Mortality rate attributable to ventilator-associated nosocomial pneumonia in an adult intensive care unit: A prospective case-control study

Objective. To evaluate the mortality rate attributable to nosocomial ventilator-associated pneumonia in an intensive care unit. Design. Prospective, matched, risk-adjusted cohort study. Setting: A 18-bed adult medical-surgical intensive care unit in a 1,100-bed regional and teaching hospital in France. Patients. From January 1, 1996, to April 30, 1999, 135 patients who developed nosocomial pneumonia were matched with 135 control patients without nosocomial pneumonia. Interventions. None. Measurements and Main Results. Nosocomial pneumonia was identified an the basis of results of distal bronchial samples. The matching process was conducted according to the following primary criteria: cause of admission, indication for ventilatory support, immunologic status, cardiac status, probability of death (+/-5%), Glasgow Coma Scale score (+/-2 points), age (+/-7 yrs), and duration of exposure to risk. When possible, case and control patients were matched according to five secondary criteria: respiratory and alcoholism status before admission, diagnosis categories, surgical procedure or not, and gender. The mortality rates were compared between case and control patients by using the Kaplan-Meier estimate and the log-rank test. The influence of nosocomial pneumonia on mortality rate then was tested by adjusting for the secondary criteria and other possible confounding factors by using the Cox proportional-hazards model. The matching process was successful for 1,080 of 1,080 primary criteria. The crude intensive care unit mortality rate was higher in patients with nosocomial pneumonia than in control patients (41 vs. 14%; p < .0001). In actuarial survival analysis, the probability of intensive care unit death was higher in the case patients (odds ratio = 2.7, 95% confidence interval = 1.8-3.1, p = .028). After adjustment, the occurrence of nosocomial pneumonia remained an independent risk factor of death (odds ratio = 2.1, 95% confidence interval = 1.2-3.6, p = .008). Nosocomial pneumonia attributable to multiresistant microorganisms was significantly associated with death (odds ratio = 2.6, 95% confidence interval = 1.1-5.8, p = .02). The length of intensive care unit stay was higher in case than in control patients (31 +/- 19 vs. 26 +/- 17 days, P < .0001). Conclusions. Nosocomial pneumonia is independently associated with death in the intensive care unit. In addition, it increases the length of intensive care unit stay.