MedCompanion

Hospital-Acquired & Ventilator-Associated Pneumonia

Pneumonia

Source: IDMP View on spectrum chart →
Patient + scenario modifiers
Patient
Clinical scenario / source

Annotations only — chart still shows the full chemistry of each drug.

1. Clinical Syndrome

Pneumonia developing ≥48 h after hospital admission (HAP) or ≥48 h after intubation (VAP) — new or progressive infiltrate with clinical signs (fever, leukocytosis, purulent secretions, worsening oxygenation).

Excludes: community-acquired pneumonia (use CAP — different empirics); aspiration pneumonia in community-dwelling (overlapping — defer to v2 with cross-link); tracheobronchitis without infiltrate.

2. Pathogens

Consider the patient: Length of stay (≥5 days raises MDRO risk), prior IV antibiotics ≤90 days, structural lung disease, immunocompromise, ICU vs ward, prior MDRO colonization.

Consider the case: Local antibiogram (unit-specific MRSA / Pseudomonas / MDR-Acinetobacter prevalence drives empirics), severity (vasopressors, bilateral infiltrates), source samples (BAL > tracheal aspirate for VAP).

3. Empiric Therapy

TierFirst choiceAlternativesDurationComments
Admitted to ward
  • 4.5 g · q6h (extended infusion) · IV · 7 days
  • 2 g · q8h · IV · 7 days
  • 750 mg · daily · IV · 7 days

    Monotherapy reasonable if no MDRO risk and local susceptibility allows.

7 days (IDSA/ATS 2016)HAP without MDRO risk factors and not septic. Single anti-pseudomonal β-lactam.
Admitted to ICU
  • load + AUC-guided · — · IV
  • 4.5 g · q6h (extended infusion) · IV
  • load + AUC-guided · — · IV
  • 2 g · q8h · IV

    Linezolid alternative to vanco if AKI or vanco intolerance.

7 daysVAP, septic, or HAP with MDRO risk factors. MRSA + Pseudomonas covered empirically.
ICU — Pseudomonas risk
  • load + AUC-guided · — · IV
  • 2 g · q8h · IV
  • Tobramycin 5–7 mg/kg · q24h · IV

    Or gentamicin, or ciprofloxacin 400 mg IV q8h. Double GN for empirics, narrow at 48 h.

  • 1 g · q8h (extended infusion) · IV

    If ESBL risk.

7 days; longer for non-fermenter or complicated course≥5 days hospitalization, prior MDR Pseudomonas, structural lung disease, septic shock.

Add coverage if:

MRSA coverage
  • Prior MRSA respiratory isolate
  • IV antibiotics ≤90 days
  • Hospitalization in unit with >20% MRSA prevalence
  • Septic shock

Add:

  • load + AUC-guided · — · IV

    Linezolid 600 mg IV/PO q12h alternative — superior pulmonary penetration, useful in AKI / persistent vanco failure.

MRSA nasal PCR has high NPV in pneumonia — de-escalate vanco/linezolid if negative.

Pseudomonas coverage
  • Prior Pseudomonas respiratory isolate
  • Structural lung disease (bronchiectasis, severe COPD, CF)
  • Septic shock
  • ≥5 days hospitalization
  • IV antibiotics ≤90 days
  • Recent invasive mechanical ventilation

Add:

  • 2 g · q8h · IV
  • Tobramycin 5–7 mg/kg · q24h · IV

    Add for double-coverage in suspected MDR Pseudomonas; narrow at 48 h.

4. Directed Therapy

De-escalate aggressively at 48 h based on cultures:

  • MSSA: cefazolin or nafcillin (link to S. aureus bacteremia syndrome)
  • MRSA: continue vancomycin or linezolid
  • Susceptible Pseudomonas: narrow to single agent (cefepime, pip-tazo, cipro per susceptibility); consider IV-to-PO if cipro susceptible
  • MRSA nasal PCR negative: drop empiric MRSA coverage (high NPV)
  • Cultures negative + clinical improvement: consider 5–7 day total course

5. Monitoring

Resolution: defervescence, oxygenation improvement (PaO2/FiO2), repeat imaging at 48–72 h if not improving. Procalcitonin trend can support de-escalation.

Toxicity: vanco AUC + Cr; aminoglycoside levels + Cr + audiometry; cefepime mental status; pip-tazo + vanco AKI (consider switching β-lactam to cefepime per ACORN findings).

Pearls

ACORN trial (JAMA 2023) — cefepime had less AKI than pip-tazo with similar mortality; weigh in patients on vanco. MRSA nasal PCR negative has ~99% NPV for ruling out MRSA pneumonia → de-escalate. 7 days is enough for most HAP/VAP per IDSA 2016 — even Pseudomonas (no benefit from longer course; more antibiotic-resistance selection).

References