Management of the infected median sternotomy wound with muscle flaps. The Emory 20-year experience.

OBJECTIVE: The purpose of the study is to define those patient variables that contribute to morbidity and mortality of median sternotomy wound infection and the results of treatment by debridement and closure by muscle flaps.
BACKGROUND: Infection of the median sternotomy wound after open heart surgery is a devastating complication associated with significant mortality. Twenty years ago, these wounds were treated with either open packing or antibiotic irrigation, with a mortality approaching 50% in some series. In 1975, the authors began treating these wounds with radical sternal debridement followed by closure using muscle or omental flaps. The mortality of sternal wound infection has dropped to < 10%.
METHODS: The authors' total experience with 409 patients treated over 20 years is described in relation to flap choices, hospital days after sternal wound closure, and incidence rates of morbidity and mortality. One hundred eighty-six patients treated since January 1988 were studied to determine which patient variables had impact on rates of flap closure complications, recurrent sternal wound infection, or death. Variables included obesity, history of smoking, hypertension, diabetes, poststernotomy septicemia, internal mammary artery harvest, use of intra-aortic balloon pump, and perioperative myocardial infarction and were analyzed using chi square tests. Fisher's exact tests, and multivariable logistic regression analysis.
RESULTS: The mortality rate over 20 years was 8.1% (33/49). Additional procedures for recurrent sternal wound infection were necessary in 5.1% of patients. Thirty-one patients (7.6%) required treatment for hematoma, and 11 patients (2.7%) required hernia repair. Among patients treated since 1988, variables strongly associated with mortality were septicemia (p < 0.00001), perioperative myocardial infarction (p = 0.006), and intra-aortic balloon pump (p = 0.0168). Factors associated with wound closure complications were intra-aortic balloon pump (p = 0.0287), hypertension (p = 0.0335), and history of smoking (p = 0.0741). Factors associated with recurrent infection were history of sternotomy (p = 0.008) and patients treated for sternal wound infection from 1988 to 1992 (p = 0.024). Mean hospital stay after sternal wound reconstruction declined from 18.6 days (1988-1992) to 12.4 days (1993-1996) (p = 0.005). To clarify management decisions of these difficult cases, a classification of sternal wound infection is presented.
CONCLUSIONS: Using the principles of sternal wound debridement and early flap coverage, the authors have achieved a significant reduction in mortality after sternal wound infection and have reduced the mean hospital stay after sternal wound closure of these critically ill patients. Further reductions in mortality will depend on earlier detection of mediastinitis, before onset of septicemia, and ongoing improvements in the critical care of patients with multisystem organ failure.