A Randomized Controlled Trial of High-Flow Nasal Oxygen (Optiflow) as Part of an Enhanced Recovery Program After Lung Resection Surgery.

ABSTRACT

BACKGROUND: Patients undergoing thoracic surgery are at risk of postoperative pulmonary complications, which are associated with increased morbidity and mortality. High-flow nasal oxygen therapy delivers humidified, warmed positive airway pressure but has not been tested routinely after thoracic surgery. METHODS: We performed a randomized, controlled, blinded study. Patients undergoing elective lung resection were randomly assigned to either high-flow nasal oxygen or standard oxygen therapy. Patients were otherwise treated within an established enhanced recovery program. The primary outcome was the difference between the preoperative and postoperative 6-minute walk test. Secondary outcomes included hospital length of stay, spirometry, and patient-reported outcomes measured using the Postoperative Quality of Recovery Scale. RESULTS: Fifty-nine patients were randomly assigned to either high-flow nasal oxygen (n = 28) or standard oxygen (n = 31) therapy. We found no difference in the 6-minute walk test outcome or spirometry; however, length of hospital stay was significantly lower in the high-flow nasal oxygen group, median 2.5 days (range, 1 to 22), compared with the standard oxygen group, median 4.0 days (range, 2 to 18); geometric mean ratio was 0.68 (95% confidence interval 0.48 to 0.86, p = 0.03). No significant differences in recovery domains were found, but patients in the high-flow nasal oxygen group reported significantly higher satisfaction (p = 0.046). CONCLUSIONS: Prophylactic high-flow nasal oxygen therapy, when incorporated into an enhanced recovery program, did not improve 6-minute walk test results but was associated with reduced length of hospital stay and improved satisfaction after lung resection, compared with standard oxygen. This finding has implications for reduced costs and better service provision, and a multicenter trial powered for length of stay is required.