A Comparison of the Mortality Risk Associated With Ventilator-Acquired Bacterial Pneumonia and Nonventilator ICU-Acquired Bacterial Pneumonia

Crit Care Med. 2019 Mar;47(3):345-352. doi: 10.1097/CCM.0000000000003553.


Objectives: To investigate the respective impact of ventilator-associated pneumonia and ICU-hospital-acquired pneumonia on the 30-day mortality of ICU patients.

Design: Longitudinal prospective studies.

Setting: French ICUs.

Patients: Patients at risk of ventilator-associated pneumonia and ICU-hospital-acquired pneumonia.

Interventions: The first three episodes of ventilator-associated pneumonia or ICU-hospital-acquired pneumonia were handled as time-dependent covariates in Cox models. We adjusted using the case-mix, illness severity, Simplified Acute Physiology Score II score at admission, and procedures and therapeutics used during the first 48 hours before the risk period. Baseline characteristics of patients with regard to the adequacy of antibiotic treatment were analyzed, as well as the Sequential Organ Failure Assessment score variation in the 2 days before the occurrence of ventilator-associated pneumonia or ICU-hospital-acquired pneumonia. Mortality was also analyzed for Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter species(ESKAPE) and P. aeruginosa pathogens.

Measurements and main results: Of 14,212 patients who were admitted to the ICUs and who stayed for more than 48 hours, 7,735 were at risk of ventilator-associated pneumonia and 9,747 were at risk of ICU-hospital-acquired pneumonia. Ventilator-associated pneumonia and ICU-hospital-acquired pneumonia occurred in 1,161 at-risk patients (15%) and 176 at-risk patients (2%), respectively. When adjusted on prognostic variables, ventilator-associated pneumonia (hazard ratio, 1.38 (1.24-1.52); p < 0.0001) and even more ICU-hospital-acquired pneumonia (hazard ratio, 1.82 [1.35-2.45]; p < 0.0001) were associated with increased 30-day mortality. The early antibiotic therapy adequacy was not associated with an improved prognosis, particularly for ICU-hospital-acquired pneumonia. The impact was similar for ventilator-associated pneumonia and ICU-hospital-acquired pneumonia mortality due to P. aeruginosa and the ESKAPE group.

Conclusions: In a large cohort of patients, we found that both ICU-hospital-acquired pneumonia and ventilator-associated pneumonia were associated with an 82% and a 38% increase in the risk of 30-day mortality, respectively. This study emphasized the importance of preventing ICU-hospital-acquired pneumonia in nonventilated patients.

Publication types

  • Comparative Study
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Aged
  • Cross Infection / mortality*
  • Female
  • France / epidemiology
  • Hospital Mortality
  • Humans
  • Intensive Care Units / statistics & numerical data*
  • Male
  • Middle Aged
  • Organ Dysfunction Scores
  • Pneumonia, Bacterial / microbiology
  • Pneumonia, Bacterial / mortality*
  • Pneumonia, Ventilator-Associated / microbiology
  • Pneumonia, Ventilator-Associated / mortality*
  • Proportional Hazards Models
  • Prospective Studies
  • Risk Factors
  • Simplified Acute Physiology Score