I now have comprehensive, high-quality data from authoritative sources. Let me compile the full management protocol.
ICU Sepsis Management Protocol
Based on Surviving Sepsis Campaign 2021 Guidelines, Harrison's 22nd Edition (2025), Goldman-Cecil Medicine, Barash Clinical Anesthesia 9e, and current ESICM guidelines (2024–2025)
1. Definitions (Sepsis-3)
| Entity | Criteria |
|---|
| Sepsis | Life-threatening organ dysfunction caused by dysregulated host response to infection: SOFA score increase ≥2 from baseline |
| Septic Shock | Sepsis + vasopressor requirement to maintain MAP ≥65 mmHg AND serum lactate >2 mmol/L despite adequate fluid resuscitation |
| qSOFA (screening) | RR ≥22/min + GCS <15 + SBP ≤100 mmHg — ≥2 criteria warrants high suspicion |
2. Immediate Recognition & Triage
- Implement institutional sepsis screening protocols (e.g., EHR-based alerting, NEWS, MEWS)
- qSOFA ≥2 → escalate to full SOFA assessment and clinical evaluation
- Target ICU admission within 6 hours of diagnosis for all critically ill or shock patients
- Obtain baseline labs: CBC, BMP, LFTs, coagulation panel, procalcitonin, lactate, ABG/VBG, urinalysis
3. The First Hour — Priority Actions
3a. Blood Cultures & Diagnostics
- Draw ≥2 sets of blood cultures (aerobic + anaerobic) before antibiotics — do not delay antibiotics >45 minutes waiting for cultures
- Additional cultures from suspected source: urine, sputum, wound, CSF (if indicated)
- Prompt imaging to identify and localize source (CXR, CT abdomen/pelvis as needed)
3b. Antibiotics
- Septic shock → antibiotics within 1 hour of recognition (every 1-hour delay = ~7–8% increase in mortality)
- Sepsis without shock → empiric antibiotics within 3 hours if no alternative diagnosis identified
Empiric Antibiotic Selection by Scenario:
| Scenario | Preferred Agents |
|---|
| Undifferentiated sepsis, low Pseudomonas risk | Ceftriaxone or cefotaxime |
| Pseudomonas likely (structural lung disease, neutropenia, recent abdominal surgery) | Cefepime, piperacillin-tazobactam, or carbapenem (meropenem/imipenem) |
| Highly resistant GNR risk (prior MDR colonization/infection) | Two empiric gram-negative agents |
| MRSA risk (frequent healthcare exposure, hospital-onset) | Add vancomycin or linezolid |
| Candida risk (recent abdominal surgery, TPN, liver failure, multifocal Candida colonization) | Echinocandin (micafungin, caspofungin) |
| SARS-CoV-2 | Add remdesivir |
| Influenza | Add oseltamivir |
Optimization: Administer β-lactams before vancomycin; consider prolonged β-lactam infusions (extended/continuous infusion) for PK/PD optimization — consult Pharmacy/ID.
3c. Source Control
- Identify and control all surgically or procedurally amenable sources as rapidly as possible
- Examples: drain abscesses, resect/repair perforated viscus, debride necrotizing fasciitis, relieve biliary obstruction (cholangitis), remove infected catheters
- Remove all potentially infected indwelling catheters (central lines, Foley, drains) when infected source suspected
4. Hemodynamic Resuscitation
4a. Fluid Resuscitation
- Initial bolus: 30 mL/kg IV crystalloid within first 3 hours (target MAP ≥65 mmHg)
- Preferred fluids: balanced crystalloids (Lactated Ringer's, Plasma-Lyte) — avoid 0.9% NaCl in large volumes (hyperchloremic acidosis risk)
- Do not use hydroxyethyl starch (HES/hetastarch) — associated with acute kidney injury and increased mortality
- Consider adding albumin when large crystalloid volumes are required
- After initial bolus: re-assess with dynamic fluid responsiveness testing (passive leg raise + cardiac output measurement, pulse pressure variation, stroke volume variation) rather than static CVP/PCWP
- Serial lactate measurements to guide adequacy of resuscitation — target lactate clearance ≥10–20% per 2 hours; lactate >2 mmol/L despite resuscitation = persistent tissue hypoperfusion
- Avoid fluid overload — associate de-resuscitation (judicious diuresis) in the post-resuscitation phase
4b. Vasopressors
Initiate when MAP <65 mmHg persists despite initial fluid resuscitation; can be given via peripheral IV initially.
