The Septic Workup: A Rational Approach to Fever in the Hospitalized Patient
A Comprehensive Guide for Critical Care Practitioners
Dr Neeraj Manikath , claude.ai
Abstract
Fever in hospitalized patients presents a diagnostic challenge that requires systematic evaluation, judicious testing, and rational antibiotic stewardship. This review synthesizes current evidence on the approach to suspected sepsis, from redefining fever and systemic inflammatory response syndrome (SIRS) in the post-Sepsis-3 era to optimizing diagnostic workup and empiric antibiotic selection. We emphasize practical clinical pearls, common pitfalls ("oysters"), and evidence-based shortcuts ("hacks") to guide postgraduate trainees in critical care medicine.
Keywords: Sepsis, fever, diagnostic stewardship, procalcitonin, empiric antibiotics, blood cultures
Introduction
Fever in hospitalized patients triggers approximately 20-30% of infectious disease consultations and drives substantial antibiotic use.¹ The challenge lies not in recognizing fever, but in distinguishing infectious from non-infectious causes, identifying the source, and initiating appropriate therapy while avoiding diagnostic overreach and antimicrobial overuse. With the evolution from SIRS-based sepsis definitions to Sepsis-3 criteria emphasizing organ dysfunction, clinicians must recalibrate their approach to the febrile hospitalized patient.²
This review provides a structured framework for the septic workup, integrating traditional clinical skills with modern biomarkers and institutional antibiograms.
Defining "Fever" and "SIRS" in the Modern Era
What Constitutes Fever?
Traditional teaching defines fever as core temperature ≥38.3°C (101°F), but this threshold oversimplifies clinical reality.³ Temperature varies by:
- Measurement site: Rectal > core > oral > axillary (approximately 0.5°C differences)
- Circadian rhythm: Lower in morning, peaks late afternoon
- Age: Blunted response in elderly and immunosuppressed
- Antipyretics: May mask but don't eliminate infection
🔑 Pearl: A temperature of 38.0°C (100.4°F) in an elderly nursing home patient or neutropenic individual warrants the same concern as 38.5°C in a young healthy adult.
🦪 Oyster: Don't dismiss "low-grade" fevers (37.8-38.2°C) in high-risk populations. Studies show mortality increases even with temperatures >37.5°C in septic patients.⁴
Hypothermia as a Sepsis Marker
Temperature <36°C carries worse prognosis than fever in sepsis, with mortality rates of 40-60% in hypothermic septic patients.⁵ Hypothermia suggests:
- Severe physiologic decompensation
- Elderly or malnourished patients
- Overwhelming infection
- Poor cardiovascular reserve
⚡ Hack: In elderly patients with "failure to thrive" and hypothermia, think sepsis first—particularly urinary or intra-abdominal sources.
The Death of SIRS?
The Sepsis-3 consensus (2016) removed SIRS criteria from sepsis definitions, recognizing that SIRS:
- Occurs in 90% of ICU patients regardless of infection²
- Has poor specificity (approximately 50%) for infection⁶
- Led to overdiagnosis and excessive antibiotics
Current Paradigm: Sepsis = life-threatening organ dysfunction caused by dysregulated host response to infection (SOFA score increase ≥2 points).²
However, SIRS remains useful for:
- Initial bedside screening (high sensitivity ~97% when ≥2 criteria present)⁷
- Risk stratification in emergency departments
- Teaching differential diagnosis of acute inflammation
Clinical Framework:
| Condition | Temperature | SIRS | Organ Dysfunction | Infection Confirmed |
|---|---|---|---|---|
| SIRS | Variable | ≥2 criteria | No | No |
| Infection | +/− | Variable | No | Yes |
| Sepsis | Usually + | Usually + | Yes (SOFA ≥2) | Yes (suspected/confirmed) |
| Septic Shock | Variable | + | Yes + hypotension | Yes |
🔑 Pearl: Use qSOFA (quick SOFA: altered mentation, SBP ≤100, RR ≥22) for bedside screening, but don't let a negative qSOFA exclude sepsis—it's designed for prognostication, not diagnosis.⁸
The Art of the History and Physical: Finding the Source
Source identification drives appropriate antibiotic selection and source control interventions. Yet the source remains clinically unidentifiable in 30-40% of culture-negative sepsis cases.⁹
The Targeted History
Key Questions:
- Timeline: When did fever start? Acute (<24h) vs. subacute (days-weeks)?
- Prior antibiotics: What, when, and response? (Previous therapy selects for resistance)
- Immunosuppression: HIV, chemotherapy, biologics, chronic steroids?
- Recent procedures: Lines, catheters, surgery, endoscopy?
- Exposures: Travel, animals, water, sick contacts?
- Baseline status: Functional decline? New confusion?
🦪 Oyster: Don't forget to ask about prior cultures. A patient with MRSA bacteremia six months ago has ~30% chance of recurrent MRSA bacteremia.¹⁰
The Systematic Physical Examination
Most infections have localizing signs if sought systematically:
HEAD-TO-TOE APPROACH:
Head & Neck:
- Oropharynx: dental abscesses, thrush, Lemierre syndrome
- Sinuses: tenderness, purulent drainage
- Neck: meningismus, lymphadenopathy, JVD
🔑 Pearl: Check the external auditory canals in diabetics (malignant otitis externa from Pseudomonas) and examine the hard palate in neutropenic patients (invasive fungal infection).
Cardiovascular:
- New murmur: endocarditis until proven otherwise
- Osler nodes, Janeway lesions, splinter hemorrhages
- Roth spots on fundoscopy
Pulmonary:
- Asymmetric breath sounds, dullness, egophony
- Consider aspiration in altered patients
Abdominal:
- Systematic palpation all four quadrants
- Murphy's sign, McBurney's point, CVA tenderness
- Rectal exam: prostate, perirectal abscess
🦪 Oyster: "Abdominal exam negative" in the septic ICU patient is inadequate. Perform serial exams, consider CT imaging, and remember that peritoneal signs may be absent in immunosuppressed or elderly patients.
Skin & Soft Tissue:
- Cellulitis, abscess, necrotizing fasciitis (pain out of proportion, crepitus, hemorrhagic bullae)
- Catheter sites: erythema, purulence, tracking
- Pressure ulcers: stage, depth, exposed bone
🔑 Pearl: "Pan-scan the pressure points" in obtunded patients: sacrum, heels, occiput, scapulae. Unstageable pressure ulcers with necrotic eschar harbor osteomyelitis in 30-40% of cases.¹¹
Lines & Devices:
- Central lines: insertion site examination (remove dressing!)
- Foley catheters: when placed? Indication?
- ET tubes: purulent secretions, VAP criteria
- Surgical drains: character of output
⚡ Hack: Use the "48-hour rule" for central lines. If fever develops >48 hours after line placement in the absence of another source, the line is guilty until proven innocent.¹²
The "Occult" Sources
Source-Unknown Sepsis Checklist:
□ Sinusitis (especially in intubated patients >7 days)¹³
□ Acalculous cholecystitis (ICU patients, TPN, prolonged fasting)
□ C. difficile (recent antibiotics, anyone—yes, even on antibiotics currently)
□ Device infections (pacemakers, prosthetic joints, vascular grafts)
□ Dental abscess (poor dentition, immunosuppressed)
□ Epidural abscess (back pain + fever = MRI spine)
□ Endocarditis (especially in IVDU, structural heart disease, persistent bacteremia)
🔑 Pearl: In the ICU patient with unexplained fever >5 days, order maxillofacial CT to evaluate for sinusitis. Nasogastric and nasotracheal tubes increase risk 5-fold.¹³
Choosing Wisely: Which Cultures, When?
Diagnostic stewardship aims to minimize low-yield testing while maximizing pathogen identification. The mantra: culture what you'll treat, treat what you culture.¹⁴
Blood Cultures: The Gold Standard
Indications for Blood Cultures:
- Temperature >38.3°C or <36°C
- Sepsis or septic shock (by definition)
- Suspected endocarditis or line infection
- Neutropenic fever
- Unexplained hemodynamic instability
**Optimal Technique:**¹⁵
- Two sets (one set = one aerobic + one anaerobic bottle)
- Different sites (both arms, or arm + central line)
- 20 mL total volume (10 mL per set, split between aerobic/anaerobic)
- Before antibiotics whenever possible
- Adequate skin prep: chlorhexidine >30 seconds, allow to dry
🔑 Pearl: Blood culture volume is the single most important variable for yield. Each 1 mL increase in volume improves detection by 2-5%.¹⁶ Pediatric bottles in adults due to "difficult stick" significantly reduces sensitivity.
