Supplementary Material for: Endotracheal Intubation and In-Hospital Mortality after Intracerebral Hemorrhage

Background: Many patients with acute intracerebral hemorrhages (ICHs) undergo endotracheal intubation with subsequent mechanical ventilation (MV) for “airway protection” with the intent to prevent aspiration, pneumonias, and its related mortality. Conversely, these procedures may independently promote pneumonia, laryngeal trauma, dysphagia, and adversely affect patient outcomes. The net benefit of intubation and MV in this patient cohort has not been systematically investigated. Methods: We conducted a large single-center observational cohort study to examine the independent association between endotracheal intubation and MV, hospital-acquired pneumonia (HAP), and in-hospital mortality (HM) in patients with ICH. All consecutive patients admitted with a primary diagnosis of a spontaneous ICH to a tertiary care hospital in Boston, Massachusetts, from June 2000 through January 2014, who were ≥18 years of age and hospitalized for ≥2 days were eligible for inclusion. Patients with pneumonia on admission, or those having brain or lung neoplasms were excluded. Our exposure of interest was endotracheal intubation and MV during hospitalization; our primary outcomes were incidence of HAP and HM, ascertained using International Classification of Diseases-9 and administrative discharge disposition codes, respectively, in patients who underwent endotracheal intubation and MV versus those who did not. Multivariable logistic regression was used to control for confounders. Results: Of the 2,386 hospital admissions screened, 1,384 patients fulfilled study criteria and were included in the final analysis. A total of 507 (36.6%) patients were intubated. Overall 133 (26.23%) patients in the intubated group developed HAP versus 41 (4.67%) patients in the non-intubated group (p < 0.0001); 195 (38.5%) intubated patients died during hospitalization compared to 48 (5.5%) non-intubated patients (p < 0.0001). After confounder adjustments, OR for HAP and HM, were 4.23 (95% CI 2.48–7.22; p < 0.0001) and 4.32 (95% CI 2.5–7.49; p < 0.0001) with c-statistics of 0.79 and 0.89, in the intubated versus non-intubated patients, respectively. Conclusion: In this large hospital-based cohort of patients presenting with an acute spontaneous ICH, endotracheal intubation and MV were associated with increased odds of HAP and HM. These findings urge further examination of the practice of intubation in prospective studies.