 (1.09-1.57) for each 10-year increase. 
    • Chronic cardiovascular disease: 1.76 (1.08-2.86).
    • Chronic pulmonary disease: 2.94 (1.48-5.84).
    • Elevated IL-6: 1.11 (1.02-1.20).
    • Elevated D-dimer: 1.10 (1.01-1.19).

Study design

  • Prospective, observational, cohort study, including critically ill adults hospitalized with laboratory-confirmed COVID-19, March 2-April 1, 2020 (follow-up ended April 28).
  • Funding: National Institute of Allergy and Infectious Diseases; others.

Limitations

  • Limited generalizability. 
  • Unexplored confounders.
  • Underestimated in-hospital mortality rate.