🦪 Oyster: "Pan-culturing" (reflexive blood, urine, sputum cultures with every fever) leads to:
- Overtreatment of colonization and contamination
- Unnecessary antibiotics for asymptomatic bacteriuria
- Cost: $50-200 per set, thousands per contaminated culture
- False positive rate: 0.6-6% depending on technique¹⁷
⚡ Hack: Use the "Shapiro rule": Blood cultures are low yield (<2%) when all of the following are absent:¹⁸
- Temperature >39.4°C
- Indwelling vascular catheter
- WBC >18,000 or <4,000
- Clinical suspicion for endocarditis
Central Line Cultures: To Draw or Not to Draw?
Peripheral vs. Central Cultures:
- Simultaneous peripheral + central (quantitative or time-to-positivity): Diagnostic for CLABSI if central grows ≥3 hours before peripheral OR ≥3-fold higher colony count¹⁹
- Central alone: Contaminates more frequently, cannot distinguish CLABSI from catheter colonization
- Never replace cultures through guidewire (same contamination)
When to Remove the Line:
- Purulence at exit site
- Tunnel infection
- Hemodynamic instability attributed to line
- S. aureus or Candida bloodstream infection
- Septic thrombophlebitis
🔑 Pearl: In stable patients with positive blood cultures, you can often observe for 72 hours after starting appropriate antibiotics. If repeat cultures clear, the line can stay; if persistently positive, remove the line.²⁰
Urine Cultures: The Most Over-Sent Test
True Indications:
- Urinary symptoms (dysuria, urgency, frequency)
- Flank pain or CVA tenderness
- Sepsis without alternative source
- Neutropenic fever
- Recent urologic procedure
- Pregnancy
NOT Indications:
- Altered mental status alone (unless no other cause)
- Presence of Foley catheter + fever (Foley bacteriuria ≠ UTI)
- Cloudy or malodorous urine alone
- Positive urinalysis without symptoms
🦪 Oyster: Asymptomatic bacteriuria (ASB) is present in:
- 50% of nursing home residents
- 100% of chronic Foley patients >30 days
- 10-15% of diabetics
Treating ASB does NOT improve outcomes (except pregnancy and pre-urologic surgery).²¹ It does select for resistant organisms and C. difficile.
⚡ Hack: Before sending urine culture in catheterized patients, ask: "If this grows pan-sensitive E. coli, will I treat it?" If the patient has no symptoms and an alternative fever source, the answer is no—don't send the culture.
Sputum Cultures: The Murky Middle
Challenge: Distinguishing colonization from infection
When Sputum Cultures Are Useful:
- VAP workup (with quantitative cultures: BAL or mini-BAL)
- Community-acquired pneumonia requiring ICU admission
- Suspected resistant organisms (MRSA, Pseudomonas)
- Immunocompromised patients
- Mycobacterial or fungal suspicion
When They're NOT Useful:
- Routine aspiration pneumonia (unlikely to change management)
- Lack of infiltrate on imaging
- Specimen quality: <25 PMNs and >10 epithelial cells/lpf (represents oropharyngeal contamination, not lower respiratory tract)²²
🔑 Pearl: For VAP diagnosis, quantitative cultures with thresholds (BAL ≥10⁴ CFU/mL, mini-BAL ≥10³ CFU/mL) reduce antibiotic overuse compared to non-quantitative cultures.²³
⚡ Hack: MRSA nasal PCR screening: Negative predictive value >99% for MRSA pneumonia. Negative test allows de-escalation from vancomycin in suspected VAP.²⁴
Other Focused Cultures
Wound/Soft Tissue:
- Aspirate or tissue biopsy preferred (swabs grow everything and nothing useful)
- Reserved for purulence or systemic signs
- Consider anaerobic culture for deep wounds
Stool:
- C. difficile testing only with diarrhea (≥3 loose stools/24h)
- Don't test formed stool or test of cure (remains positive for weeks)
- Nucleic acid amplification tests (NAAT) preferred over toxin EIA²⁵
CSF:
- Meningitis: headache + fever + meningismus (only 44% have all three)²⁶
- Altered mental status + fever with no other cause warrants consideration
- CT head before LP only if: mass lesion concern, immunocompromised, new seizure, focal deficit
Fungal Cultures:
- Candida risk factors: TPN, broad-spectrum antibiotics, immunosuppression, GI surgery
- (1,3)-β-D-glucan and galactomannan adjuncts for invasive candidiasis/aspergillosis
The Role of Biomarkers: Procalcitonin vs. CRP in Guiding Therapy
Biomarkers supplement but never replace clinical judgment. Understanding their kinetics, performance characteristics, and proper applications optimizes utility.
Procalcitonin (PCT)
Biology:
- Precursor of calcitonin, produced by C-cells of thyroid
- In bacterial infection, induced by bacterial endotoxin and inflammatory cytokines (IL-6, TNF-α)
- Half-life: 24-30 hours
- Rises in 4 hours, peaks at 24 hours
**Performance Characteristics:**²⁷,²⁸
| Cut-off | Interpretation | Sensitivity | Specificity |
|---|---|---|---|
| <0.25 μg/L | Bacterial infection unlikely | 85-90% | 40-50% |
| 0.25-0.5 | Possible bacterial infection | — | — |
| >0.5 | Likely bacterial infection | 60-75% | 70-80% |
| >2.0 | Sepsis/severe infection | 50-60% | 85-90% |
Clinical Applications:
- Antibiotic Initiation Decision: Low PCT (<0.25 μg/L) in stable patients supports withholding antibiotics in:
- Acute bronchitis/COPD exacerbation
- Possible CAP with equivocal findings
- Antibiotic De-escalation: Multiple trials show PCT-guided protocols reduce antibiotic duration without increasing mortality:
- Stop if PCT decreases >80% from peak OR <0.5 μg/L
- Meta-analyses show 2-3 days shorter duration²⁹,³⁰
🔑 Pearl: PCT-guided therapy reduces antibiotic exposure by 20-30% without increasing mortality. The PRORATA trial showed 2.7-day reduction in antibiotic duration using stop criteria: PCT <0.5 μg/L or 80% decrease from peak.²⁹
Limitations of PCT:
False Elevations (non-bacterial):
- Severe trauma, surgery, burns (first 48 hours)
- Small cell lung cancer, medullary thyroid cancer
- Severe cardiogenic shock
- Massive pulmonary embolism
- Heat stroke
- Severe pancreatitis
False Negatives:
- Localized infections without systemic involvement
- Early infection (<4-6 hours)
- Atypical bacteria (Legionella, Mycoplasma, Chlamydia)
- Some fungal and viral infections
- Immunosuppressed patients with blunted response
🦪 Oyster: PCT >5 μg/L in a patient without obvious infection should prompt search for occult malignancy (particularly small cell or medullary thyroid carcinoma).³¹
C-Reactive Protein (CRP)
Biology:
- Acute phase reactant produced by hepatocytes
- Induced by IL-6
- Half-life: 19 hours
- Rises in 6-12 hours, peaks at 48 hours
Performance:
- More sensitive (90-95%) but less specific (50-60%) than PCT for bacterial infection³²
- Elevated in virtually all inflammatory conditions
- Cutoff >50-100 mg/L suggests bacterial infection
Advantages Over PCT:
- Less expensive ($10 vs. $25-50)
- Widely available
- Longer track record
Disadvantages:
- Elevated in: autoimmune disease, malignancy, tissue injury, viral infections
- Slower kinetics (less useful for acute decision-making)
- Less validated for antibiotic stewardship
🔑 Pearl: Serial CRP measurements outperform single values. Failure of CRP to decline by 50% at 48-72 hours suggests:
- Treatment failure
- Wrong antibiotic
- Undrained collection
- Alternative diagnosis³³
PCT vs. CRP: Head-to-Head Comparison
| Feature | Procalcitonin | CRP |
|---|---|---|
| Specificity for bacterial infection | Higher (70-80%) | Lower (50-60%) |
| Time to elevation | 4 hours | 6-12 hours |
| Peak | 24 hours | 48 hours |
| Half-life | 24-30 hours | 19 hours |
| Cost | Higher | Lower |
| Validated for antibiotic stewardship | Yes (strong evidence) | Limited evidence |
| Affected by renal failure | Yes (accumulates) | No |
⚡ Hack: Use PCT for decision-making (start/stop antibiotics) and CRP for monitoring (treatment response, locating collections). The combination outperforms either alone.³⁴
Other Emerging Biomarkers
Presepsin (sCD14-ST):
- Rises earlier than PCT (2-3 hours)
- Higher specificity than CRP
- Limited availability, higher cost
- Promising but requires more validation³⁵
Interleukin-6:
- Very early marker (1-2 hours)
- High sensitivity but poor specificity
- Expensive, not widely available
Lactate:
- Not specific for infection (elevated in any shock, seizures, ischemia, medications)
- Prognostic value: Lactate >2 mmol/L defines sepsis-induced hypoperfusion
- Clearance matters: Failure to decrease by 10% at 6 hours associated with increased mortality³⁶
Empiric Antibiotics: Tailoring Your Choice to the Patient and the Hospital's Bug Map
Empiric antibiotic selection is both art and science, balancing the need for adequate initial coverage (mortality increases 7% per hour of delay in septic shock)³⁷ against antimicrobial stewardship principles.
