To assess the effectiveness of noninvasive ventilation in patients with acute respiratory failure and do-not-intubate or comfort-measures-only orders.Data Sources:
MEDLINE, EMBASE, CINAHL, Scopus, and Web of Science from inception to January 1, 2017.Study Selection:
Studies of all design types that enrolled patients in the ICU or hospital ward who received noninvasive ventilation and had preset do-not-intubate or comfort-measures-only orders.Data Extraction:
Data abstraction followed Meta-analysis of Observational Studies in Epidemiology guidelines. Data quality was assessed using a modified Newcastle-Ottawa Scale.Data Synthesis:
Twenty-seven studies evaluating 2,020 patients with do-not-intubate orders and three studies evaluating 200 patients with comfort-measures-only orders were included. In patients with do-not-intubate orders, the pooled survival was 56% (95% CI, 49–64%) at hospital discharge and 32% (95% CI, 21–45%) at 1 year. Hospital survival was 68% for chronic obstructive pulmonary disease, 68% for pulmonary edema, 41% for pneumonia, and 37% for patients with malignancy. Survival was comparable for patients treated in a hospital ward versus an ICU. Quality of life of survivors was not reduced compared with baseline, although few studies evaluated this. No studies evaluated quality of dying in nonsurvivors. In patients with comfort-measures-only orders, a single study showed that noninvasive ventilation was associated with mild reductions in dyspnea and opioid requirements.Conclusions:
A large proportion of patients with do-not-intubate orders who received noninvasive ventilation survived to hospital discharge and at 1 year, with limited data showing no decrease in quality of life in survivors. Provision of noninvasive ventilation in a well-equipped hospital ward may be a viable alternative to the ICU for selected patients. Crucial questions regarding quality of life in survivors, quality of death in nonsurvivors, and the impact of noninvasive ventilation in patients with comfort-measures-only orders remain largely unanswered.