1. Clinical Assessment of Infection in the ICU

Early recognition of sepsis relies on a constellation of findings, including vital sign derangements, evidence of end-organ dysfunction, and biomarkers that reflect systemic inflammation. Prompt evaluation is critical to initiating timely and appropriate antimicrobial therapy.

A. Triggers for Infection Evaluation

• Vital Signs: Tachycardia (> 90 bpm), fever (> 38.3 °C) or hypothermia (< 36 °C), and hypotension (SBP < 90 mmHg or MAP < 65 mmHg) are classic signs. A rising serum lactate (> 2 mmol/L) is a crucial early marker of tissue hypoperfusion.
• Organ Dysfunction: Clinicians must actively track signs of organ injury, including declining urine output, altered mental status, falling platelet counts, and rising bilirubin or creatinine levels.

B. Key Inflammatory Biomarkers

Biomarkers provide objective data to support a diagnosis of bacterial infection and monitor the response to therapy.

2. Interpretation of Laboratory and Microbiological Data

Accurate cultures and susceptibility data are the backbone of targeted antimicrobial therapy. Best practices in specimen collection, combined with an understanding of MIC breakpoints and the strategic use of rapid diagnostics, are essential for effective stewardship.

A. Specimen Collection and Susceptibility Testing

• Specimen Integrity: Obtain at least two sets of blood cultures (one aerobic, one anaerobic bottle per set) from separate venipuncture sites before administering antibiotics. Meticulous skin antisepsis with chlorhexidine is crucial to prevent contamination. For suspected ventilator-associated pneumonia, quantitative bronchoalveolar lavage is preferred.
• Susceptibility Interpretation: Minimum Inhibitory Concentrations (MICs) are interpreted using CLSI or EUCAST breakpoints to classify isolates as Susceptible, Intermediate/Susceptible Dose-Dependent (SDD), or Resistant. This guides the selection of an effective agent.
• Pharmacodynamic Targets: For time-dependent agents like β-lactams, the goal is to maintain the free drug concentration above the MIC for at least 50–70% of the dosing interval (ƒT>MIC). For concentration-dependent agents like aminoglycosides, the goal is a high peak-to-MIC ratio (Cₘₐₓ/MIC > 8–10).

3. Initial Therapeutic Drug Monitoring (TDM) to Guide Therapy

TDM is essential for optimizing efficacy and minimizing toxicity for drugs with narrow therapeutic windows, particularly given the highly variable pharmacokinetics observed in critically ill patients.

4. Severity Scoring Systems for Antimicrobial Decision-Making

Severity of illness scores like the Sequential Organ Failure Assessment (SOFA) and APACHE II are used to quantify physiologic derangement and predict mortality risk. This data helps guide the urgency and breadth of initial empiric antimicrobial coverage.

A. SOFA and APACHE II Scores

• SOFA Score (0–24 points): Assesses dysfunction across six organ systems. A change in SOFA score of ≥2 points from baseline is a key component of the Sepsis-3 definition and correlates with an in-hospital mortality of ≥10%. Higher scores warrant broader empiric coverage.
• APACHE II Score (0–71 points): A more complex score incorporating 12 acute physiologic variables, age, and chronic health status. A score >25 predicts a mortality rate over 50% and helps inform discussions about goals of care and the intensity of interventions.

B. Limitations and Context

Scoring systems are tools, not replacements for clinical judgment. Clinicians must consider patient-specific factors:

• Special Populations: Immunocompromised or neutropenic patients may fail to mount a fever or leukocytosis. Patients on ECMO have altered hemodynamics that can make the cardiovascular SOFA score difficult to interpret.
• Pharmacokinetic Variability: Conditions common in the ICU, such as obesity and augmented renal clearance (ARC), require individualized dosing that goes beyond standard protocols.