Clinical risk predictors associated with cardiac mortality following vascular surgery in South African patients.

Clinical risk prediction is important in the prognostication of peri-operative cardiac complications and the management of high-risk cardiac patients for major non-cardiac surgery. However, the current pre-operative clinical risk indices have been derived in European and American patients and not validated in South African patients. The purpose of this study was to evaluate the utility of the clinical risk predictors identified in Lee's revised cardiac risk index and in the African arm of the INTERHEART study, in predicting cardiac mortality following vascular surgery in South African patients. A retrospective cohort study was conducted of all patients undergoing elective or urgent vascular surgery at Inkosi Albert Luthuli Central Hospital over a three-year period. All in-hospital deaths were identified and classified into cardiac or non-cardiac deaths by an investigator blinded to the patients' pre-operative clinical risk predicators. A second investigator blinded to the cause of death identified the following clinical risk predictors: history of ischaemic heart disease, congestive cardiac failure and cerebrovascular accident, presence of diabetes, hypertension and obesity (BMI > 30 kg.m(-2)), elevated serum creatinine (> 180 micromol.l(-1)), positive smoking history and ethnicity. The main finding was that a serum creatinine level of greater than 180 micromol.l(-1) and a positive smoking history were significantly associated with cardiac death (p = 0.012, p = 0.012, respectively). Multivariate analyses using a backward stepwise modeling technique found only a serum creatinine of > 180 micromol.l(-1) and a positive smoking history to be significantly associated with cardiac mortality (p = 0.038, 0.035, respectively) with an odds ratio and 95% confidence interval of 3.02 (1.06-8.59) and 3.40 (1.09-10.62), respectively. All other clinical predictors were not significantly different between the two groups. However, based on the sample size of this study, a type 2 or b error may have resulted in the other risk predictors not being identified as important clinical predictors of cardiac mortality. Therefore, until such time as a study of adequate power is conducted, a history of ischaemic heart disease, congestive cardiac failure, diabetes and cerebrovascular accidents should still be considered to be important clinical risk predictors in South African surgical patients. In conclusion, an elevated serum creatinine and a positive history for smoking are important clinical predictors of cardiac mortality in South African patients following elective or urgent vascular surgery.