JAMA Netw Open. 2026 Apr 1;9(4):e265447. doi: 10.1001/jamanetworkopen.2026.5447.
ABSTRACT
IMPORTANCE: High-flow nasal oxygen therapy (HFNOT) is used for noninvasive respiratory support following cardiac surgery despite uncertainty about its clinical effectiveness or associated costs.
OBJECTIVE: To determine whether prophylactic HFNOT in patients at increased risk of respiratory complications following cardiac surgery has clinical benefits compared with standard oxygen therapy (SOT).
DESIGN, SETTING, AND PARTICIPANTS: This adaptive, parallel group, randomized clinical trial collected and analyzed data from 17 cardiac surgery centers in 3 countries between October 7, 2020, and June 19, 2024. Eligible participants included adults undergoing nonemergent cardiac surgery with any of the following risk factors for pulmonary complications: chronic obstructive pulmonary disease, asthma, lower respiratory tract infection in the last 4 weeks, a body mass index of 35 or greater, or currently or recently smoking for longer than 10 pack-years. Outcome assessors were blinded. A preplanned sample size re-estimation was conducted after 300 participants completed the 90-day follow-up.
INTERVENTION: Participants were randomized at a 1:1 ratio with concealed allocation to HFNOT or SOT administered for at least 16 hours immediately after postoperative extubation.
MAIN OUTCOMES AND MEASURES: The primary effectiveness outcome was days alive and at home (DAH) without increased support compared with baseline in the first 90 days (DAH90). Any day of increased support, including at home, would provide a value of 0 for that day. Secondary outcomes included DAH90 without considering the additional support component.
RESULTS: A total of 1280 patients were recruited (mean [SD] age, 62.9 [10.5] years; 892 [69.7%] men; 640 in each group), of whom 1224 (95.6%) had complete DAH90 data. The primary outcome of median DAH90 was 0 (IQR, 0-79) for the HFNOT group and 0 (IQR, 0-87) for the SOT group (median difference, 0 [95% CI, 0-0]; P = .75). Secondary clinical outcomes, including DAH90 without considering whether additional support was required, were similar between groups.
CONCLUSIONS AND RELEVANCE: In this randomized clinical trial of HFNOT in patients at increased risk of postoperative pulmonary complications after nonemergent cardiac surgery, HFNOT did not improve DAH90 without increased support. These findings do not support the implementation of routine prophylactic HFNOT after cardiac surgery.
TRIAL REGISTRATION: isrctn.org Identifier: ISRCTN14092678.
PMID:41949866 | DOI:10.1001/jamanetworkopen.2026.5447