| Step | Agent | Dose | Notes |
|---|
| 1st line | Norepinephrine | Start 0.01–0.1 μg/kg/min, titrate | Preferred first-line — potent α + β adrenergic effects; reduces mortality vs. dopamine |
| 2nd line | Vasopressin | Add at 0.03 units/min (fixed dose) | Add when NE ≥0.25–0.5 μg/kg/min; do not titrate; spare catecholamines |
| 3rd line | Epinephrine | 0.01–0.3 μg/kg/min | Add when hypotension persists despite NE + vasopressin |
| Low CO state | Dobutamine | 2.5–20 μg/kg/min | Add to NE, or use epinephrine alone, if persistent low cardiac output despite adequate preload |
| Avoid | Dopamine | — | Use only in highly selected cases (bradycardia without tachyarrhythmia risk); higher arrhythmia and mortality |
| Avoid | Terlipressin, levosimendan | — | Not recommended per SSC 2021 |
- Target MAP ≥65 mmHg (higher targets, e.g., 80–85 mmHg, not shown to improve outcomes and increase arrhythmias)
- Monitor for signs of end-organ perfusion: urine output (target ≥0.5 mL/kg/hr), mental status, lactate clearance
5. Respiratory Support
- Supplemental O₂ to target SpO₂ ≥94%
- Low threshold for early endotracheal intubation in sepsis-induced respiratory failure
Mechanical Ventilation (Sepsis-induced ARDS):
| Parameter | Target |
|---|
| Tidal volume | 6 mL/kg ideal body weight (lung-protective) |
| Plateau pressure | ≤30 cmH₂O |
| PEEP | Titrate to oxygenation (higher PEEP for moderate-severe ARDS) |
| FiO₂ | Minimize — target SpO₂ 92–96% |
| Head-of-bed elevation | 30–45° (semi-recumbent) — reduces VAP |
| Prone positioning | ≥12 hrs/day for moderate-severe ARDS (P/F ratio <150) |
| NMB | Use intermittent bolus dosing to facilitate proning in moderate-severe ARDS |
| Pulmonary artery catheter | Avoid routine use in ALI/ARDS |
| ECMO (V-V) | Consider for severe ARDS refractory to ventilation — at experienced centers |
6. Adjunctive & Supportive Therapies
6a. Corticosteroids
- Indication: Septic shock refractory to adequate fluid resuscitation and vasopressor therapy
- Regimen: Hydrocortisone 200 mg/day IV (either 50 mg q6h or continuous infusion) ± fludrocortisone 50 mcg PO daily
- Do not use in mild septic shock or sepsis without shock
- Do not use corticotropin stimulation testing to guide steroid use — does not predict responders
- Taper once vasopressors are no longer required
6b. Blood Transfusion
- Transfusion threshold: Hgb <7 g/dL (restrictive strategy) in absence of tissue hypoperfusion, acute coronary syndrome, or active hemorrhage
- Target Hgb 7–9 g/dL
- Do not target supranormal oxygen delivery values
6c. Glycemic Control
- Target blood glucose 140–180 mg/dL (7.8–10 mmol/L) using insulin infusion protocol
- Avoid hypoglycemia (associated with increased mortality)
- Monitor glucose every 1–2 hours until stable, then every 4 hours
6d. Renal Replacement Therapy (RRT)
- Initiate when AKI results in life-threatening fluid overload, severe acidosis (pH <7.2 in setting of AKI), refractory hyperkalemia, or uremia
- Sodium bicarbonate: consider for pH <7.2 in setting of acute kidney injury
- No proven benefit to early vs. late initiation in stable (non-emergent) AKI — individualize
6e. Anticoagulation & DVT Prophylaxis
- Pharmacologic DVT prophylaxis (UFH or LMWH) for all patients without contraindication
- Add mechanical prophylaxis (sequential compression devices) when pharmacologic contraindicated
- Avoid routine therapeutic anticoagulation for sepsis-induced coagulopathy or DIC without confirmed thrombosis
