Objective: Mechanical ventilation is a common therapy used in caring for critically ill patients, but its epidemiology is poorly understood. We describe population-based, temporal trends in the incidence, survival, and hospital bed utilization of mechanically ventilated, noncardiac surgery adult patients.
Design: Retrospective, observational cohort study using linked administrative databases.
Setting: Province of Ontario, Canada.
Patients: Subjects were 150,755 unique patients who received mechanical ventilation between 1992 and 2000.
Interventions: None.
Measurements: Annual measures of mechanical ventilation incidence, 30-day patient mortality rate, and number of mechanical ventilation days and inpatient days for mechanically ventilated patients as a proportion of total adult inpatient bed days.
Main results: From 1992 to 2000, the crude and age- and gender-adjusted incidence of mechanical ventilation increased 9% (p <.001) and 2% (p <.027), respectively, to 217 per 100,000 adults. Crude mortality rate 30 days after initiation of mechanical ventilation increased from 27% to 32% (p <.001). Significant predictors of 30-day mortality rate (adjusted hazard ratio, 95% confidence interval) were calendar year (1.03, 1.02-1.03), age >80 yrs (2.3, 2.2-2.3), Charlson score 3+ (2.0, 2.0-2.1), and specific diagnosis. From 1992 to 2000, the number of mechanical ventilation days and inpatient days for mechanically ventilated patients, as a proportion of total adult inpatient bed days, increased 69% and 30% (both p <.001), respectively, to 1.8% and 6.2%.
Conclusions: There was a small, but important, increase in mechanical ventilation incidence and a substantial increase in the proportion of inpatient bed days used by mechanically ventilated patients in Ontario during the 1990s. These trends are important in planning for expansion of health care resources to meet the needs of the aging population. The increase, over time, in risk-adjusted mortality rate of mechanically ventilated patients is concerning and requires further investigation.