COVID-19: no mortality benefit seen with hydroxychloroquine alone or with azithromycin

  • Rosenberg ES & al.
  • JAMA
  • 11 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

  • Hydroxychloroquine, azithromycin, or the combination was not associated with significantly lower in-hospital mortality in this US cohort study of hospitalized patients with COVID-19.               

Why this matters

Key results

  • 1438 patients included; most were male, age >45 years.
  • Treatments:
    • Hydroxychloroquine+azithromycin: 51.1% (735). 
    • Hydroxychloroquine alone: 18.8% (271).
    • Azithromycin alone: 14.7% (211).
    • Neither: 15.4% (221).
  • At 21 days, estimated mortality averaged across all observed patient covariate patterns (95% CIs):
    • Hydroxychloroquine+azithromycin: 22.5% (19.7%-25.1%).
    • Hydroxychloroquine alone: 18.9% (14.3%-23.2%).
    • Azithromycin alone: 10.9% (5.8%-15.6%).
    • Neither: 17.8% (11.1%-23.9%).
  • Adjusted HRs (95% CIs) for mortality were not significant vs use of neither drug:
    • Combination: 1.35 (0.76-2.40).
    • Hydroxychloroquine alone: 1.08 (0.63-1.85).
    • Azithromycin alone: 0.56 (0.26-1.21).
  • Mortality did not differ significantly between hydroxychloroquine only vs azithromycin only (adjusted HR, 1.92; 95% CI, 0.99-3.74).
  • Compared with patients receiving neither drug, cardiac arrest was more likely with the combination (aOR, 2.13; 95% CI, 1.12-4.05), but not with hydroxychloroquine alone (1.91; 0.96-3.81) or azithromycin alone (0.64; 0.27-1.56).

Study design

  • Retrospective, multicenter cohort study of hydroxychloroquine and azithromycin prescribing patterns in patients hospitalized with COVID-19 in New York State.
  • Funding: None disclosed.

Limitations

  • Retrospective, observational.
  • Uncaptured readmissions.
  • Residual confounding likely.