Mortality risk model for heart valve surgery in China

The study aim was to construct a scoring system for the prediction of early mortality in heart valve surgery patients in China, on the basis of objective risk factors.

Data from 43 Chinese medical centers recorded between January 2007 and December 2008 were analyzed for 13,353 heart valve surgery patients aged > or = 18 years. There were 2,505 cases of aortic valve surgery, 6,996 cases of mitral valve surgery, and 3,852 cases of double-valve surgery (concomitant aortic valve surgery with mitral valve surgery). The EuroSCORE performance was first studied for valve procedures, and a logistic regression then used to examine the relationship between risk factors and in-hospital mortality.

The overall mortality was 1.95%. The EuroSCORE has a low discrimination ability for valve surgery, and sensibly overpredicts risk. The risk index contains a total of 10 risk factors, including age, body mass index (BMI), prior valve operation, serum creatinine level, and NYHA functional class. The mathematical models were highly significant predictors of the outcome and in-hospital mortality, and the results were in general agreement with those reported by others. The risk model exhibited a good predictive ability (Hosmer-Lemeshow test, p = 0.97) and discriminated between high- and low-risk patients reasonably well (area under the receiver operating characteristic curve = 0.76).

Results and methods are presented for use in clinical practice to calculate patient-specific in-hospital mortality after valve surgery, either by applying the logistic equation for each model or by using a simple scoring system with a look-up table for mortality rate.