Impact of the ventilator bundle on ventilator-associated pneumonia in intensive care unit

Abstract

Objectives: The ventilator bundle is being promoted to prevent adverse events in ventilated patients including ventilator-associated pneumonia (VAP). We aimed to: (i) examine adoption of the ventilator bundle elements; (ii) determine effectiveness of individual elements and setting characteristics in reducing VAP; (iii) determine effectiveness of two infection-specific elements on reducing VAP; and, (iv) assess crossover effects of complying with VAP elements on central line-associated bloodstream infections. Design: Cross-sectional survey. Setting: Four hundred and fifteen ICUs from 250 US hospitals. Participants: Managers/directors of infection prevention and control departments. Interventions: Adoption and compliance with ventilator bundle elements. Main Outcome Measures: VAP rates. Results: The mean VAP rate was 2.7/1000 ventilator days. Two-thirds (n = 284) reported presence of the full ventilator bundle policy. However, only 66% (n = 188/284) monitored implementation; of those, 39% (n = 73/188) reported high compliance. Only when an intensive care unit (ICU) had a policy, monitored compliance and achieved high compliance were VAP rates lower. Compliance with individual elements or just one of two infection-related element had no impact on VAP (β= -0.79, P= 0.15). There was an association between complying with two infection elements and lower rates (β = -1.81, P< 0.01). There were no crossover effects. Presence of a full-time hospital epidemiologist (HE) was significantly associated with lower VAP rates (β = -3.62, P< 0.01). Conclusions: The ventilator bundle was frequently present but not well implemented. Individual elements did not appear effective; strict compliance with infection elements was needed. Efforts to prevent VAP may be successful in settings of high levels of compliance with all infection-specific elements and in settings with full-time HEs. © The Author 2011. Published by Oxford University Press in association with the International Society for Quality in Health Care; all rights reserved.