I now have all the content needed. Let me compile the complete ICU sepsis management protocol.
ICU Sepsis Management Protocol
Based on the Surviving Sepsis Campaign (SSC) 2021 Guidelines, Harrison's Principles of Internal Medicine 22e (2025), Barash Clinical Anesthesia 9e, Fischer's Mastery of Surgery 8e, and recent ESICM 2025 guidelines.
1. Definitions (Sepsis-3)
| Condition | Criteria |
|---|
| Sepsis | Suspected/confirmed infection + acute increase in SOFA score ≥2 from baseline |
| Septic Shock | Sepsis + vasopressor requirement to maintain MAP ≥65 mmHg + serum lactate >2 mmol/L despite adequate fluid resuscitation |
Rapid screening - qSOFA (use at bedside, not in ICU):
- Respiratory rate ≥22 breaths/min
- GCS <15
- Systolic BP ≤100 mmHg
Score ≥2: high risk - escalate monitoring and initiate workup.
2. Immediate Recognition & Workup (Hour 0)
Order simultaneously:
- Lactate (serum) - recheck in 2-4 h if initial >2 mmol/L
- Blood cultures x2 sets (aerobic + anaerobic) - BEFORE antibiotics
- CBC, CMP, coagulation panel (PT/INR, fibrinogen), LFTs
- Procalcitonin (baseline; useful for antibiotic de-escalation, NOT for starting therapy)
- ABG/VBG, chest X-ray
- Urine culture, urinalysis
- Consider: LP (if CNS source), wound culture, imaging (CT abdomen/pelvis, ultrasound) to identify source
ICU admission: target within 6 hours of diagnosis for all critically ill or shocked patients.
3. The 1-Hour Bundle (SSC 2021)
| Action | Target Time | Details |
|---|
| Measure lactate | Hour 0 | Recheck if >2 mmol/L |
| Blood cultures | Before antibiotics | 2 sets aerobic + anaerobic |
| Broad-spectrum antibiotics | Within 1 hour (shock) / Within 3 h (no shock, uncertain diagnosis) | See Section 4 |
| IV crystalloid | Begin immediately if hypotensive or lactate ≥4 mmol/L | 30 mL/kg in first 3 h |
| Vasopressors | If MAP <65 mmHg during/after fluids | Norepinephrine first-line |
4. Antimicrobial Therapy
Timing
- Septic shock: immediate empiric antibiotics within 1 hour of recognition - each 1-hour delay increases mortality by ~7-8%
- Sepsis without shock: if alternative diagnosis not identified within 3 h, start empiric antibiotics
Empiric Regimens by Source
| Site | Regimen | Notes |
|---|
| Undifferentiated / Unknown source | Ceftriaxone + metronidazole ± vancomycin | Add antipseudomonal coverage if risk factors present |
| Pulmonary (CAP) | Beta-lactam (ampicillin-sulbactam, ceftriaxone) + azithromycin OR respiratory fluoroquinolone (levofloxacin) | Add MRSA/Pseudomonas coverage if risk factors |
| HAP/VAP | Vancomycin or linezolid + antipseudomonal beta-lactam (piperacillin-tazobactam, cefepime, meropenem) | Two antipseudomonal agents from different classes if prior IV abx within 90 days |
| Intra-abdominal | Piperacillin-tazobactam OR meropenem; add vancomycin if MRSA risk | Source control mandatory |
| Urinary tract | Ceftriaxone (outpatient); cefepime or meropenem (healthcare-associated) | |
| CNS (community meningitis) | Ceftriaxone + vancomycin + dexamethasone ± ampicillin (if Listeria risk) | |
| Healthcare-associated ventriculitis | Vancomycin + cefepime or meropenem | |
| Skin/soft tissue - NSTI | Vancomycin + piperacillin-tazobactam + clindamycin (toxin suppression) | Surgical debridement urgent |
Antifungal Considerations
- Routine empiric antifungal: NOT recommended in undifferentiated sepsis
- Add empiric echinocandin if: recent abdominal surgery, TPN, liver failure, diabetes, or Candida colonization at multiple sites
Antibiotic Optimization