6f. Stress Ulcer Prophylaxis
- Proton pump inhibitor or H₂ antagonist for patients with risk factors (mechanical ventilation >48h, coagulopathy, high-dose steroids, prior GI bleed)
6g. Nutrition
- Start enteral nutrition within 24–48 hours of ICU admission when hemodynamically stable and GI tract functional
- Avoid parenteral nutrition in the first 7 days if early EN is feasible
- Target 25–30 kcal/kg/day; avoid overfeeding
7. Antibiotic De-escalation & Stewardship
- Reassess antibiotic spectrum at 48–72 hours once culture/sensitivity data available
- Narrow spectrum as soon as pathogen and sensitivities are known
- Procalcitonin-guided de-escalation/discontinuation is supported — do not use PCT to start antibiotics, but use serial PCT to guide stopping
- Typical course: 7–10 days for most sepsis sources; shorter courses may be appropriate in many infections (e.g., uncomplicated bacteremia 7 days; pneumonia 5–7 days)
- Longer courses for immunosuppressed patients, slow clinical response, or undrained infection foci
8. Monitoring & Targets Summary
| Parameter | Target |
|---|
| MAP | ≥65 mmHg |
| Urine output | ≥0.5 mL/kg/hr |
| Lactate | Normalize; clearance ≥10%/2h |
| SpO₂ | ≥94% (92–96% on ventilator) |
| Blood glucose | 140–180 mg/dL |
| Hemoglobin | 7–9 g/dL |
| Platelet transfusion threshold | <10,000/μL (prophylactic); <50,000/μL if bleeding/procedure |
9. De-escalation of Care & ICU Liberation
Once hemodynamically stable:
- ABCDEF Bundle: Assess analgesia, target light sedation (RASS –1 to 0), daily SAT/SBT, early mobilization
- Daily sedation interruption when on continuous infusion sedation
- Ventilator weaning via daily spontaneous breathing trials (SBT)
- Remove indwelling lines/catheters (CVC, Foley) as soon as no longer needed
- Judicious diuresis for fluid overload in post-resuscitation phase
- DVT prophylaxis, early physical/occupational therapy
- Screen for ICU-acquired weakness, delirium, post-ICU PTSD
10. Post-ICU Discharge Planning
- Screen for ongoing physical, cognitive, and psychological sequelae (post-intensive care syndrome — PICS)
- Address economic and social support needs
- Establish follow-up with providers capable of assessing long-term complications
- Counsel on elevated risk of rehospitalization (50% within 1 year), cardiovascular events, and recurrent infection
Evidence Notes (2024–2025 Updates)
- ESICM Fluid Guidelines 2024–2025 (PMID 38771364, 40163133): Balanced crystalloids preferred; restrict cumulative volumes; reassess with dynamic parameters rather than static filling pressures.
- ESICM Hemodynamic Monitoring Guidelines 2025 (PMID 41236566): Updated guidance on circulatory shock assessment and vasopressor monitoring.
- S3 Sepsis Guideline Update 2025 (PMID 40824313): European update reinforcing SSC 2021 core recommendations with refinements in fluid management and antibiotic stewardship.
- No FDA-approved host-targeted therapies (anti-TNF, activated protein C, IV immunoglobulin, vitamin C, polymyxin-B hemoperfusion) have proven mortality benefit — none recommended.
Sources: Harrison's Principles of Internal Medicine 22E (2025); Barash, Cullen & Stoelting's Clinical Anesthesia 9e; Goldman-Cecil Medicine; Rosen's Emergency Medicine; Current Surgical Therapy 14e — cross-referenced with Surviving Sepsis Campaign 2021 Guidelines and ESICM 2024–2025 practice guidelines.