The Framework: Four Key Questions
1. Where did they acquire this infection?
- Community-acquired
- Healthcare-associated
- Hospital-acquired (>48 hours)
2. What's the likely source?
- Determines bacterial spectrum (gram-positive, gram-negative, anaerobes)
3. What are the patient risk factors for resistance?
- Prior antibiotics (especially within 90 days)
- Prior resistant organisms
- Chronic care facility
- Immunosuppression
- Recent hospitalization
4. What is my hospital's resistance profile?
- Local antibiogram
- Unit-specific patterns (ICU vs. floor)
- Endemic organisms (e.g., regional Candida auris)
Common Scenarios and Empiric Choices
Scenario 1: Community-Acquired Pneumonia (Sepsis)
Low Risk for MRSA/Pseudomonas:
- β-lactam + macrolide: Ceftriaxone 2g IV daily + azithromycin 500mg IV daily
- Respiratory fluoroquinolone: Levofloxacin 750mg IV daily (if no contraindications)
Risk factors for MRSA (prior MRSA, IVDU, recent influenza, necrotizing pneumonia):
- Add vancomycin 15-20 mg/kg IV q8-12h (target trough 15-20 mcg/mL)
- Consider linezolid 600mg IV q12h if vancomycin MIC >1.5 mg/L
Risk factors for Pseudomonas (structural lung disease, recent antibiotics, bronchiectasis):
- Antipseudomonal β-lactam: Piperacillin-tazobactam 4.5g IV q6h, cefepime 2g IV q8h, or meropenem 1g IV q8h
- Consider dual gram-negative coverage (add aminoglycoside or fluoroquinolone) if septic shock
🔑 Pearl: De-escalate MRSA coverage if MRSA nasal PCR negative (NPV 99%) and patient improving at 48-72 hours.²⁴
Scenario 2: Intra-Abdominal Infection
Community-acquired, mild-moderate:
- Ceftriaxone 2g IV daily + metronidazole 500mg IV q8h
- Ertapenem 1g IV daily
Healthcare-associated or severe:
- Piperacillin-tazobactam 4.5g IV q6h (extended infusion)
- Cefepime 2g IV q8h + metronidazole 500mg IV q8h
- Meropenem 1g IV q8h (if ESBL risk high)
Add empiric Candida coverage if:
- Recent abdominal surgery
- Anastomotic leak
- Recurrent perforation
- Upper GI source with severe sepsis → Fluconazole 800mg load, then 400mg daily OR echinocandin (if azole-resistant risk or unstable)³⁸
🦪 Oyster: Don't forget source control—antibiotics alone won't cure undrained abscesses or perforated viscus. "No antimicrobial can sterilize a collection that requires drainage."
Scenario 3: Urosepsis
Community-acquired:
- Ceftriaxone 2g IV daily
- If "worried" (septic shock, recent hospitalization): Piperacillin-tazobactam 3.375g IV q6h
Healthcare-associated:
- Risk factors for ESBL: Recent fluoroquinolone or cephalosporin use, recent hospitalization, chronic care facility → Carbapenem (ertapenem 1g IV daily if mild; meropenem 1g IV q8h if severe)
Obstructive uropathy:
- Broader spectrum (Piperacillin-tazobactam or carbapenem)
- Urgent drainage (nephrostomy or ureteral stent)
- Consider Enterococcus coverage if risk factors (add ampicillin)
🔑 Pearl: Fluoroquinolones have excellent urinary penetration but resistance rates now exceed 30% in many areas for E. coli. Review your local antibiogram before using empirically.³⁹
Scenario 4: Skin and Soft Tissue Infection
Cellulitis (non-purulent):
- Cefazolin 2g IV q8h (for streptococci)
- Vancomycin if MRSA risk (see below)
Purulent/abscess:
- MRSA coverage: Vancomycin 15-20 mg/kg IV q8-12h
- Alternative: Daptomycin 6 mg/kg IV daily (not for pneumonia—inactivated by surfactant)
Necrotizing fasciitis:
- Surgical emergency first
- Polymicrobial: Piperacillin-tazobactam 4.5g IV q6h + vancomycin 15-20 mg/kg IV q8-12h + clindamycin 900mg IV q8h
- Clindamycin for toxin suppression (especially Group A Strep toxic shock)⁴⁰
MRSA Risk Factors:
- Prior MRSA infection/colonization
- Injection drug use
- Incarceration
- Close contact with MRSA case
- Abscess requiring I&D
- Local prevalence >30%
⚡ Hack: In necrotizing soft tissue infections, the WBC can be paradoxically normal or low (median 11,000). Don't let "reassuring labs" delay surgical exploration.⁴¹
Scenario 5: Central Line-Associated Bloodstream Infection (CLABSI)
Empiric coverage:
- Vancomycin 15-20 mg/kg IV q8-12h (for coagulase-negative staph, MRSA)
- Gram-negative coverage based on risk:
- Low risk: Ceftriaxone
- Immunosuppressed, neutropenic, or ICU: Cefepime or antipseudomonal penicillin
Consider Candida coverage if:
- TPN
- Immunosuppression
- Broad-spectrum antibiotics >7 days
- Persistent fever despite antibiotics
- GI surgery/perforation → Echinocandin (micafungin 100mg IV daily, caspofungin, anidulafungin) preferred over fluconazole for empiric therapy⁴²
🔑 Pearl: If blood cultures grow coagulase-negative staphylococci in 1 of 2 sets, it's usually a contaminant (90% PPV if both sets positive, 10% if single).⁴³ Repeat cultures before committing to 7-14 days of antibiotics.
Understanding Your Hospital's "Bug Map" (Antibiogram)
The antibiogram is your local resistance guide, typically updated annually. Key elements:
How to Read It:
- Rows: Organisms
- Columns: Antibiotics
- Numbers: % susceptibility
Red Flags:
- MRSA prevalence >30%: Consider empiric vancomycin for suspected staph infections
- Fluoroquinolone resistance >20% in E. coli: Avoid empiric FQ for UTI
- ESBL prevalence >10%: Lower threshold for carbapenem use
⚡ Hack: Create a pocket card of your ICU's top 5 organisms and their susceptibilities. Update yearly. This becomes your mental "empiric therapy database."
Special Populations
Neutropenic Fever (ANC <500)
- High risk: Prolonged neutropenia expected, significant comorbidity
- Empiric: Antipseudomonal β-lactam monotherapy (cefepime, meropenem, piperacillin-tazobactam)
- Add vancomycin ONLY if: skin/soft tissue infection, hemodynamic instability, pneumonia, suspected CLABSI, mucositis
- Add mold-active azole (voriconazole/posaconazole) if persistent fever >4-7 days despite antibiotics⁴⁴
🦪 Oyster: Don't routinely add vancomycin to empiric neutropenic fever regimens. It doesn't reduce mortality and selects for VRE.⁴⁵
Immunocompromised (Non-Neutropenic)
Considerations:
- Solid organ transplant: CMV, pneumocystis, listeria, cryptococcus, aspergillus
- Biologics (TNF-α inhibitors): TB reactivation, fungal infections
- High-dose steroids: Strongyloides hyperinfection, pneumocystis
Expand differential beyond bacteria; lower threshold for fungal coverage and atypical cultures.
Antibiotic Dosing Pearls
β-Lactams:
- Extended/continuous infusions improve outcomes in severe sepsis: Piperacillin-tazobactam 3.375g over 4 hours q8h vs. 4.5g over 30 min q6h⁴⁶
- Time-dependent killing: Keep free drug concentration above MIC for 50-70% of dosing interval
Vancomycin:
- AUC-guided dosing preferred over trough monitoring (target AUC/MIC >400)⁴⁷
- If using troughs: 15-20 mcg/mL for serious infections (endocarditis, osteomyelitis, pneumonia)
- Load with 25-30 mg/kg for rapid attainment in septic shock
Aminoglycosides:
- Once-daily dosing: 5-7 mg/kg gentamicin/tobramycin, 15-20 mg/kg amikacin
- Primarily for synergy (endocarditis) or gram-negative coverage
- Monitor peaks (severe infections: gentamicin 20-30 mcg/mL) and troughs (<1-2 mcg/mL to reduce nephrotoxicity)
Fluoroquinolones:
- Concentration-dependent killing: Maximize Cmax/MIC ratio
- Levofloxacin 750mg daily superior to 500mg for severe infections
🔑 Pearl: In septic shock with capillary leak, use loading doses for hydrophilic antibiotics (β-lactams, vancomycin, aminoglycosides). Volume of distribution increases 30-50%, reducing peak concentrations.⁴⁸
De-escalation: The Other Half of Stewardship
Starting broad is often necessary; staying broad is malpractice.
