Effect of chlorhexidine skin disinfection and retrosternal gentamicin sponge on post-sternotomy mediastinitis: results from a prospective controlled registry of 2340 patients

Abstract

Reducing post-sternotomy mediastinitis (PSM) requires incorporating multiple methods. However, the independent effects of these measures are not well studied. To evaluate the independent effect of preoperative disinfection using isopropyl alcohol (IPA)-chlorhexidine gluconate (CHG) and the topical application of a retrosternal gentamicin collagen sponge at wound closure on reducing PSM. From October 2012 to August 2014, 2340 patients were included in this prospective, controlled registry. Patients were divided into four groups. In groups 1 and 2, the skin was disinfected with IPA, and in groups 3 and 4, the skin was disinfected with IPA-CHG. A retrosternal gentamicin collagen sponge was used in groups 2 and 4. Freedom from PSM up to the 30th postoperative day was the primary endpoint. The secondary endpoint was freedom from any surgical site dehiscence. A stepwise regression model was made to reveal the independent factors associated with lower incidence of PSM. There were significant differences in outcome among the groups (P<0.0001). Primary healing was highest in group 4 (91.4%), which showed the lowest rate for mediastinitis (0.9%). Multivariate analysis showed that the use of CHG and a gentamicin sponge was statistically significant (P=0.026 and 0.013, respectively). The other significant independent factors were valve operation (P=0.001), body mass index >30kg/m2 (P=0.001), preoperative stroke (P=0.005), and blood transfusion (P=0.022). Preoperative skin disinfection with IPA-CHG is superior to only IPA, and it should be recommended. The addition of a retrosternal gentamicin-releasing sponge further reduces the rate of mediastinitis.