A look into Lee's score: peri-operative cardiovascular risk assessment in non-cardiac surgeries-usefulness of revised cardiac risk index

Objective The revised cardiac risk index (RCRI/Lee's score) was designed for peri-operative risk assessment before elective major non-cardiac surgeries. Through this article, we report the usefulness of RCRI in our daily practice, while evaluating patients undergoing surgeries of varying risk. Methods Only referred patients, aged ≥ 40 years, were included. Risk stratification was done using RCRI scoring system. Patients were categorised into 4 classes depending on 0, 1, 2, and ≥3 risk predictors (risk predictors were high-risk surgery, history of ischaemic heart disease (IHD), diabetes on insulin, history of stroke (cerebrovascular accident [CVA]), history of congestive heart failure (CHF) and serum creatinine of >2 mg%). Electrocardiograms (ECG) were done in all patients, while troponin I in intermediate and high-risk patients, and in others if symptomatic. Perioperative cardiovascular events were managed appropriately. Results Of the 920 patients included, only 853 patients were analysed as 67 patients were not operated upon. The mean age was 59 ± 11years and 46% of the patients were women. Two hundred and ninety-two underwent high-risk surgeries, 97 patients had history of IHD, 89 had history of CHF, 36 gave history of CVA, 269 patients were diabetics on insulin and 68 had serum creatinine >2 mg%. Number of patients in Lee's classes I, II, III, and IV were 311, 347, 150, and 52, respectively. 26 out of 853 patients had peri-operative events. Of the six variables in RCRI, only history of IHD was an independent predictor of events. Event rates increased as the RCRI class increased, i.e. 1.7%, 2.0%, 6.7%, and 7.7% for classes I–IV, respectively. Age >70 years, poor general medical condition, emergency surgery and left bundle branch block (LBBB) on ECG, were significantly associated with peri-operative events. Conclusion The RCRI is a useful tool in pre-operative risk stratification. It should perhaps be further updated to improve its predictive accuracy.