RESUMO
BACKGROUND: Fentanyl is one of the most frequently administered intraoperative drugs and may increase the risk of postoperative respiratory complications (PRCs). METHODS: We performed a pre-specified analysis of 145 735 adult non-cardiac surgical cases under general anaesthesia. Using multivariable logistic regression, we evaluated the association of intraoperative fentanyl dose and PRCs within 3 days after surgery (defined as reintubation, respiratory failure, pneumonia, pulmonary oedema, or atelectasis). We examined effect modification by patient characteristics, surgical site, and anaesthetics used. RESULTS: PRCs within 3 days after surgery occurred in 18 839 (12.9%) patients. In comparison with high intraoperative fentanyl doses [median: 3.85; inter-quartile range (IQR): 3.42-4.50 µg kg-1, quartile 4 (Q4)], low intraoperative fentanyl dose [median: 0.80, IQR: 0.00-1.14 µg kg-1, quartile 1 (Q1)] was significantly associated with lower odds of PRCs [Q1 vs Q4: 10.9% vs 16.2%; adjusted odds ratio (aOR) 0.79; 95% confidence intervals (CI) 0.75-0.84; P<0.001; adjusted absolute risk difference (aARD) -1.7%]. This effect was augmented by thoracic surgery (P for interaction <0.001; aARD -6.2%), high doses of inhalation anaesthetics (P for interaction=0.016; aARD -2.2%) and neuromuscular blocking agents (NMBAs) (P for interaction=0.001; aARD -3.4%). Exploratory analysis demonstrated that compared with no fentanyl, low-dose fentanyl was associated with lower rates of PRCs (decile 2 vs decile 1: aOR 0.82, CI 0.75-0.89, P<0.001). CONCLUSIONS: Intraoperative low-dose fentanyl (about 60-120 µg for a 70 kg patient) was associated with lower risk of postoperative respiratory complications compared with both no fentanyl and high-dose fentanyl. Beneficial effects of low-dose fentanyl were magnified in specific patient subgroups. CLINICAL TRIAL REGISTRATION: NCT03198208.