Use the Revised Cardiac Risk Index (RCRI/Lee Index) for cardiac risk assessment in most perioperative settings, while SORT is better suited for predicting overall 30-day mortality risk.
Primary Recommendation
The RCRI (Modified Cardiac Lee Index) should be your primary tool for assessing perioperative cardiac complications (MI, cardiac arrest, pulmonary edema, complete heart block), as it is specifically recommended by major cardiovascular guidelines 1. The 2009 and 2014 ESC/ESA guidelines give it a Class I, Level A/B recommendation for perioperative cardiac risk stratification 1.
When to Use Each Tool
Use RCRI (Lee Index) when:
- Your primary concern is cardiac complications (myocardial infarction, cardiac arrest, pulmonary edema, ventricular fibrillation, complete heart block) 2
- You need guideline-endorsed risk stratification for cardiovascular management decisions 1, 3
- You're determining need for further cardiac testing or perioperative beta-blockers
- You want a simple, validated 6-variable tool that takes seconds to calculate 3
The RCRI includes: history of ischemic heart disease, heart failure, cerebrovascular disease, insulin-dependent diabetes, creatinine >177 μmol/L (>2 mg/dL), and high-risk surgery 3.
Performance caveat: The RCRI discriminates moderately well for mixed non-cardiac surgery (AUC 0.75) but performs worse in vascular surgery patients (AUC 0.64) and poorly predicts death (median AUC 0.62) 2, 4.
Use SORT when:
- Your primary outcome of interest is 30-day mortality rather than cardiac complications specifically 5, 6
- You need better overall mortality prediction (SORT shows superior discrimination with AUC 0.91 vs RCRI's 0.62 for death) 5
- You want to predict postoperative morbidity, not just cardiac events 7
- You're counseling patients about overall surgical risk and mortality
SORT uses 6 variables: ASA Physical Status grade, urgency of surgery, surgical specialty risk, surgical severity, presence of cancer, and age ≥65 years 5.
Complementary Use Strategy
The 2014 ESC/ESA guidelines explicitly state that newer tools like NSQIP (similar conceptually to SORT) and the Lee index "provide complementary prognostic perspectives" 2. This suggests using both when comprehensive risk assessment is needed:
- Calculate RCRI first - it's guideline-mandated for cardiac risk 1
- Add SORT when you need mortality prediction or when RCRI shows limitations (vascular surgery, predicting death)
- For shared decision-making, SORT may provide clearer mortality communication to patients 3, 5
Critical Limitations to Avoid
RCRI weaknesses:
- Suboptimal for patients with multiple risk factors 1
- Poor discrimination in vascular surgery 2, 4
- Doesn't predict death well 2
- Only classifies surgery as "high-risk" vs "other" - too crude 1
SORT weaknesses:
- Not specifically designed for cardiac complications - misses pulmonary edema and heart block that RCRI captures 2
- Less validation in cardiovascular guideline literature
- Requires more detailed surgical classification
Practical Algorithm
For elective non-cardiac surgery in patients ≥45 years or with known cardiovascular disease 3:
Always calculate RCRI (guideline-mandated) 1
- 0-1 points = low cardiac risk, proceed with surgery
- ≥2 points = elevated cardiac risk, consider further evaluation
Add SORT when:
- Patient/surgeon requests mortality estimate
- RCRI suggests intermediate-high risk but you need overall mortality context
- Vascular surgery (where RCRI performs poorly)
- Discussing whether to proceed with surgery at all
For emergency surgery: Neither tool should delay surgery 3 - proceed immediately
The 2024 AHA/ACC guidelines now also recommend structured functional capacity assessment (Duke Activity Status Index) alongside risk indices 8, as poor functional capacity (<4 METs) independently predicts adverse outcomes.
Bottom line: RCRI remains the guideline-endorsed standard for cardiac risk; add SORT when you need mortality prediction or when RCRI's limitations apply.