- Prolonged/extended infusion of beta-lactams improves pharmacodynamic target attainment (JAMA 2024, PMID 38864162 - meta-analysis showing mortality benefit)
- Administer beta-lactams before vancomycin when both are ordered
- De-escalate at 48-72 h based on culture results and clinical response
- Procalcitonin-guided duration reduces antibiotic exposure without worsening outcomes
- Target 7-10 days total for most infections (shorter for rapidly improving, controlled source)
5. Fluid Resuscitation
Initial Resuscitation
- 30 mL/kg of IV crystalloid in the first 3 hours for sepsis-induced hypoperfusion (MAP <65 or lactate ≥4 mmol/L)
- Balanced crystalloids (lactated Ringer's, PlasmaLyte) preferred over normal saline to reduce hyperchloremic acidosis risk
- Do not use hydroxyethyl starch (hetastarch) - associated with AKI and increased mortality
- Albumin (4-5%): consider when large volumes of crystalloid are required; NOT first-line but reasonable adjunct (Chest 2024, PMID 38447639)
Ongoing Resuscitation - Fluid Responsiveness Assessment
Beyond initial 30 mL/kg, use dynamic markers rather than static CVP:
- Passive leg raise (PLR) + cardiac output measurement
- Pulse pressure variation (PPV) or stroke volume variation (SVV) in mechanically ventilated patients with regular rhythm
- Point-of-care ultrasound (POCUS): IVC collapsibility, LV/RV function
- Lactate clearance: target ≥10-20% reduction per 2 hours as resuscitation endpoint
- Reassess after each 500 mL bolus; stop if no hemodynamic improvement or signs of fluid overload (pulmonary edema, rising CVP without MAP response)
Prolonged positive fluid balance in critically ill patients is independently associated with worse outcomes - avoid excessive fluid accumulation.
6. Vasopressors & Hemodynamic Support
Vasopressor Ladder
| Agent | Dose | Role |
|---|
| Norepinephrine | 0.01-0.5 mcg/kg/min (titrate) | First-line - maintain MAP ≥65 mmHg |
| Vasopressin | 0.03-0.04 units/min (fixed rate) | Add when NE ≥0.25-0.5 mcg/kg/min; do NOT use as sole agent |
| Epinephrine | 0.01-0.3 mcg/kg/min | Third-line, or alternative to vasopressin |
| Dopamine | 2-20 mcg/kg/min | Use only in highly selected circumstances (e.g., bradycardia with low cardiac output); generally avoid |
| Dobutamine | 2.5-10 mcg/kg/min | Add to NE in low cardiac output/myocardial dysfunction |
Agents NOT recommended: levosimendan, terlipressin (per SSC 2021).
MAP Target
- Standard target: MAP ≥65 mmHg
- Higher targets (70-80 mmHg) may be appropriate in patients with chronic hypertension or known severe atherosclerosis - individualize
- Do NOT target supranormal oxygen delivery values
Hemodynamic Monitoring
- Arterial line: place early in all vasopressor-dependent patients
- Central venous access: required for vasopressors
- Pulmonary artery catheter: not routinely recommended; consider in refractory mixed septic/cardiogenic shock
- Bedside echocardiography: highly recommended to assess cardiac function and guide therapy
7. Corticosteroids
Indication: vasopressor-dependent septic shock persisting despite adequate volume resuscitation
- Hydrocortisone 200 mg/day IV: given as 50 mg IV every 6 h (or continuous infusion 200 mg/24 h)
- Do NOT use ACTH stimulation test to determine eligibility
- Taper when vasopressors are no longer required
- Avoid high-dose corticosteroids (>300 mg/day hydrocortisone equivalent) - no benefit, increased harm
Evidence note: ADRENAL and APROCCHSS trials showed hydrocortisone shortens time to shock reversal and vasopressor weaning, with possible mortality benefit in most severe patients.
