Tracheostomy Avoidance in Flap Reconstruction of the Upper Aerodigestive Tract is Safe in Selected Patients

Abstract

AbstractObjectiveAssess the impact of tracheostomy on postoperative morbidity and mortality and examine factors that predicted tracheostomy in head and neck flap patients.Study DesignRetrospective cohort January 2017 to December 2021.SettingSingle tertiary center.MethodsAdult patients undergoing head and neck flap reconstruction were included, excluding cutaneous reconstruction, laryngectomy, or with tracheostomy present at surgery. Patients were routinely extubated immediately following surgery. Univariate and multivariable analyses examined associations between tracheostomy and postoperative outcomes and identified predictors of tracheostomy.ResultsWe included 193 patients, 69 (35.8%) with tracheostomy and 124 (64.2%) without. Tracheostomy avoidance was associated with no difference in 30‐day mortality (P = .531) and with decreased likelihood of 30‐day complications (0.019), unplanned return to the operating room (0.021), and discharge to a facility (<0.001) and with 2.2 decrease in length of stay on multivariable analysis (confidence interval [CI]: 0.62‐3.82, P = .007). Four patients (2.1%) required reactive tracheostomy with no resulting mortalities. Significant associations with tracheostomy on multivariable analysis included bilateral neck dissections (odds ratio [OR]: 3.13, CI 1.12‐9.06, P = .030), number of resected mandibular subsites (OR: 2.65, CI: 1.50‐5.67, P = .0023), specimen volume (OR: 1.87, CI: 1.29‐2.71, P = .001), body mass index < 20 (OR: 3.80, CI: 1.24‐11.64, P = .019), mandibulectomy (OR: 0.04, CI: 0.01‐0.22, P < .001), forearm flap (OR: 0.15, CI: 0.05‐0.41, P < .001), oral cavity site (OR: 0.21, CI: 0.06‐0.73, P = .014), and age > 70 (OR: 0.33, CI: 0.14‐0.81, P = .016).ConclusionTracheostomy avoidance is safe in properly selected patients undergoing head and neck flap reconstruction. Multiple factors predicted tracheostomy, which may guide patient selection at other centers.