Fellow's quick read · Infectious Diseases
Hospital-acquired & ventilator-associated pneumonia
IDSA/ATS 2016 (Kalil et al, CID 2016;63:e61–111) · reviewed 2026-06-25
Personal study digest for a new ID fellow. Recommendations are from the IDSA/ATS 2016 HAP/VAP guideline. The “What’s changed since 2016” section is reviewer synthesis of newer evidence, each claim cited. Not a substitute for the full guideline.
In one line
HAP (no ventilator) and VAP are two distinct entities and HCAP is abolished. Every empiric regimen covers S. aureus + P. aeruginosa + other gram-negatives; add MRSA coverage and a second anti-pseudomonal agent only when risk factors or the local antibiogram demand it. Treat 7 days, then de-escalate.◆When to suspect / diagnose
- VAP = pneumonia >48 h after intubation. HAP = pneumonia ≥48 h after admission, not ventilator-associated. The 2016 guideline treats them as two separate groups.
- HCAP is gone. The category was removed — drive MRSA/Pseudomonas coverage off MDR risk factors, not prior healthcare contact.
- Diagnosis is clinical (new/progressive infiltrate + fever, leukocytosis, purulent secretions, worsening oxygenation). Use clinical criteria ALONE to START antibiotics — not serum PCT (strong rec against), CRP, sTREM-1, or modified CPIS.
- VAP micro: noninvasive sampling (endotracheal aspirate) with semiquantitative culture is preferred over bronchoscopic/quantitative sampling (weak). Suspected non-VAP HAP: treat per noninvasive respiratory + blood cultures rather than empirically when feasible.
◆Workup the fellow drives
- Cultures before antibiotics, but don’t delay therapy: a respiratory sample (endotracheal aspirate for VAP; sputum for HAP) plus blood cultures.
- Pull the unit-specific antibiogram — it sets the empiric choices and the MRSA / double-gram-negative thresholds (strong rec that every hospital generate one, ideally ICU-specific).
- Ventilator-associated tracheobronchitis (VAT) — signs of airway infection without a new infiltrate: do NOT give antibiotics (weak).
◆Empiric therapy — build the regimen
| Component | Include when… | Agents (representative doses) |
|---|---|---|
| Anti-pseudomonal β-lactam (always) | Every empiric regimen — covers P. aeruginosa + GNB | Pip-tazo 4.5 g IV q6h · cefepime/ceftazidime 2 g q8h · meropenem 1 g q8h · imipenem 500 mg q6h · aztreonam 2 g q8h |
| MRSA agent | VAP: IV abx ≤90 d, unit MRSA >10–20%, or prevalence unknown. HAP: those + high mortality risk (vent support for HAP or septic shock); threshold >20% | Vancomycin 15 mg/kg IV q8–12h (now AUC-guided) or linezolid 600 mg IV q12h |
| 2nd anti-pseudomonal (different class) | VAP: MDR risk factor, unit >10% resistant to the chosen agent, or no local data. HAP: high mortality risk or IV abx ≤90 d. Avoid two β-lactams; don’t use an aminoglycoside as the sole agent. | Cipro 400 mg q8h / levofloxacin 750 mg daily; or amikacin 15–20 mg/kg / gentamicin or tobramycin 5–7 mg/kg daily; or colistin / polymyxin B |
- The shared MDR red flag is prior IV antibiotics within 90 days — it independently triggers MRSA, MDR-Pseudomonas, and MDR-HAP coverage. VAP-specific MDR factors (Table 2): septic shock at VAP onset, ARDS preceding VAP, ≥5 days hospitalization, acute renal-replacement therapy before VAP.
- Structural lung disease (bronchiectasis, cystic fibrosis) → use two anti-pseudomonal agents regardless.
- If MRSA coverage is omitted, the regimen must still cover MSSA (pip-tazo, cefepime, levofloxacin, imipenem, or meropenem all do).
◆Definitive (culture-directed) therapy
- MRSA HAP/VAP: vancomycin or linezolid (strong, moderate) — no other agent preferred.
- P. aeruginosa: choose by susceptibility. Monotherapy with one active agent if not in septic shock / low death risk; reserve combination (two active agents) for patients still in septic shock or at high death risk. Never aminoglycoside monotherapy. Drop back to a single agent once shock resolves.
- Acinetobacter: carbapenem or ampicillin-sulbactam if susceptible; if only polymyxin-susceptible, IV colistin/polymyxin B; do not add rifampicin; avoid tigecycline.
- Dose by PK/PD (extended/continuous β-lactam infusions, weight-based, levels) rather than the package insert (weak).
◆Duration & stopping
- 7 days for both VAP (strong, moderate) and HAP (strong, very-low), regardless of pathogen — a deliberate 2016 simplification from the old 8–15 d. Lengthen/shorten only for the rate of clinical, radiologic, and laboratory improvement.
- De-escalate the empiric regimen (narrow the agent / combination → monotherapy) rather than holding fixed (weak).
- PCT + clinical criteria may guide STOPPING (weak) — note benefit is unproven when courses are already ≤7 days. Do NOT use CPIS to decide discontinuation.