De-escalation Checklist (at 48-72 hours):
□ Culture results available? Narrow to organism-specific therapy
□ Patient improving clinically? Consider shorter duration
□ PCT decreased >80% or <0.5? Consider stopping
□ Source controlled? Confirm drainage, line removal if needed
□ Negative cultures + alternative diagnosis? Stop antibiotics
**Typical Duration Targets:**⁴⁹
| Infection | Duration |
|---|---|
| Uncomplicated UTI | 5-7 days (catheter-associated) |
| CAP | 5 days if afebrile 48h |
| Intra-abdominal (source control) | 4-7 days |
| Gram-negative bacteremia (uncomplicated) | 7 days |
| S. aureus bacteremia | 14 days (uncomplicated); 4-6 weeks (complicated) |
| VAP | 7-8 days (15% non-inferiority margin) |
🦪 Oyster: "Completing the course" of a 14-day antibiotic prescribed in the ICU, continued despite negative cultures and clinical improvement, is a cognitive bias. Duration should be re-evaluated daily,not predetermined.⁵⁰
Practical Pearls, Oysters, and Hacks: A Synthesis
Diagnostic Pearls 💎
-
The "Sepsis Six" in the First Hour:
- Blood cultures (before antibiotics)
- Lactate measurement
- Broad-spectrum antibiotics
- Fluid resuscitation (30 mL/kg crystalloid if hypotensive or lactate >4)
- Vasopressors (if hypotensive despite fluids)
- Reassess perfusion markers⁵¹
-
The "3-Day Rule": If a patient isn't improving by 72 hours, revisit the entire workup:
- Wrong bug (resistant organism, inadequate spectrum)
- Wrong drug (inadequate dosing, poor penetration)
- Wrong source (missed abscess, undrained collection)
- Wrong diagnosis (non-infectious mimic)
-
Fever Patterns Can Guide Diagnosis:
- Continuous/sustained: Gram-negative sepsis, typhoid
- Intermittent/hectic: Abscess, malignancy
- Double quotidian (2 peaks/day): Visceral leishmaniasis, Still's disease
- Pel-Ebstein (cyclic): Hodgkin lymphoma (Pattern recognition has low sensitivity but high specificity when present)
-
The "Culture-Negative Sepsis" Differential:
- Prior antibiotics (most common—80% of cases)
- Fastidious organisms: HACEK, Brucella, Legionella, mycobacteria
- Intracellular: Rickettsia, Ehrlichia, Anaplasma
- Fungal: Endemic mycoses (histoplasma, blastomyces)
- Viral: Influenza, CMV, HSV, VZV
- Non-infectious: Drug fever, PE, MI, pancreatitis, adrenal crisis
-
The Paradoxical Hypothermic Patient: Mortality in sepsis with hypothermia (<36°C) approaches 50-60% vs. 25-30% with fever.⁵ Hypothermia suggests:
- Extreme age (elderly, neonates)
- Severe malnutrition/cirrhosis
- Overwhelming infection (toxic shock syndrome)
- Environmental exposure → Requires aggressive resuscitation and source control
Critical Oysters 🦪 (Common Pitfalls)
-
"The Urine Culture Made Me Do It" Syndrome: Treating asymptomatic bacteriuria is the #1 cause of inappropriate antibiotic use in hospitalized patients.²¹ If the patient has no urinary symptoms and another source of fever, don't send the culture.
-
The "Pan-Culture" Reflex: Reflexively ordering blood/urine/sputum cultures with every temperature spike leads to:
- Overdiagnosis of colonization (especially sputum, catheter urine)
- Treatment of contaminants (especially single-bottle coagulase-negative staph)
- Unnecessary antibiotic exposure → Apply clinical reasoning before ordering cultures
-
Mistaking Colonization for Infection:
- Sputum growing MRSA in a patient with CHF exacerbation (no pneumonia)
- Foley urine growing Candida in asymptomatic patient
- Central line tip culture positive (don't routinely culture tips—only if clinical suspicion)⁵² → Cultures must be interpreted in clinical context
-
The "Fever = Infection" Fallacy: Non-infectious causes of fever in hospitalized patients:
- Drug fever: β-lactams, sulfonamides, anticonvulsants, allopurinol (typically 7-10 days after starting)
- Venous thromboembolism: PE, DVT (10-15% have fever)
- Acute MI: Transmural infarction with pericarditis
- Acute pancreatitis: Inflammatory, not infected (unless necrotizing >7 days)
- Gout/pseudogout: Crystal arthropathy mimics septic joint
- Adrenal insufficiency: Often with hypothermia, but can present with fever
- Malignancy: Lymphoma, renal cell carcinoma (paraneoplastic)
- Transfusion reaction: Febrile non-hemolytic, TRALI
- Neuroleptic malignant syndrome / serotonin syndrome
- Thyroid storm
⚡ Hack: Use the "Fever Workup Timeout" at 72 hours. If antibiotics started empirically, ask: "Do I have evidence of infection?" If no—stop antibiotics and search for non-infectious causes.
-
Underdosing in Critical Illness: Pharmacokinetics are profoundly altered in sepsis:
- Increased volume of distribution (capillary leak, fluid resuscitation)
- Augmented renal clearance in younger patients (CrCl >130 mL/min despite "normal" creatinine)
- Hypoalbuminemia (reduced protein binding of drugs) → Standard doses may be subtherapeutic. Consider loading doses and therapeutic drug monitoring.⁴⁸
-
The "Good Pressors, Bad Bugs" Paradox: Patients can be hemodynamically stable on vasopressors while harboring inadequately treated infection. Don't be reassured by "stable on minimal pressors"—if source control is inadequate or antibiotics wrong, outcome remains poor. Clinical improvement requires both hemodynamic stabilization AND infection control.
-
Delayed Source Control: "Time is tissue" applies to sepsis source control as much as antibiotics:
- Intra-abdominal abscess >3 cm: Drain within 12-24 hours
- Necrotizing soft tissue infection: Surgery within 6 hours (mortality increases 9% per hour delay)⁵³
- Obstructive pyelonephritis: Urgent decompression
- Empyema with loculations: Chest tube ± VATS → Antibiotics are adjunctive; source control is definitive
Evidence-Based Hacks ⚡
-
The "Antibiotic Timeout" at 48-72 Hours: Implement a mandatory reassessment:
- Are cultures positive? De-escalate to targeted therapy
- Are cultures negative? Consider stopping if alternative diagnosis
- Is patient improving? Consider shorter course
- Is PCT declining? Use to guide duration This simple intervention reduces antibiotic days by 20-25% in most studies.⁵⁴
-
The MRSA Nasal PCR Strategy: In suspected pneumonia or bacteremia:
- Negative MRSA nasal screen → 99% NPV for MRSA infection²⁴
- Use to de-escalate vancomycin at 48 hours if patient improving
- Saves average 3-4 days of vancomycin per patient
- Cost: $20-30 test vs. $200-400 in vancomycin + monitoring costs
-
The "Double-Cover Smart" Approach: Dual gram-negative coverage (β-lactam + aminoglycoside or fluoroquinolone) improves outcomes ONLY in:
- Septic shock with suspected gram-negative infection
- Neutropenic fever with severe sepsis
- Infections with high resistance risk (nosocomial Pseudomonas)
Duration: De-escalate to monotherapy within 3-5 days once organism identified and susceptibilities known. Prolonged dual therapy increases toxicity without benefit.⁵⁵
-
The "Short-Course Revolution": Recent trials support shorter antibiotic durations:
- VAP: 7 days non-inferior to 14 days (except non-fermenting GNR)⁵⁶
- Intra-abdominal infection with source control: 4 days if adequate drainage and clinical improvement⁵⁷
- Uncomplicated gram-negative bacteremia: 7 days vs. 14 days (ongoing trials suggest equipoise)⁵⁸ → Shift from "complete the course" to "treat until clinical resolution + margin"
-
The "Organ Penetration Pearl": Match antibiotic choice to site of infection based on penetration:
Site Good Penetration Poor Penetration CNS Fluoroquinolones, ceftriaxone, meropenem Vancomycin (use high dose), aminoglycosides Bone Fluoroquinolones, rifampin, linezolid Aminoglycosides, β-lactams (need high dose) Prostate Fluoroquinolones, TMP-SMX β-lactams, aminoglycosides Lung abscess Linezolid, clindamycin, metronidazole Vancomycin, aminoglycosides Urine Fluoroquinolones, nitrofurantoin Tigecycline, daptomycin -
The "Biomarker-Guided De-escalation Protocol": Implement PCT-guided stopping rules:
- Stop antibiotics if:
- PCT <0.5 μg/L OR decreased >80% from peak, AND
- Clinical improvement (afebrile >24h, hemodynamically stable), AND
- Source controlled
- Reduces antibiotic duration by 2.3 days on average without increasing mortality²⁹,³⁰
- Stop antibiotics if:
-
The "ID Phone-a-Friend" Threshold: Call infectious disease consultation for:
- Persistent bacteremia (>72 hours of appropriate therapy)
- S. aureus bacteremia (endocarditis evaluation)
- Candida bloodstream infection (source control + duration)
- Multidrug-resistant organisms (ESBL, CRE, XDR-Pseudomonas)
- Immunocompromised patients with fever of unknown origin
- Consideration of oral step-down for prolonged IV therapy (OPAT)
Studies show ID consultation improves outcomes and reduces mortality by 25-56% in S. aureus bacteremia.⁵⁹
-
The "Daily Foley Audit": Every morning, ask: "Does this patient still need a urinary catheter?"
