BACKGROUND AND AIM OF THE STUDY
Valvular endocarditis occurring in patients with dialysis-dependent renal failure increases the risk of mortality following valve surgery, but few specific data are available to guide surgical selection. The study aim was to develop a simple risk scoring system to identify high-risk dialysis patients undergoing valve surgery for endocarditis.
METHODS
Using STS data from between January 1994 and December 2003, a total of 1,862 valvular procedures was recorded in dialysis patients with endocarditis. Isolated mitral procedures were performed in 718 patients, isolated aortic in 656, double valves in 386, isolated tricuspid in 55, and triple valves in 47. Logistic regression analysis was performed relating baseline variables to hospital mortality. Points were assigned to each significant risk factor by rounding regression coefficients to integers. An estimate of risk was obtained for each patient by averaging the predicted mortality among all patients having the same number of total points.
RESULTS
In the logistic regression, significant variables (all p < 0.001) and effect estimates (odds ratios; points) were: cardiogenic shock or salvage status (2.77; 3), double valve (2.37; 3), age > or = 60 years (2.02; 2), isolated mitral valve (1.89; 2), body surface area > 2.1 m2 (1.86; 2), arrhythmia (1.56; 1), active endocarditis (1.54; 1), and female gender (1.53; 1), with a C-statistic of 0.705. Mortality increased exponentially as a function of points, with operative mortality exceeding 70% at 10 points.
CONCLUSION
The risk of mortality for dialysis patients having valve surgery for endocarditis is high, yet depends on a variety of factors. The proposed risk scoring system successfully discriminates between higher and lower risk patients, and could contribute to better decision making.