Trends and Variations in Tracheal Intubation for Acute Respiration Failure in Critically Ill Patients in the United States.

BACKGROUND

Acute respiratory failure (ARF) is a common indication for tracheal intubation in critically ill patients. This study sought to describe temporal trends and hospital variation in tracheal intubation among patients with ARF, identify factors associated with its use, and examine the relationship between tracheal intubation and hospital mortality.

METHODS

We performed an observational cohort study of adult hospitalizations (2016-2019) for ARF using the National Inpatient Sample. We identified critically ill patients with ARF who required tracheal intubation and assessed outcomes, temporal trends in the incidence, and associated patient and hospital characteristics. Multivariable logistic regression was used to examine factors associated with the performance of tracheal intubation and to characterize the association between mortality and intubation status while controlling for type of respiratory failure, demographics, and patient comorbidities.

RESULTS

A total of 2017,349 admissions with a diagnosis of ARF were identified, representing 10,086,744 (95% confidence interval [CI], 9957,967-10,215,520) cases nationwide. Of those, 2153,460 (21.3%) required tracheal intubation. Intubated patients were more likely to be nonwhite men, have a higher comorbidity burden, be insured by Medicaid, and be hospitalized in large urban, teaching hospitals. Although ARF cases increased from 1872,519 in 2016 to 3194,075 in 2019, the proportion of patients undergoing tracheal intubation declined from 23.8% to 19.8% during the same period (P < .0001). Mortality rates were higher in intubated than in nonintubated patients (P < 001). Mortality remained stable in intubated patients (30.6% [95% CI, 30.2-31.1] in 2016, versus 31.0% [95% CI, 30.5-31.4] in 2019) but decreased in nonintubated patients during the same time period (8.1% [95% CI, 7.9-8.3] versus 6.6% [95% CI, 6.5-6.8]). 10.5% of the variability in intubation rates was attributable to the admitting hospital, and sepsis was the patient factor most strongly associated with tracheal intubation.

CONCLUSIONS

In the National Inpatient Sample, tracheal intubation in critically ill patients with ARF has declined over time with no change in mortality. In the same period, mortality rates in nonintubated patients have fallen. Further work is needed to better understand factors affecting the decision to intubate for respiratory failure in critically ill patients.