Review antibiotics daily, narrow once cultures return, use procalcitonin with clinical judgement to stop, and prefer short courses (often 5-7 days).
At the bedside
Timed steps- 1
Antibiotic time-out
DailyEvery round: Is this infection? Which source? Are cultures back? Can I narrow, switch to oral, or stop? Stop empirical antibiotics if another cause of illness is shown or strongly suspected (SSC 2026).
- 2
Narrow to the culture result
48-72 hSwitch a carbapenem to ceftriaxone or another narrower agent if susceptible (SSC 2026 recommends de-escalation once the organism and susceptibilities are known). Stop MRSA cover if cultures (and nasal MRSA screen, if done) are negative.
- 3
Stop double cover
48-72 hDrop the second gram-negative agent (e.g. amikacin) once susceptibility is known.
- 4
Consider oral step-down
Day 3-5Stable, absorbing, source controlled, and an oral option with good bioavailability exists.
- 5
Use procalcitonin to support stopping
Decision pointsWhere the right duration is unclear, use procalcitonin with clinical review to decide when to stop (SSC 2026, conditional): stop when PCT falls ≥ 80% from its peak or to ≤ 0.5 ng/mL, with clinical improvement (SAPS).
- 6
Short durations
EndMost sepsis 5-7 days (ICMR: 7-10). Uncomplicated gram-negative bacteraemia 7 days (BALANCE). Intra-abdominal after source control about 4 days. Longer for S. aureus bacteraemia (≥ 14 days), endocarditis, osteomyelitis or undrained abscess.
Typical durations (source controlled, clinically improving)
| Infection | Duration | Basis |
|---|---|---|
| Community-acquired pneumonia | 5-7 days | Guideline consensus |
| Hospital or ventilator-associated pneumonia | 7 days | IDSA/ATS 2016 |
| Complicated UTI or pyelonephritis | 7 days | RCTs |
| Intra-abdominal after source control | ~4 days | STOP-IT |
| Gram-negative bacteraemia, uncomplicated | 7 days | BALANCE 2025 |
| S. aureus bacteraemia | ≥ 14 days IV (4-6 weeks if complicated) | Guidelines |
References
- 1
- 2
- 3
- 4
- 5
Daily antibiotic review
Step 1 of ~3
Decision: Is infection still the likely diagnosis?
Based on Prescott HC, Antonelli M, Alhazzani W, et al; de Jong E, van Oers JA, Beishuizen A, et al. Educational use; follow local protocol.
Evidence
- SSC 2026: re-evaluate continuously and stop empirical therapy if another diagnosis is shown or strongly suspected (good practice); recommends de-escalation when organism and susceptibilities are known (strong, very low certainty); suggests de-escalation even when final cultures are negative (conditional); shorter over longer courses after adequate source control (conditional); PCT plus clinical evaluation to decide when to stop if the duration is unclear (conditional).
- Selective decontamination of the digestive tract (new, conditional, moderate certainty): suggested for ventilated septic patients only in units with a low prevalence of antimicrobial resistance.
- SAPS (2016): PCT-guided stopping reduced antibiotic duration and was associated with lower mortality in Dutch ICUs.
- BALANCE (2025, n ≈ 3600): 7 days was non-inferior to 14 days for bloodstream infection in hospitalised patients, including ICU patients (S. aureus and some others excluded).
- STOP-IT (2015): a fixed ~4 days after source control for intra-abdominal infection.
Controversies
- De-escalation with no organism grown: SSC 2026 now suggests it (conditional, very low certainty); many clinicians already narrow if the patient is improving.
- PCT is less reliable in kidney failure, after surgery or trauma, and in fungal or atypical infections.
🇮🇳 India notes
- Indian stewardship programmes have reduced carbapenem and polymyxin use without harm by using daily antibiotic time-outs.
- PCT can be expensive in many labs; use it at decision points (day 3-5) rather than daily.
- ICMR 2019 suggests 7-10 days for most sepsis, with longer courses for slow response, undrainable foci or S. aureus bacteraemia.
- SSC 2026's suggestion for selective digestive decontamination applies only to low-resistance units; most Indian ICUs report high carbapenem resistance, so it will rarely apply. Consultant to decide whether to mention it in the app.
References
- 1
- 2
- 3
- 4
- 5
Test yourself
Module quiz · score 70% to complete the module
Educational reference for trained clinicians. Check doses against your hospital protocol and the patient in front of you.