Algorithm for primary closure in sternal wound infection: a single institution 10-year experience.

Objectives: To evaluate a simple treatment algorithm in sternal wound infection (SWI) allowing for primary closure and to describe the different surgical techniques and their associated morbidity and mortality.

Methods: A retrospective analysis of all patients operated on between 1996 and 2004 in a single tertiary care institution. All epidemiological and surgical data were prospectively collected in our database. Univariate and multivariate analysis were used to determine preoperative and perioperative risks factors for 90-day and long-term mortality.

Results: Out of 5905 procedures, 146 sternal wound infections were documented (2.4%). The respective incidence of SWI for CABG, isolated valve, or combined procedures were 2.8%, 1.1%, and 3.2%. Pathogens involved were S. epidermidis (44.5%), S. aureus (31.5%), and gram-negative rods (19.2%). Re-operation was required in 131/146 patients. Mean time to the first re-operation was 17.3+/-12 days. Modalities of treatment consisted of drainage alone (44 patients), rewiring (25 patients), debridement, rewiring and mediastinal lavage (52 patients), and partial/complete sternal resection (10 patients). Additional procedures were required in 49 patients (37.7%). The 90-day mortality for uninfected patients and patients with superficial SWI were 4.4% and 2.8% (p=0.78) whereas for patients with deep SWI, 90-day mortality was 14.5% (DSWI vs others, p<0.0001).

Conclusions: Deep sternal wound infection (DSWI) remains a dreadful complication in contemporary cardiac surgery while risk factors are currently well defined. Using a simple approach of primary closure together with liberal use of vascularized flaps has allowed us to achieve satisfactory short-term outcome in this subset of patients.