- Design
- International, randomised, open-label trial
- Population
- 508 stable adults with left-sided endocarditis (S aureus, E faecalis, streptococci)
- Primary outcome
- Days alive without antibiotics at 6 months; death, surgery or embolism
- Effect
- +13 days (95% CI 12-13); safety 8.2% vs 10.7%; relapse 5.1% vs 1.6%
POET II, an international open-label trial, enrolled 508 adults in stable condition with left-sided infective endocarditis due to Staphylococcus aureus, Enterococcus faecalis or streptococci. All had received at least 2 to 4 weeks of antibiotics and met stabilisation criteria. They were then randomised to stop antibiotics or continue to a standard total of 4 to 6 weeks. Published in the New England Journal of Medicine on 28 August 2026.
Tailored therapy gave a median of 13 more days alive without antibiotics over six months (183 vs 169). The safety composite of death, unplanned cardiac surgery or symptomatic embolism occurred in 8.2% versus 10.7%, meeting non-inferiority. But relapse of bacteraemia or endocarditis was 5.1% with tailored therapy and 1.6% with standard duration.
The six-week standard rests largely on expert opinion, so testing it was overdue. The result is mixed: shortening looks safe on major events but carries a real cost in relapse.
For now, standard durations remain the default, with shortening a case-by-case decision for clearly stable patients with reliable follow-up. Guidelines have not changed on the basis of this trial.
- Keep standard 4 to 6 week courses as the default for left-sided endocarditis
- Consider earlier stopping only in clearly stable patients, after discussion with the endocarditis team
- Arrange early follow-up blood cultures if antibiotics are stopped early
- Warn patients to report fever promptly after any shortened course
- Note that the trial excluded unstable patients and other organisms
Why it matters
The six-week standard was tested head to head, and the answer is a trade-off rather than a free win.
Don't overread it
Non-inferiority on the safety composite does not mean shorter courses are as good; relapse was higher.
The statistics, in plain English
The safety difference (-2.4 points, 95% CI -7.7 to 2.7) stayed inside the pre-set 7.5-point margin, so tailored therapy was judged non-inferior. Relapse was a secondary outcome with small numbers (13 vs 4), but the gap reached significance and is clinically important.
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