- Indications: Strict I/O monitoring in shock, urologic surgery/obstruction, palliative care
- NOT indications: Incontinence, convenience, "still on pressors" (if stable)
- Remove catheter → reduces CAUTI risk by 50% and eliminates the temptation to send urine cultures⁶⁰
-
The "Antibiogram App": Create unit-specific antibiograms accessible on smartphones:
- Top 10 organisms with susceptibility percentages
- Update quarterly (more responsive than annual hospital-wide)
- Include Sankey diagrams showing resistance trends → Enables real-time evidence-based empiric choices at the bedside
-
The "Procalcitonin-Plus-Lactate" Strategy: Combined biomarker approach:
- High PCT (>2) + High lactate (>4): Aggressive resuscitation + broad antibiotics
- High PCT + Normal lactate: Infection present but compensated → empiric antibiotics, may avoid ICU
- Low PCT (<0.5) + High lactate: Consider non-infectious shock (cardiogenic, hemorrhagic)
- Low PCT + Normal lactate: Infection unlikely, search for alternative diagnosis
This 2×2 matrix improves diagnostic accuracy and triage decisions.⁶¹
Algorithmic Approach: Putting It All Together
The Systematic "Sepsis Evaluation Protocol"
STEP 1: IMMEDIATE ASSESSMENT (0-15 minutes)
- Vital signs including temperature trend
- Quick history: immunosuppression, recent antibiotics, procedures
- Focused exam for source (skin, lungs, abdomen, lines)
- Labs: CBC, CMP, lactate, blood cultures × 2 (before antibiotics)
- qSOFA/SOFA score
Decision Point: Does patient meet Sepsis-3 criteria (infection + SOFA ≥2)?
STEP 2: SOURCE IDENTIFICATION (15-60 minutes)
- Targeted cultures based on suspected source:
- Respiratory: CXR, sputum (if infiltrate), consider MRSA nasal PCR
- Urinary: UA with reflex culture (ONLY if symptoms)
- Intra-abdominal: CT abdomen/pelvis with IV contrast
- Soft tissue: Imaging for deep space infection/necrotizing features
- Central line: Examine site, consider paired cultures if CLABSI suspected
- Unknown: Consider CT chest/abdomen/pelvis, LP if altered mental status
Decision Point: Is source control needed emergently? (abscess drainage, line removal, surgery)
STEP 3: RISK STRATIFICATION FOR RESISTANCE (15-30 minutes)
Patient Factors:
□ Recent antibiotics within 90 days?
□ Recent hospitalization/nursing home?
□ Known colonization with MDR organism?
□ Immunosuppression?
□ Structural lung disease (bronchiectasis, CF)?
□ Prior MDR infection?
Institutional Factors:
□ High local MRSA prevalence (>30%)?
□ High ESBL rates (>10-15%)?
□ Endemic CRE or XDR organisms?
Decision Point: Low risk → standard empiric therapy; High risk → escalate coverage
STEP 4: EMPIRIC ANTIBIOTIC SELECTION (Goal: <60 minutes from recognition)
Use Decision Matrix:
| Source | Community | Healthcare-Associated | Add if Septic Shock |
|---|---|---|---|
| Pneumonia | Ceftriaxone + azithro | Pip-tazo or cefepime ± vanco | Dual GN coverage |
| Urine | Ceftriaxone | Pip-tazo or carbapenem | Amp for enterococcus |
| Abdomen | Ceftriaxone + metro | Pip-tazo or meropenem | Echinocandin if risk |
| Skin/soft tissue | Cefazolin | Vancomycin ± pip-tazo | Clinda (toxin suppression) |
| Line | Vancomycin + cefepime | Vancomycin + cefepime | Echinocandin if TPN/IC risk |
| Unknown | Vancomycin + pip-tazo | Vancomycin + meropenem | Dual GN + fungal coverage |
Load appropriately: Vancomycin 25-30 mg/kg, β-lactams standard doses
Decision Point: Document indication, expected duration, and plan for reassessment in 48-72h
STEP 5: ADJUNCTIVE MEASURES
□ Lactate clearance monitoring (repeat at 2-6 hours)
□ Procalcitonin baseline (for future de-escalation)
□ Resuscitation: 30 mL/kg crystalloid if lactate ≥4 or hypotensive
□ Vasopressors: Norepinephrine first-line (target MAP ≥65)
□ Source control: Expedite procedures within 6-12 hours
□ Consider stress-dose steroids if refractory shock
STEP 6: THE 48-72 HOUR REASSESSMENT (MANDATORY)
Clinical Response:
□ Afebrile >24 hours?
□ Hemodynamically improved (off pressors or weaning)?
□ Lactate normalized?
□ Source controlled?
Microbiologic Data:
□ Cultures finalized?
□ Organism susceptibilities available?
□ Contaminants vs. pathogens clarified?
Biomarkers:
□ PCT decreased >80% or <0.5?
□ CRP decreased >50%?
DECISION ALGORITHM:
Cultures POSITIVE
├─ Susceptibilities available?
│ ├─ Yes → DE-ESCALATE to narrow-spectrum targeted therapy
│ └─ No → Continue empiric, call micro for preliminary results
└─ Clinical improvement?
├─ Yes → Continue therapy, set duration goal
└─ No → Consider wrong bug/drug/source, ID consult
Cultures NEGATIVE
├─ Alternative diagnosis identified?
│ ├─ Yes → STOP antibiotics
│ └─ No → Continue below
└─ Clinical improvement?
├─ Yes + PCT <0.5 → Consider STOPPING antibiotics
├─ Yes + PCT 0.5-2 → Short course (5-7 days)
└─ No → Broaden ddx (atypicals, fungal, non-infectious), ID consult
STEP 7: DURATION AND DISCHARGE PLANNING
Set Target Duration Based on:
- Source (see earlier table)
- Organism (S. aureus 14d minimum, others 5-7d)
- Response (PCT/CRP kinetics)
- Complications (undrained abscess = longer)
Plan Antibiotic Transition:
- IV to PO when: Afebrile >24h, hemodynamically stable, GI function intact, bioavailability adequate
- High bioavailability options: Fluoroquinolones (100%), linezolid (100%), metronidazole (100%), fluconazole (90%)
- Consider OPAT (outpatient parenteral therapy) for prolonged courses: Ceftriaxone, ertapenem, daptomycin
Special Considerations: The Sepsis Mimics
Not every fever with organ dysfunction is sepsis. Beware these common mimics:
1. Acute Pancreatitis
- Fever in 30-40%, leukocytosis, SIRS criteria met
- Early phase is sterile inflammation
- Don't give antibiotics empirically unless infected necrosis suspected (typically >7 days, gas in pancreas, persistent SIRS)⁶²
2. Acute Coronary Syndrome
- Transmural MI can cause fever (24-48 hours post-infarction)
- Inflammatory response, not infection
- Troponin, ECG, and imaging clarify
3. Pulmonary Embolism
- Fever in 10-15% of PE cases
- Tachycardia, hypoxia, elevated WBC mimic sepsis
- D-dimer, CT angiography diagnostic
4. Drug Fever
- Onset typically 7-10 days after starting culprit drug
- Fever pattern: Often high-grade, may have relative bradycardia
- Eosinophilia (30% of cases), rash (20%)
- Common culprits: β-lactams, sulfonamides, phenytoin, allopurinol
- Diagnosis of exclusion: Defervescence within 72 hours of stopping drug⁶³
⚡ Hack: Consider drug fever in any patient who's been on antibiotics >7 days with persistent fever despite "appropriate" therapy and negative repeat cultures.
5. Adrenal Insufficiency
- Fever (often low-grade), hypotension, altered mental status
- Hyponatremia, hyperkalemia, hypoglycemia, eosinophilia
- Random cortisol <5 μg/dL diagnostic; 5-15 equivocal (ACTH stim test)
- Risk factors: Chronic steroids, septic shock, hemorrhage (Waterhouse-Friderichsen)
6. Thyroid Storm
- High fever (often >40°C), tachycardia, agitation, diarrhea
- Triggers: Infection, surgery, iodine load, MI
- TSH suppressed, free T4/T3 elevated
- Treatment: PTU/methimazole + β-blocker + steroids + iodine (1h after thionamide)
7. Neuroleptic Malignant Syndrome / Serotonin Syndrome
- NMS: Antipsychotic exposure, rigidity, hyperthermia, CK elevation
- Serotonin syndrome: Serotonergic drug, clonus, hyperreflexia, agitation
- Both can mimic sepsis with fever + altered mental status
- Treatment: Stop offending agent, supportive care, bromocriptine (NMS), cyproheptadine (serotonin syndrome)
Antibiotic Stewardship: Beyond the Individual Patient
Antibiotic resistance is a public health crisis. Every prescription has consequences beyond the individual patient.
