Standardised procalcitonin monitoring and antibiotic duration in community-acquired sepsis: a pragmatic randomised controlled trial

Prolonged antibiotic therapy drives antimicrobial resistance, yet most evidence for procalcitonin (PCT)-guided stewardship in sepsis comes from protocol-driven trials in intensive care. We conducted a pragmatic, open-label, randomised controlled trial at a tertiary emergency department in Bangkok, Thailand (February 2025–March 2026). Adults with community-acquired sepsis (infection plus two or more systemic inflammatory response syndrome criteria) were randomised 1:1 to scheduled PCT measurement on days 0, 2 and 5, with results returned through routine channels, or to standard care; treating physicians retained full discretion. The primary outcome was total antibiotic duration. Of 182 randomised participants, 130 were analysed after post-randomisation withdrawals (PCT n = 62, control n = 68). Antibiotic duration was shorter with PCT monitoring (median 7 days, interquartile range 6–10, versus 9 days, 7–12.5; Mann–Whitney exact P = 0.028; mean difference − 1.34 days, 95% confidence interval − 2.60 to − 0.08). Twenty-eight-day mortality was 4/62 (6.5%) versus 3/68 (4.4%) (risk ratio 1.46, 95% confidence interval 0.34 to 6.28) and, among all randomised participants with known status, 12/87 (13.8%) versus 16/92 (17.4%) (risk ratio 0.79, 0.40 to 1.58); the trial was not powered for mortality and cannot exclude harm. Guidance-only PCT monitoring shortened antibiotic duration in emergency department sepsis; its safety requires confirmation in a larger trial. Trial registration: Thai Clinical Trial Registry no. TCTR20250216013 (https://www.thaiclinicaltrials.org)