◆Key decisions a fellow owns
- Start every empiric call from the unit antibiogram; it sets the MRSA and single-vs-double gram-negative thresholds.
- At the bedside, decide MRSA coverage and one-vs-two anti-pseudomonals from Table 2 risk factors + local rates — not reflex “vanc + 2 anti-pseudomonals” for everyone.
- De-escalate at 48–72 h on cultures; for confirmed Pseudomonas, narrow to a single susceptible agent once out of shock.
- Stop at 7 days in responders — set the stop date when you start.
What’s changed since 2016
Reviewer synthesis of newer evidence — not the guideline. Each claim cited; flags where the 2016 IDSA/ATS guideline lags current practice.
- Novel β-lactam agents now carry resistant gram-negative HAP/VAP — polymyxins demoted. Three phase-3 nosocomial-pneumonia RCTs, all noninferior in ventilated patients: high-dose ceftolozane-tazobactam 3 g (ASPECT-NP, 28-d mortality 24.0% vs 25.3%), cefiderocol (APEKS-NP, day-14 mortality 12.4% vs 11.6%), and imipenem-relebactam (RESTORE-IMI 2, 28-d mortality 15.9% vs 21.3%). The IDSA 2024 AMR gram-negative guidance (not the HAP/VAP guideline) is now the operative source for ESBL-E, CRE, DTR-Pseudomonas, CRAB, and Stenotrophomonas, preferring these β-lactam/BLIs over the colistin-centric pathogen sections of 2016. ASPECT-NP, Lancet Infect Dis 2019;19:1299–1311 (PMID 31563344) · APEKS-NP, Lancet Infect Dis 2021;21:213–225 (PMID 33058798) · RESTORE-IMI 2, CID 2021;73:e4539–e4548 (PMID 32785589) · Tamma, CID 2024 (PMID 39108079)
- Duration — 7 days holds, individualized courses can go shorter, but Pseudomonas is the caveat. REGARD-VAP: individualized short course (median 6 d, as short as 3–5 d) noninferior to usual care (median 14 d) for 60-d death/recurrence (41% vs 44%), with far fewer antibiotic side-effects (8% vs 38%). BUT iDIAPASON: for P. aeruginosa VAP, 8 d FAILED to prove noninferiority vs 15 d (composite 35% vs 26%; recurrence 17% vs 9%; underpowered, stopped early). REGARD-VAP, Lancet Respir Med 2024;12:399–408 (PMID 38272050) · iDIAPASON, Intensive Care Med 2022;48:841–849 (PMID 35552788)
- Vancomycin is AUC-guided now, not trough 15–20. The 2020 revised ASHP/IDSA/PIDS/SIDP consensus targets AUC/MIC 400–600 to cut nephrotoxicity; the 2016 table’s trough-15–20 goal is outdated. Rybak, Am J Health Syst Pharm 2020;77:835–864 (PMID 32191793)
- Procalcitonin stance unchanged. Clinical criteria alone to START; PCT + clinical to help STOP — both 2016 positions remain current.
- No newer IDSA/ATS HAP/VAP guideline — 2016 is still operative. The peer-body ERS/ESICM/ESCMID/ALAT European guideline appeared ~1 year later; the 2022 SHEA/IDSA compendium update covers prevention, not treatment. Torres, Eur Respir J 2017;50:1700582 (PMID 28890434)
- Prevention advance (newly addressed): 3 days of inhaled amikacin in patients ventilated ≥72 h reduced VAP at 28 days (15% vs 22%). This is prevention, not treatment. AMIKINHAL, NEJM 2023;389:2052–2062 (PMID 37888914)
◆Anki cards minted this run
- 7-day duration for both HAP and VAP, regardless of pathogen.
- Procalcitonin — STOP not START (clinical criteria alone to initiate).
- Pseudomonas VAP — the duration exception (iDIAPASON, 8 d not noninferior to 15 d).
- High-dose ceftolozane-tazobactam 3 g for HAP/VAP (ASPECT-NP).
- Held (cap / already in deck): empiric MRSA agents, antibiogram-based empiric therapy, double-anti-pseudomonal triggers, HCAP abolished, and the CID 2026 newer-vs-generic agents meta-analysis are already carded; REGARD-VAP short-course held under the 4-card cap.
Sources: IDSA/ATS 2016 HAP/VAP guideline — Kalil, CID 2016;63:e61–111 (doi:10.1093/cid/ciw353); ASPECT-NP (PMID 31563344); APEKS-NP (PMID 33058798); RESTORE-IMI 2 (PMID 32785589); REGARD-VAP (PMID 38272050); iDIAPASON (PMID 35552788); IDSA 2024 AMR gram-negative guidance — Tamma, CID 2024 (PMID 39108079); vancomycin AUC consensus — Rybak, AJHP 2020;77:835–864 (PMID 32191793); European HAP/VAP guideline — Torres, Eur Respir J 2017;50:1700582 (PMID 28890434); AMIKINHAL (PMID 37888914).