Intrahospital transport of critically ill ventilated patients: a risk factor for ventilator-associated pneumonia--a matched cohort study.

Objective: To evaluate the impact of intrahospital transport of critically ill ventilated patients on the acquisition of ventilator-associated pneumonia.

Design: An exposed/unexposed matched cohort study.

Setting: An 18-bed adult medical-surgical intensive care unit in a 1,100-bed regional and teaching hospital in France.

Patients: From January 1, 2001, to December 31, 2002, 118 of 228 ventilated patients transported out of the intensive care unit (exposed patients) were matched with 118 unexposed patients selected among 295 ventilated patients who did not undergo intrahospital transport.

Interventions: None.

Measurements And Main Results: The matching process was conducted according to six criteria: duration of mechanical ventilation, duration of antibiotherapy, indication for ventilatory support, age, probability of death, and surgical procedures or not during intensive care unit stay. The rates of ventilator-associated pneumonia (as defined by usual clinical and biological criteria plus positive culture of bronchoscopy directed catheter) acquisition between exposed and unexposed patients were compared by univariate analysis and then by multivariate analysis (conditional logistic regression and Cox's proportional-hazards model) to account for potential confounding factors. The ventilator-associated pneumonia rate was 26% in exposed patients compared with 10% in the matched unexposed patients. Using conditional logistic regression, two factors were independently associated with ventilator-associated pneumonia: intrahospital transport (odds ratio, 3.1; 95% confidence interval, 1.4-6.7) and the need for reintubation. Using Cox's model, three independent risk factors were identified: the need for reintubation, enteral nutrition, and intrahospital transport (odds ratio, 2.9; 95% confidence interval, 1.4-5.7). The intensive care unit mortality rate was similar (p > .1) in exposed (35%) and unexposed patients (26%).

Conclusions: Intrahospital transport appears to be a significant risk factor for ventilator-associated pneumonia. However, the respective roles of intrahospital transport and of the cause that leads clinicians to transport patients (mainly for radiographic examinations) are difficult to dissociate even after multiple statistical adjustments. When intrahospital transport is needed, very cautious measures must be taken before and during intrahospital transport to prevent ventilator-associated pneumonia. In addition, in the few days after intrahospital transport, intensive search for ventilator-associated pneumonia is justified.