A decade of non-inferiority trials has shown that many standard antibiotic courses were longer than they needed to be. Matching duration to the evidence reduces toxicity, cost and resistance pressure without raising failure rates.
Several durations are now well supported: around five days for community-acquired pneumonia in a patient improving by then, five to six days for uncomplicated cellulitis, seven days for uncomplicated Gram-negative bacteraemia and for pyelonephritis with an active agent, and short courses for many intra-abdominal infections once source control is achieved. The common thread is a patient who is responding and has had the source controlled.
The habit worth keeping is to set a stop date or review date when starting antibiotics, and to ask at each review whether the evidence supports stopping rather than continuing by default.
- Around five days suffices for community-acquired pneumonia in a responding patient.
- Five to six days is enough for uncomplicated cellulitis.
- Seven days treats uncomplicated Gram-negative bacteraemia and pyelonephritis with an active drug.
- Short courses work for many intra-abdominal infections once source control is achieved.
- Set a stop or review date at the start and justify any continuation.
Why it matters
Longer courses add toxicity and resistance pressure without improving cure in most responding patients.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for infectious diseases, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free