The Collateral Damage of Antibiotics
Patient-Level:
- C. difficile infection: Risk increases 2-10× with any antibiotic; highest with fluoroquinolones, clindamycin, cephalosporins⁶⁴
- Selection of resistant organisms: ESBL, VRE, MRSA colonization
- Adverse effects: Renal toxicity (vanco, aminoglycosides), QT prolongation (fluoroquinolones, azithromycin), tendonopathy, seizures
Population-Level:
- Resistance transmission: MDR organisms spread in ICUs and hospitals
- Microbiome disruption: Long-term dysbiosis linked to recurrent infections, metabolic disorders
- Healthcare costs: Each C. difficile infection costs $10,000-20,000; MDR infections cost 2-3× more than susceptible
Stewardship Principles
-
Audit and Feedback: Regular review of antibiotic prescribing with individualized feedback reduces inappropriate use by 20-30%⁶⁵
-
Prospective Authorization: Restrict broad-spectrum agents (carbapenems, anti-MRSA, antifungals) to appropriate indications
-
IV-to-PO Conversion: Transition when clinically appropriate (saves costs, reduces line complications)
-
Dose Optimization: TDM for vancomycin, aminoglycosides; extended infusions for β-lactams
-
De-escalation Culture: Make narrow-spectrum therapy the default, not the exception
🔑 Pearl: Stewardship is not about "restricting antibiotics" but about optimizing appropriateness—right drug, right dose, right duration, right patient.
Future Directions: Emerging Diagnostics and Therapies
Rapid Diagnostics
Multiplex PCR Panels:
- Blood culture identification (BC-GPC, BC-GN panels): Organisms identified in 1-2 hours vs. 24-48 hours
- Respiratory panels: Detects viruses + atypical bacteria
- GI panels: Identifies C. difficile, bacterial enteropathogens
- Impact: Earlier targeted therapy, reduced broad-spectrum use⁶⁶
T2 Candida:
- Direct detection of Candida from blood (bypasses culture)
- Results in 3-5 hours vs. 2-5 days for culture
- Sensitivity 91%, specificity 99%
- Useful in high-risk patients (TPN, immunosuppression)⁶⁷
Metagenomic Next-Generation Sequencing (mNGS):
- Unbiased pathogen detection from any specimen
- Identifies rare, fastidious, or unexpected organisms
- Cost and turnaround time improving
- Current role: Diagnostic odysseys, immunocompromised patients⁶⁸
Novel Therapeutics
New Antibiotics for MDR Organisms:
- Ceftazidime-avibactam: CRE, ESBL, AmpC
- Ceftolozane-tazobactam: XDR Pseudomonas
- Meropenem-vaborbactam: CRE (including KPC)
- Plazomicin: Aminoglycoside-resistant GNRs
- Cefiderocol: Siderophore cephalosporin for XDR organisms
Immunomodulation:
- Selective immunosuppression in hyperinflammatory sepsis (cytokine storm)
- IL-1 antagonists, GM-CSF, IFN-γ supplementation
- Personalized approaches based on immune phenotyping⁶⁹
Precision Medicine in Sepsis
Emerging paradigm: "Sepsis" is not one disease but a heterogeneous syndrome
- Genomic markers: Polymorphisms predicting response (e.g., TLR4 variants)
- Transcriptomic phenotypes: SRS1 (hyperinflammatory) vs. SRS2 (immunosuppressed)⁷⁰
- Metabolomic profiling: Predicting need for specific interventions
- Future goal: Theranostic approach—diagnose sepsis subtype, match to targeted therapy
Conclusion
The approach to fever in the hospitalized patient requires synthesis of clinical acumen, evidence-based diagnostics, and rational therapeutics. Key principles include:
- Fever definitions are nuanced; hypothermia carries worse prognosis than fever
- SIRS has been deemphasized but remains a sensitive screening tool; sepsis requires organ dysfunction
- History and physical examination remain the cornerstone of source identification
- Diagnostic stewardship mandates judicious culture ordering—not reflexive pan-culturing
- Biomarkers (especially procalcitonin) guide antibiotic duration and de-escalation
- Empiric antibiotics must be tailored to patient risk factors and local resistance patterns
- 48-72 hour reassessment is mandatory for de-escalation or escalation
- Source control is as important as antibiotics in many infections
- Shorter durations are non-inferior for most infections when source controlled
- Antimicrobial stewardship protects both individual patients and future populations
As critical care practitioners, we must balance the urgency of treating sepsis with the long-term responsibility of preserving antibiotic effectiveness. The judicious diagnostician considers not only "What infection does this patient have?" but equally "Could this be something other than infection?" and "How little can I do to achieve optimal outcomes?"
Mastering the septic workup is not about memorizing algorithms, but about developing clinical reasoning that integrates physiology, microbiology, pharmacology, and epidemiology at the bedside. The art lies in knowing when to escalate and when to de-escalate—and having the wisdom to distinguish between the two.
References
-
Eliakim-Raz N, et al. Fever in hospitalized patients: aetiology, prevalence and outcome. J Hosp Infect. 2014;88(4):188-195.
-
Singer M, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016;315(8):801-810.
-
Mackowiak PA, et al. A critical appraisal of 98.6°F, the upper limit of the normal body temperature, and other legacies of Carl Reinhold August Wunderlich. JAMA. 1992;268(12):1578-1580.
-
Rumbus Z, et al. Fever Is Associated with Reduced, Hypothermia with Increased Mortality in Septic Patients: A Meta-Analysis of Clinical Trials. PLoS One. 2017;12(1):e0170152.
-
Marik PE, et al. Hypothermia and cytokines in septic shock. Norasept II Study Investigators. Intensive Care Med. 2000;26(6):716-721.
-
Churpek MM, et al. Quick Sepsis-related Organ Failure Assessment, Systemic Inflammatory Response Syndrome, and Early Warning Scores for Detecting Clinical Deterioration in Infected Patients outside the ICU. Am J Respir Crit Care Med. 2017;195(7):906-911.
-
Kaukonen KM, et al. Systemic inflammatory response syndrome criteria in defining severe sepsis. N Engl J Med. 2015;372(17):1629-1638.
-
Seymour CW, et al. Assessment of Clinical Criteria for Sepsis: For the Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016;315(8):762-774.
-
Phua J, et al. Characteristics and outcomes of culture-negative versus culture-positive severe sepsis. Crit Care. 2013;17(5):R202.
-
Kaasch AJ, et al. Staphylococcus aureus bloodstream infection: how best to manage? Curr Opin Infect Dis. 2014;27(6):523-529.
-
Lyder CH, et al. Hospital-acquired pressure ulcers: results from the national Medicare Patient Safety Monitoring System study. J Am Geriatr Soc. 2012;60(9):1603-1608.
-
Mermel LA, et al. Clinical practice guidelines for the diagnosis and management of intravascular catheter-related infection: 2009 Update by IDSA. Clin Infect Dis. 2009;49(1):1-45.
-
Rouby JJ, et al. Nosocomial maxillary sinusitis in mechanically ventilated patients. J Infect Dis. 1994;169(6):1261-1264.
-
Morgan DJ, et al. Diagnostic stewardship—leveraging the laboratory to improve antimicrobial use. JAMA. 2017;318(7):607-608.
-
Lee A, et al. Improved blood culture contamination rates after feedback from physicians and education. Am J Clin Pathol. 2007;128(1):130-133.
-
Cockerill FR 3rd, et al. Optimal testing parameters for blood cultures. Clin Infect Dis. 2004;38(12):1724-1730.
-
Bates DW, et al. Contaminant blood cultures and resource utilization. The true consequences of false-positive results. JAMA. 1991;265(3):365-369.
-
Shapiro NI, et al. Who needs a blood culture? A prospectively derived and validated prediction rule. J Emerg Med. 2008;35(3):255-264.
-
Safdar N, et al. Meta-analysis: methods for diagnosing intravascular device-related bloodstream infection. Ann Intern Med. 2005;142(6):451-466.
-
O'Grady NP, et al. Guidelines for the prevention of intravascular catheter-related infections. Am J Infect Control. 2011;39(4 Suppl 1):S1-34.
-
Nicolle LE, et al. Clinical Practice Guideline for the Management of Asymptomatic Bacteriuria: 2019 Update by IDSA. Clin Infect Dis. 2019;68(10):e83-e110.
-
Murray PR, et al. Microscopic and baceriologic analysis of expectorated sputum. Mayo Clin Proc. 1975;50(6):339-344.
