674

Abstract

Introduction: Tracheostomy (Trach) following cardiac surgery is frequently performed “late” to avoid sternal wound infection (SWI). However, the relationship between timing of Trach and survival remains uncertain in this population. Hypothesis: In postoperative cardiac surgical patients, early Trach is associated with improved survival and hospital and ICU LOS. Methods: Institutional REB approval was obtained for this observational study of all patients undergoing major cardiac surgery and who then had Trach at a single academic centre between September 1997 and October 2010. Early Trach was defined as performed between POD 2-13, and late as between POD 14-30. Primary outcome was all cause in-hospital mortality. Secondary outcomes included hospital and ICU length of stay (LOS), and incidence of SWI. Univariate analysis compared perioperative patient characteristics and Kaplan-Meier and log-rank test compared survival of the two time cohorts. A propensity score based on a priori defined predictor variables was built using logistic regression and patients were matched based on their propensity to receiving early Trach. Cox proportional hazards analysis compared survival of the time cohorts stratifying on each matched pair. Survival time was defined as date of Trach until date of death or hospital discharge. Survivors were censored on the day of discharge. Measure of association was HR with 95% CI. All analyses were performed using SAS 9.2, with statistical significance defined as p<0.05. Results: Of 18845 patients, 187 had early Trach(29.4% mortality) and 188 late Trach(33.0% mortality). Hospital and ICU LOS were significantly less in the early Trach group. There was no statistically significant difference between survival of each group (log-rank p=0.59). This lack of survival difference persisted after Cox proportional hazards analysis stratifying on each of the 109 matched pairs (HR=0.90 [0.48-1.70]). Early Trach was associated with lower risk of SWI (OR 0.48 [0.23-1.0], p=0.05). Conclusions: While early tracheostomy was associated with decreased hospital and ICU LOS, and SWI; it was not associated with mortality benefit following cardiac surgery.