8. Respiratory Support
Oxygenation Target
- SpO2 target: 90-96% (avoid hyperoxia and hypoxia)
- If neurologically intact and no contraindication: high-flow nasal cannula (HFNC) preferred over non-invasive ventilation for hypoxemia without hypercapnia
Mechanical Ventilation (Sepsis-induced ARDS)
ARDS complicates ~7% of sepsis cases. Use lung-protective ventilation:
| Parameter | Target |
|---|
| Tidal volume | 6-8 mL/kg ideal body weight |
| Plateau pressure | ≤30 cmH2O |
| PEEP | Titrate to oxygenation/compliance; higher PEEP for moderate-severe ARDS |
| FiO2 | Minimum to achieve SpO2 90-96% |
| Mode | Volume-controlled AC initially; wean with spontaneous breathing trials |
| Head of bed | Elevate 30-45° (semi-recumbent) |
Rescue Strategies for Refractory ARDS
- Prone positioning: ≥12 h/day in moderate-severe ARDS (P/F ratio <150); associated with mortality reduction in PROSEVA trial
- Neuromuscular blockade: for prone positioning facilitation; intermittent bolus preferred over continuous infusion
- Recruitment maneuvers: use cautiously
- Venovenous ECMO (VV-ECMO): for severe ARDS failing conventional ventilation, if experienced team and resources available
Liberation from Ventilation
- Daily spontaneous breathing trials (SBT) once hemodynamically stable and low FiO2
- Use sedation protocols with daily sedation interruption ("awakening trials" - SAT/SBT paired protocol per PADIS guidelines, updated 2025, PMID 39982143)
9. Source Control
- Identify source of infection promptly - imaging (CT, ultrasound) as needed
- Perform source control as soon as feasible - do not delay
- Interventions include:
- Drainage of abscesses (percutaneous or surgical)
- Debridement of necrotizing tissue (NSTI - emergent OR)
- Removal of infected devices (IV catheters, urinary catheters, prosthetics)
- Bowel resection/repair (perforation)
- Decompression of biliary obstruction (ERCP, PTC, or surgical)
- Remove intravascular catheters if catheter-related infection is suspected and new access can be established
10. Organ Support - Additional Considerations
Acute Kidney Injury (AKI)
- Occurs in up to 2/3 of septic patients
- Avoid nephrotoxins: aminoglycosides, NSAIDs, IV contrast (when possible)
- Maintain MAP ≥65 mmHg (higher in CKD patients)
- Renal Replacement Therapy (RRT): initiate for refractory:
- Hyperkalemia (K+ >6 mEq/L)
- Severe metabolic acidosis (pH <7.2 with acute kidney injury)
- Fluid overload with oliguria/anuria
- Uremic complications
- No survival benefit to early prophylactic RRT (STARRT-AKI trial)
- Bicarbonate: use IV sodium bicarbonate to correct pH <7.2 in setting of AKI
Glycemic Control
- Target blood glucose: 140-180 mg/dL (7.8-10 mmol/L) in ICU patients
- Use continuous insulin infusion protocols with frequent glucose monitoring (every 1-2 h)
- Avoid tight glycemic control (80-110 mg/dL) - increases risk of hypoglycemia without mortality benefit (NICE-SUGAR trial)
Hematologic
- Red blood cell transfusion: threshold hemoglobin <7 g/dL (target 7-9 g/dL) in stable patients without active ischemia or hemorrhage
- Higher threshold (8-9 g/dL) acceptable in: active CAD, acute hemorrhage, symptomatic anemia
- Do NOT use erythropoietin to treat anemia in sepsis