-
Canadian Critical Care Trials Group. A randomized trial of diagnostic techniques for ventilator-associated pneumonia. N Engl J Med. 2006;355(25):2619-2630.
-
Parente DM, et al. The Clinical Utility of Methicillin-Resistant Staphylococcus aureus (MRSA) Nasal Screening to Rule Out MRSA Pneumonia: A Diagnostic Meta-analysis With Antimicrobial Stewardship Implications. Clin Infect Dis. 2018;67(1):1-7.
-
McDonald LC, et al. Clinical Practice Guidelines for Clostridium difficile Infection in Adults and Children: 2017 Update by IDSA and SHEA. Clin Infect Dis. 2018;66(7):e1-e48.
-
Attia J, et al. The rational clinical examination. Does this adult patient have acute meningitis? JAMA. 1999;282(2):175-181.
-
Wacker C, et al. Procalcitonin as a diagnostic marker for sepsis: a systematic review and meta-analysis. Lancet Infect Dis. 2013;13(5):426-435.
-
Schuetz P, et al. Procalcitonin to initiate or discontinue antibiotics in acute respiratory tract infections. Cochrane Database Syst Rev. 2017;10:CD007498.
-
Bouadma L, et al. Use of procalcitonin to reduce patients' exposure to antibiotics in intensive care units (PRORATA trial): a multicentre randomised controlled trial. Lancet. 2010;375(9713):463-474.
-
de Jong E, et al. Efficacy and safety of procalcitonin guidance in reducing the duration of antibiotic treatment in critically ill patients: a randomised, controlled, open-label trial. Lancet Infect Dis. 2016;16(7):819-827.
-
Becker KL, et al. Procalcitonin in sepsis and systemic inflammation: a harmful biomarker and a therapeutic target. Br J Pharmacol. 2010;159(2):253-264.
-
Póvoa P. C-reactive protein: a valuable marker of sepsis. Intensive Care Med. 2002;28(3):235-243.
-
Coelho LM, et al. Usefulness of C-reactive protein in monitoring the severe community-acquired pneumonia clinical course. Crit Care. 2007;11(4):R92.
-
Memar MY, et al. Biomarkers and Pathways in Multidrug-Resistant Bacteria. Mol Biol Rep. 2020;47(12):9999-10014.
-
Ulla M, et al. Diagnostic and prognostic value of presepsin in the management of sepsis in the emergency department: a multicenter prospective study. Crit Care. 2013;17(4):R168.
-
Nguyen HB, et al. Early lactate clearance is associated with improved outcome in severe sepsis and septic shock. Crit Care Med. 2004;32(8):1637-1642.
-
Kumar A, et al. Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock. Crit Care Med. 2006;34(6):1589-1596.
-
Pappas PG, et al. Clinical Practice Guideline for the Management of Candidiasis: 2016 Update by IDSA. Clin Infect Dis. 2016;62(4):e1-50.
-
Sanchez GV, et al. Antibiotic Resistance Among Urinary Isolates from Female Outpatients in the United States in 2003 and 2012. Antimicrob Agents Chemother. 2016;60(5):2680-2683.
-
Stevens DL, et al. Practice guidelines for the diagnosis and management of skin and soft tissue infections: 2014 update by IDSA. Clin Infect Dis. 2014;59(2):147-159.
-
Wong CH, et al. The LRINEC (Laboratory Risk Indicator for Necrotizing Fasciitis) score: a tool for distinguishing necrotizing fasciitis from other soft tissue infections. Crit Care Med. 2004;32(7):1535-1541.
-
Andes DR, et al. Impact of treatment strategy on outcomes in patients with candidemia and other forms of invasive candidiasis: a patient-level quantitative review of randomized trials. Clin Infect Dis. 2012;54(8):1110-1122.
-
Beekmann SE, et al. Effects of rapid detection of bloodstream infections on length of hospitalization and hospital charges. J Clin Microbiol. 2003;41(7):3119-3125.
-
Freifeld AG, et al. Clinical practice guideline for the use of antimicrobial agents in neutropenic patients with cancer: 2010 update by IDSA. Clin Infect Dis. 2011;52(4):e56-93.
-
Paul M, et al. Systematic review and meta-analysis of the efficacy of appropriate empiric antibiotic therapy for sepsis. Antimicrob Agents Chemother. 2010;54(11):4851-4863.
-
Rhodes A, et al. Surviving Sepsis Campaign: International Guidelines for Management of Sepsis and Septic Shock: 2016. Intensive Care Med. 2017;43(3):304-377.
-
Rybak MJ, et al. Therapeutic monitoring of vancomycin for serious methicillin-resistant Staphylococcus aureus infections: A revised consensus guideline and review by the ASHP, IDSA, PIDS, and SIDP. Am J Health Syst Pharm. 2020;77(11):835-864.
-
Roberts JA, et al. Individualised antibiotic dosing for patients who are critically ill: challenges and potential solutions. Lancet Infect Dis. 2014;14(6):498-509.
-
Chastre J, et al. Comparison of 8 vs 15 days of antibiotic therapy for ventilator-associated pneumonia in adults: a randomized trial. JAMA. 2003;290(19):2588-2598.
-
Spellberg B. The New Antibiotic Mantra—"Shorter Is Better". JAMA Intern Med. 2016;176(9):1254-1255.
-
Levy MM, et al. The Surviving Sepsis Campaign Bundle: 2018 update. Intensive Care Med. 2018;44(6):925-928.
-
Safdar N, et al. Does the sterility of specimens for culture of coagulase-negative staphylococci decrease their clinical value? A systematic review of the literature. Clin Infect Dis. 2004;39(6):830-838.
-
Anaya DA, et al. Necrotizing soft-tissue infection: diagnosis and management. Clin Infect Dis. 2007;44(5):705-710.
-
Tamma PD, et al. Association of Adverse Events With Antibiotic Use in Hospitalized Patients. JAMA Intern Med. 2017;177(9):1308-1315.
-
Kumar A, et al. Early combination antibiotic therapy yields improved survival compared with monotherapy in septic shock: a propensity-matched analysis. Crit Care Med. 2010;38(9):1773-1785.
-
Singh N, et al. Short-course empiric antibiotic therapy for patients with pulmonary infiltrates in the intensive care unit. A proposed solution for indiscriminate antibiotic prescription. Am J Respir Crit Care Med. 2000;162(2 Pt 1):505-511.
-
Sawyer RG, et al. Trial of Short-Course Antimicrobial Therapy for Intraabdominal Infection. N Engl J Med. 2015;372(21):1996-2005.
-
Yahav D, et al. Seven Versus 14 Days of Antibiotic Therapy for Uncomplicated Gram-negative Bacteremia: A Noninferiority Randomized Controlled Trial. Clin Infect Dis. 2019;69(7):1091-1098.
-
López-Cortés LE, et al. Impact of an evidence-based bundle intervention in the quality-of-care management and outcome of Staphylococcus aureus bacteremia. Clin Infect Dis. 2013;57(9):1225-1233.
-
Lo E, et al. Strategies to prevent catheter-associated urinary tract infections in acute care hospitals: 2014 update. Infect Control Hosp Epidemiol. 2014;35(5):464-479.
-
Liu D, et al. Prognostic value of procalcitonin in adult patients with sepsis: A systematic review and meta-analysis. PLoS One. 2015;10(6):e0129450.
-
Tenner S, et al. American College of Gastroenterology guideline: management of acute pancreatitis. Am J Gastroenterol. 2013;108(9):1400-1415.
-
Patel RA, et al. Drug fever. Pharmacotherapy. 2010;30(1):57-69.
-
Brown KA, et al. Meta-analysis of antibiotics and the risk of community-associated Clostridium difficile infection. Antimicrob Agents Chemother. 2013;57(5):2326-2332.
-
Davey P, et al. Interventions to improve antibiotic prescribing practices for hospital inpatients. Cochrane Database Syst Rev. 2017;2:CD003543.
-
Banerjee R, et al. Randomized Trial of Rapid Multiplex Polymerase Chain Reaction-Based Blood Culture Identification and Susceptibility Testing. Clin Infect Dis. 2015;61(7):1071-1080.
-
Clancy CJ, et al. T2 Magnetic Resonance for the Diagnosis of Bloodstream Infections: A Clinical Review. J Fungi (Basel). 2018;4(1):24.
-
Wilson MR, et al. Clinical Metagenomic Sequencing for Diagnosis of Meningitis and Encephalitis. N Engl J Med. 2019;380(24):2327-2340.
-
van der Poll T, et al. The immunopathology of sepsis and potential therapeutic targets. Nat Rev Immunol. 2017;17(7):407-420.
-
Sweeney TE, et al. A comprehensive time-course-based multicohort analysis of sepsis and sterile inflammation reveals a robust diagnostic gene set. Sci Transl Med. 2015;7(287):287ra71.
