Association Between Awake Tracheal Intubation and Peri-Intubation Events in Critically Ill Patients: A Single-Center Retrospective Cohort Study

Background Tracheal intubation in critically ill adults often precipitates peri-intubation events including severe cardiovascular instability, severe hypoxemia, and cardiac arrest. Awake tracheal intubation (ATI) may mitigate these risks, yet evidence from intensive care units (ICUs) remains limited. We evaluated whether ATI is associated with fewer peri-intubation events than induction of general anesthesia (GA). Methods We performed a retrospective, single-center cohort study of consecutive adults undergoing ICU tracheal intubation from April 2018 through March 2023. The exposure was ATI versus GA. The primary outcome was a composite peri-intubation event within 30 minutes of procedure start, comprising cardiovascular instability, severe hypoxemia, or cardiac arrest. Secondary outcomes were the composite components and 28-day all-cause mortality. The primary analysis estimated adjusted marginal risks using g-computation with multivariable logistic regression and marginal standardization. Uncertainty was quantified with nonparametric bootstrap. Robustness was examined with modified Poisson regression, 1:1 propensity score matching (targeting ATT), and entropy balancing (targeting ATE). Results Of 553 screened episodes, 342 met criteria. The composite occurred in 172 of 342 patients (50.2%) overall (ATI 9 of 48 [18.8%] vs GA 163 of 294 [55.4%]). After adjustment, the estimated risk of the composite was 0.170 (95% CI, 0.069–0.296) with ATI and 0.558 (0.498–0.617) with GA, yielding an adjusted risk difference of − 0.388 (− 0.508 to − 0.251) and a risk ratio of 0.305 (0.120–0.534). Component analyses suggested lower adjusted risks of cardiovascular instability and severe hypoxemia with ATI. Findings were directionally consistent in propensity score-matched and entropy-balanced analyses. 28-day mortality did not differ significantly (adjusted risk 0.301 with ATI vs 0.181 with GA; risk difference + 0.119, 95% CI − 0.008 to + 0.252). Cardiac arrest was rare (1 event per group). No ATI-attributable safety signals were identified. Conclusions In this single-center ICU cohort, ATI was associated with substantially fewer peri-intubation events than GA, with convergent estimates across complementary analytic approaches. These hypothesis-generating data support considering ATI for physiologically fragile patients and motivate pragmatic trials to define patient selection and implementation strategies.