Practice pointer for the COVID-19 pandemic: caution urged in antibiotic prescribing for respiratory infections

  • Leis JA & al.
  • BMJ
  • 13 nov. 2020

  • de Liz Scherer
  • Clinical Essentials
El acceso al contenido completo es sólo para profesionales sanitarios registrados. El acceso al contenido completo es sólo para profesionales sanitarios registrados.


  • The clinician-led "Choosing Wisely" campaign, present in 20 countries, discourages antibiotic prescribing for upper respiratory tract infections (RTIs) in patients presenting with signs/symptoms indicative of mild, self-limiting COVID-19.

Why this matters

  • Virtual assessments during the COVID-19 pandemic have limited diagnostic capabilities and lead to antibiotic overprescribing for RTIs.
  • Judicious prescribing is recommended and in-person assessment for suspected bacterial RTIs should follow virtual consultations.
  • Provide diagnostic clarity, address patient concerns, offer a symptom management plan, and ensure that a safety net or contingency plan is in place.

Key points

  • Most RTIs are self-limiting.
  • Avoid antibiotics for suspected viral RTIs (COVID-19, influenza, influenza-like-illness).
  • Schedule in-person follow-up should the following occur:
    • Acute otitis media: symptoms >48 hours, fever ≥39°C.
    • Pharyngitis: worsening/persistent symptoms >48 hours, high suspicion of bacterial pharyngitis.
    • Sinusitis: no improvement >7 days, presence of red flags (e.g., headache, altered mental status, ophthalmic symptoms such as orbital swelling or double vision, neurologic deficits).
    • COPD: with shortness of breath that prevents activities of daily living.
    • Suspected pneumonia: in-person visit required.
    • Common cold: dehydration, red flags.

Study design

  • Practice pointer based on Medline, Cochrane randomized controlled trials, meta-analyses to summarize the best evidence supporting a standardized approach to antibiotic prescribing for RTIs in the era of COVID-19.
  • Funding: None disclosed.


  • Selection bias.
  • Inconsistent methods across studies.