In the post cardiac surgical patient, the traditionally held belief that soiling of the surgical site from a tracheal stoma increases the risk of sternal wound infection (SWI) continues to influence clinical decision-making regarding this intervention. The purpose of this study was to investigate whether tracheostomy was associated with SWI post cardiac surgery. Following institutional REB approval, all patients undergoing cardiac surgery via median sternotomy from September 1, 1997 to October 31, 2010 were included in this retrospective observational study. Patients with preoperative tracheostomy in situ and those patients receiving tracheostomy following documented SWI were excluded. Primary exposure was tracheostomy performed during ICU admission. Primary outcome was SWI. Secondary outcomes were in-hospital mortality and ICU length of stay. Perioperative patient characteristics and outcomes were compared using t-tests and chi-square tests. Variables found to be strongly associated with SWI (p<0.1) were entered into a multivariable stepwise logistic regression model, for the determination of predictors of SWI. Also included the model were variables that were repeated reported to predict SWI in literature. 411 of 18845 included patients had tracheostomy prior to SWI with a median time to tracheostomy of 14 days. The incidences of SWI in the tracheostomy and non-tracheostomy groups were 19.53%(80/411) and 0.84%(154/18434), respectively. On univariate analysis, SWI was associated with age, female gender, higher preoperative risk scores, diabetes, poor LV function, harvest of internal thoracic arteries, resternotomy, cardiopulmonary bypass time, renal replacement therapy, cardiogenic shock, reopen postoperatively, reintubation and mechanical ventilation >72 hours. Tracheostomy was associated with SWI (p<0.0001) and increased mortality (30.4%[125/411] vs. 2.6%[472/18434]; p<0.0001) and ICU length of stay (44±32.3 days vs. 2.4±3.8 days; p<0.0001). On multivariable analysis, tracheostomy was found to be an independent predictor of SWI (OR 2.38). The 1.2% (234/18845) incidence of SWI in our cohort is in keeping with previously reported rates of SWI following cardiac surgery ranging between 0.4 to 8.6%1Ann Thorac Surg. 2005 Aug; 80: 618-621PubMed Google Scholar, 2Intensive Care Med. 2008 Oct; 34: 1779-1787Crossref PubMed Scopus (177) Google Scholar, 3Ann Thorac Surg. 2007 Dec; 84: 1984-1991Abstract Full Text Full Text PDF PubMed Scopus (31) Google Scholar, 4J Thorac Cardiovasc Surg. 1996; 111: 1200-1207Abstract Full Text Full Text PDF PubMed Scopus (267) Google Scholar. However, the relationship between tracheostomy and SWI has been variably reported and remains uncertain1Ann Thorac Surg. 2005 Aug; 80: 618-621PubMed Google Scholar, 2Intensive Care Med. 2008 Oct; 34: 1779-1787Crossref PubMed Scopus (177) Google Scholar, 3Ann Thorac Surg. 2007 Dec; 84: 1984-1991Abstract Full Text Full Text PDF PubMed Scopus (31) Google Scholar, 4J Thorac Cardiovasc Surg. 1996; 111: 1200-1207Abstract Full Text Full Text PDF PubMed Scopus (267) Google Scholar. Rahmanian's multivariable analysis concluded respiratory failure, rather than tracheostomy, was predictive of SWI3Ann Thorac Surg. 2007 Dec; 84: 1984-1991Abstract Full Text Full Text PDF PubMed Scopus (31) Google Scholar. Our multivariable model found tracheostomy to be a strong and independent predictor of SWI (OR=2.38), after adjusting for the same surrogates for respiratory failure used in Rahmanian's paper.
Read full abstract