- DVT prophylaxis: pharmacologic (LMWH preferred) unless contraindicated; mechanical (sequential compression devices) if pharmacologic contraindicated
- Stress ulcer prophylaxis: proton pump inhibitor or H2 blocker for patients with bleeding risk factors (mechanical ventilation, coagulopathy)
11. Antimicrobial Stewardship & De-escalation
| Time Point | Action |
|---|
| 0-6 h | Broad-spectrum empiric coverage |
| 24-48 h | Review cultures; narrow if organism identified |
| 48-72 h | Clinical reassessment - if improving and cultures negative, consider stopping antibiotics if infection diagnosis uncertain |
| Day 5-7 | Procalcitonin-guided stop decision; total course 7-10 days for most infections |
| Ongoing | Reassess need for antifungals, antivirals, and anaerobic coverage |
12. Monitoring & Reassessment Targets
| Parameter | Target |
|---|
| MAP | ≥65 mmHg |
| Lactate | Normalization or ≥10-20% clearance per 2 h |
| Urine output | ≥0.5 mL/kg/h |
| SpO2 | 90-96% |
| Blood glucose | 140-180 mg/dL |
| Hemoglobin | ≥7 g/dL (≥8-9 g/dL in CAD/hemorrhage) |
| Plateau pressure (ventilated) | ≤30 cmH2O |
| Tidal volume | 6-8 mL/kg IBW |
13. Adjunctive & Investigational Therapies
- Vitamin C + thiamine + hydrocortisone (HAT therapy): VICTAS trial (JAMA 2021) showed no benefit in vasopressor/ventilator-free days - do NOT use routinely
- Activated protein C (drotrecogin alfa): withdrawn from market; no benefit
- IVIG: insufficient evidence for routine use; may consider in toxic shock syndrome or streptococcal NF
- Selenium, glutamine: not recommended
- Therapeutic plasma exchange: investigational; not routine
14. Goals of Care
- Discuss prognosis and goals of care with patient/family early in ICU admission
- Integrate palliative care principles - manage symptoms, family communication
- Document goals, code status, and advance directives
- Re-evaluate goals if no improvement or new organ failure develops
Quick Reference Summary
HOUR 0-1
├─ Cultures x2 sets → Antibiotics (within 1 h if shock)
├─ Lactate → Repeat if >2 mmol/L
├─ IV access (peripheral ± central)
├─ 30 mL/kg balanced crystalloid if MAP <65 or lactate ≥4
└─ Norepinephrine if MAP <65 during/after fluids
HOUR 1-6
├─ ICU admission (target <6 h)
├─ Arterial line + central line
├─ Reassess fluid responsiveness (PLR, PPV, POCUS)
├─ Reassess vasopressor needs (add vasopressin if NE ≥0.25-0.5 mcg/kg/min)
├─ Source control (imaging → drainage/surgery)
├─ Respiratory: HFNC or intubate if needed
└─ Organ support (AKI, glycemia, DVT/GI prophylaxis)
HOUR 6-72
├─ Hydrocortisone 200 mg/day if vasopressor-dependent
├─ Prone/NMB if severe ARDS
├─ De-escalate antibiotics per cultures
├─ Daily SBT + SAT (ventilated patients)
└─ Goals of care discussion
Key Sources
- Harrison's Principles of Internal Medicine, 22e (2025) - Ch. 315 (Sepsis)
- Barash, Cullen & Stoelting's Clinical Anesthesia, 9e - Table 57-4
- Fischer's Mastery of Surgery, 8e - Sepsis Management
- Surviving Sepsis Campaign International Guidelines 2021
- ESICM Guidelines on Circulatory Shock and Hemodynamic Monitoring (2025) - PMID 41236566
- Abdul-Aziz et al., Prolonged vs Intermittent beta-Lactam Infusions in Sepsis, JAMA 2024 - PMID 38864162
- Lewis et al., PADIS ICU Guidelines Update 2025, Crit Care Med - PMID 39982143