Summary: Key Takeaways for the Critical Care Fellow
The "Rule of 3s" for Sepsis Management
First 3 Hours:
- 3 assessments: Clinical (source), Laboratory (lactate), Microbiologic (cultures)
- 3 interventions: Antibiotics, Fluids (30 mL/kg if indicated), Source control planning
- 3 decisions: Empiric regimen, ICU need, Consultation triggers
At 3 Days (72 hours):
- 3 questions: Are cultures finalized? Is patient improving? Can I narrow/stop?
- 3 biomarkers: PCT trend, CRP trend, Lactate clearance
- 3 outcomes: De-escalate, Continue targeted, or Escalate with ID consult
After 3 Weeks (or discharge):
- 3 reviews: Was diagnosis correct? Was duration appropriate? What can I learn?
- 3 stewardship goals: Document indication, Record expected duration, Plan follow-up
- 3 patient safety measures: C. diff prophylaxis consideration, Antibiotic allergy verification, Resistance documentation
The Critical Care Fellow's Pocket Card
FEVER WORKUP CHECKLIST:
□ Temperature confirmed (rectal/core if questionable)
□ SIRS/qSOFA/SOFA scored (document baseline)
□ Detailed exposure history (antibiotics, procedures, travel)
□ Systematic physical exam (don't skip skin, lines, rectal)
□ Targeted cultures ONLY (not reflexive pan-culture)
□ Blood cultures × 2 BEFORE antibiotics (adequate volume)
□ Source control evaluated (imaging if needed)
□ Empiric antibiotics within 1 hour (if sepsis/shock)
□ Consult antibiogram (tailor to local resistance)
□ Baseline PCT/CRP (for future de-escalation)
□ Document plan (indication, expected duration, reassessment date)
□ 48-72h reassessment MANDATORY (calendar reminder)
Red Flags That Demand Immediate Action
🚩 Necrotizing soft tissue infection → Surgery consult NOW (not after imaging)
🚩 Obstructive pyelonephritis → Urology/IR for urgent decompression
🚩 S. aureus bacteremia → Remove lines, TEE, ID consult, minimum 14 days IV
🚩 Candida bloodstream infection → Remove lines, ophthalmology exam, echo, ID consult
🚩 Persistent bacteremia >72h → Think endocarditis, abscess, hardware infection
🚩 Hypothermia + sepsis → High mortality, aggressive resuscitation
🚩 "Improving" on broad-spectrum but cultures negative → Consider stopping antibiotics
Common Cognitive Errors in Sepsis Management
-
Anchoring bias: "The urine culture is positive, so this must be urosepsis" (ignore pneumonia on CXR)
- Remedy: Systematically consider all potential sources
-
Availability heuristic: "Last patient with fever had C. diff" (order C. diff on every fever)
- Remedy: Apply evidence-based testing criteria
-
Premature closure: "Cellulitis diagnosed, start cefazolin" (miss underlying necrotizing fasciitis)
- Remedy: Reassess if clinical trajectory doesn't match expectations
-
Commission bias: "Must do something" (start antibiotics for drug fever at day 10)
- Remedy: Consider stopping antibiotics as an active intervention
-
Sunk cost fallacy: "Already gave 10 days of antibiotics, must complete 14"
- Remedy: Duration should be dynamic based on clinical response
-
Confirmation bias: "PCT is low, must not be infection" (ignore clinical deterioration)
- Remedy: Biomarkers supplement, never replace, clinical judgment
Scenarios for Self-Assessment
Case 1: 72-year-old nursing home resident, altered mental status, T 37.6°C, hypotensive. Urine culture grows E. coli >100K. Start antibiotics?
Answer: Yes—elderly have blunted fever response; hypothermia worse than fever. However, also consider aspiration pneumonia (common with altered MS) and get CXR. Don't anchor on positive urine culture alone in catheterized patient.
Case 2: 45-year-old IVDU, T 39.5°C, new murmur. Blood culture × 1 grows coagulase-negative staph at 18 hours. True infection?
Answer: Uncertain—single bottle CoNS often contaminant (90% if 1 of 2 positive), BUT in IVDU with new murmur, must treat as endocarditis until proven otherwise. Repeat blood cultures, TEE, ID consult. Don't dismiss based on "common contaminant" in high-risk scenario.
Case 3: Post-op day 5 abdominal surgery, fever 38.7°C, no localizing signs. PCT 0.3, CRP 85. Blood/urine cultures sent. Antibiotics?
Answer: Hold antibiotics initially—post-op fever at day 5 could be atelectasis, DVT/PE, drug fever. Low PCT reassuring. Perform focused workup (exam for wound infection, Doppler for DVT, review medications). If cultures negative at 48h and PCT not rising, avoid antibiotics. Don't reflexively treat every fever.
Case 4: Septic shock, blood cultures positive for MRSA at 12h. On vancomycin, improved hemodynamics, afebrile by day 3. Repeat blood cultures at 48h negative. Remove central line?
Answer: Not necessarily—if patient improving and repeat cultures negative, can observe line. However, MRSA bacteremia warrants ID consultation. Need TEE to assess for endocarditis. If any signs of tunnel infection, purulence, or persistent fevers—remove line. Minimum 14 days of therapy regardless.
Case 5: Community-acquired pneumonia, started ceftriaxone + azithromycin. Day 3: afebrile, improved oxygenation, PCT decreased from 2.5 to 0.4. Continue antibiotics?
Answer: Yes, but plan short course—5 days total if remains afebrile and improving. PCT decline >80% supports shorter duration. Can transition to PO when tolerating, consider levofloxacin 750mg daily × 2 more days to complete 5-day course. Document rationale for short course.
Final Thoughts: The Wisdom of Restraint
The most sophisticated sepsis management is not about maximizing interventions, but about precise calibration—knowing when to act aggressively and when to exercise restraint.
The Modern Sepsis Paradox:
- We must start broad-spectrum antibiotics rapidly (every hour counts in septic shock)
- Yet we must also avoid unnecessary antibiotics (every day counts in resistance development)
The resolution lies in dynamic reassessment. The initial choice is often empiric and broad by necessity, but the 48-72 hour evaluation is where clinical excellence is demonstrated. The best intensivists are not those who start the most antibiotics, but those who know when to stop them.
Three Principles of Antibiotic Wisdom:
- Confidence in starting (when sepsis suspected, don't hesitate)
- Humility in continuing (always ask "do I still need this?")
- Courage in stopping (cessation is an active clinical decision, not passive omission)
As you manage each septic patient, remember: You are simultaneously caring for the individual before you and steward of antibiotics for future patients who will depend on these drugs remaining effective. Both responsibilities are sacred.
The septic workup is not a checklist to complete, but a clinical reasoning process to master—one patient, one culture, one decision at a time.
Acknowledgments: The authors thank the numerous critical care fellows, infectious disease specialists, clinical microbiologists, and antimicrobial stewardship teams whose collective wisdom and daily clinical excellence inform this review.
Conflicts of Interest: None declared.
Funding: None.
Suggested Further Reading
-
Sepsis Definitions & Epidemiology:
- Singer M, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016.
- Fleischmann C, et al. Assessment of Global Incidence and Mortality of Hospital-treated Sepsis. Am J Respir Crit Care Med. 2016.
-
Antibiotic Stewardship:
- Barlam TF, et al. Implementing an Antibiotic Stewardship Program: Guidelines by IDSA and SHEA. Clin Infect Dis. 2016.
- Spellberg B. The New Antibiotic Mantra—"Shorter Is Better". JAMA Intern Med. 2016.
-
Biomarkers:
- Schuetz P, et al. Biomarkers to improve diagnostic and prognostic accuracy in systemic infections. Curr Opin Crit Care. 2017.
- Póvoa P, et al. C-reactive protein as a marker of infection in critically ill patients. Clin Microbiol Infect. 2005.
-
Empiric Therapy Guidelines:
- Rhodes A, et al. Surviving Sepsis Campaign: International Guidelines for Management of Sepsis and Septic Shock: 2016.
- IDSA Clinical Practice Guidelines (multiple, source-specific): www.idsociety.org/practice-guideline
-
Diagnostic Stewardship:
- Morgan DJ, et al. Diagnostic Stewardship—Leveraging the Laboratory to Improve Antimicrobial Use. JAMA. 2017.
- Fabre V, et al. A Review of Antibiotic Use in the Intensive Care Unit and Opportunities for Improvement. Curr Infect Dis Rep. 2019.
Online Resources:
- Surviving Sepsis Campaign: www.survivingsepsis.org
- IDSA Guidelines: www.idsociety.org
- CDC Antibiotic Resistance Resources: www.cdc.gov/drugresistance
- Sanford Guide to Antimicrobial Therapy: www.sanfordguide.com (subscription)
This review represents the current state of knowledge as of 2025 and should be supplemented with emerging evidence and local institutional guidelines. Clinical judgment must always be individualized to the specific patient presentation.
