Objectives: Continuous or bilevel positive airway pressure ventilation, called noninvasive ventilation (NIV), is a controversial therapy for acute decompensated heart failure (ADHF). While NIV is considered safe and effective in patients with chronic obstructive pulmonary disease (COPD), clinical trial data that have addressed safety in ADHF patients are limited, with some suggestion of increased mortality. The objective of this study was to assess mortality outcomes associated with NIV and to determine if a failed trial of NIV followed by endotracheal intubation (ETI) (NIV failure) is associated with worse outcomes, compared to immediate ETI.
Methods: This was a retrospective analysis of the Acute Decompensated Heart Failure National Registry (ADHERE), which enrolls patients with treatment for, or with a primary discharge diagnosis of, ADHF. The authors compared characteristics and outcomes in four groups: no ventilation, NIV success, NIV failure, and ETI. One-way analysis of variance or Wilcoxon testing was performed for continuous data, and chi-square tests were used for categorical data. In addition, multivariable logistic regression was used to adjust mortality comparisons for risk factors.
Results: Entry criteria were met by 37,372 patients, of which 2,430 had ventilation assistance. Of the ventilation group, 1,688 (69.5%) were deemed NIV success, 72 (3.0%) were NIV failures, and 670 (27.6%) required ETI. The NIV failure group had the lowest O(2) saturation (SaO(2)) (84 +/- 16%), compared to either NIV success (89.6 +/- 10%) or ETI (88 +/- 13%; p = 0.017). ETI patients were more likely to receive vasoactive medications (p < 0.001) than the NIV success cohort. When comparing NIV failures to ETI, there were no differences in treatment during hospitalization (p > 0.05); other than that the NIV failure group more often received vasodilators (68.1% vs. 54.3%; p = 0.026). In-hospital mortality was 7.9% with NIV, 13.9% with NIV failure, and 15.4% with ETI. After risk adjustment, the mortality odds ratio for NIV failure versus ETI increased to 1.43, although this endpoint was not statistically significant.
Conclusions: In this analysis of ADHF patients receiving NIV to date, patients placed on NIV for ADHF fared better than patients requiring immediate ETI. Patients who failed NIV and required ETI still experienced lower mortality than those initially placed on ETI. Thus, while the ETI group may be more severely ill, starting therapy with NIV instead of immediate ETI will likely not harm the patient. When ETI is required, mortality and length of stay may be adversely affected. Since a successful trial of NIV is associated with improved outcomes in patients with ADHF, application of this therapy may be a reasonable treatment option.