POSTSURGICAL MEDIASTINITIS - A CASE-CONTROL STUDY

We report the results of a case-control study of postsurgical mediasti nitis (PSM) that we conducted from 1985 to 1993. The incidence of PSM was 2.2% (81 of 3,711 cases who underwent sternotomy); we analyzed the findings for 73 cases and 73 controls. Univariate analysis revealed t hat the risk factors for PSM were emergency surgery (27% of cases vs. 13% of controls), New York Heart Association functional class IV (46.5 % vs. 21.9%), heart transplantation (12% vs. 0), and coronary artery b ypass graft (CABG) surgery (60% vs. 41%). The incidences of fever, reo peration for bleeding, pacemaker placement, use of vasoactive drugs, p rolonged mechanical ventilation, use of central lines, and treatment i n the intensive care unit were also higher for cases. Multivariate ana lysis identified the following independent risk factors for PSM: reope ration (risk ratio [RR], 9.2), need for vasoactive drugs (RR, 3.5), CA BG surgery (RR, 3.2), and fever that persisted after the third postsur gical day (RR, 406). The related mortality was 13.7%, and death was si gnificantly more frequent among cases (17.7%) than among controls (2.7 %). Multivariate analysis identified the following independent risk fa ctors for mortality: bacteremia (RR, 21.5), the use of an intraaortic balloon (RR, 14.9), advanced age (RR, 1.14 per year), and prolonged me chanical ventilation (RR, 1.1 per day).