Early- and late-onset ventilator-associated pneumonia acquired in the intensive care unit: comparison of risk factors

J Crit Care. 2008 Mar;23(1):27-33. doi: 10.1016/j.jcrc.2007.08.005. Epub 2007 Dec 11.

Abstract

Purpose: To compare risk factors of early- (E) and late-onset (L) ventilator-associated pneumonia (VAP).

Materials and methods: An epidemiological survey based on a nosocomial infection surveillance program of 11 intensive care units (ICUs) of university teaching hospitals in Lyon, France, was conducted. A total of 7236 consecutive ventilated patients, older than 18 years and hospitalized in ICUs for at least 48 hours, were studied between 1996 and 2002. Data during ICU stay, patient-dependent risk factors, device exposure, nosocomial infections occurrence, and outcome were collected. The cutoff point definition between E-VAP (<or=6 days) and L-VAP (>six days) was based on the daily hazard rate of VAP.

Results: The VAP incidence rate was 13.1%, 356 (37.6%) E-VAP (within 6 days of admission) and 590 (62.4%) L-VAP were reported. Independent risk factor for E-VAP vs L-VAP was surgical diagnostic category (odds ratio [OR], 1.49 [95% confidence interval, 1.07-2.07]), whereas independent risk factors for L-VAP vs E-VAP were older age (OR, 1.01 [1.01-1.02]), high Simplified Acute Physiology Score II (OR, 1.01 [1.00-1.02]), infection on admission (OR=2.22 [1.61-3.03]), another nosocomial infection before VAP (OR, 5.88 [3.33-11.11]), and exposure to central venous catheter before VAP (OR, 4.76 [1.04-20.00]).

Conclusions: E-VAP and L-VAP have different risk factors, highlighting the need for developing specific preventive measures.

Publication types

  • Comparative Study

MeSH terms

  • Adult
  • Aged
  • Anti-Bacterial Agents / therapeutic use
  • Chi-Square Distribution
  • Female
  • France / epidemiology
  • Humans
  • Incidence
  • Intensive Care Units*
  • Male
  • Middle Aged
  • Pneumonia, Ventilator-Associated / drug therapy
  • Pneumonia, Ventilator-Associated / epidemiology
  • Pneumonia, Ventilator-Associated / etiology*
  • Population Surveillance
  • Prospective Studies
  • Risk Factors
  • Statistics, Nonparametric

Substances

  • Anti-Bacterial Agents