Clinical background
Preoperative cardiovascular risk stratification before non-cardiac surgery aims to distinguish patients who can be operated on without specific cardiological investigation from those who require optimisation, monitoring or, sometimes, a change of strategy. The decision is difficult because the individual patient's risk depends on a combination of patient-related factors and the nature of the procedure, and because overinvestigation is costly and can delay necessary surgery.
The established tool, the Revised Cardiac Risk Index (RCRI), has been criticised for underestimating risk in modern surgical practice, particularly in high-risk groups such as vascular and emergency surgery, and because its derivation cohort was narrow: 4,315 patients at a single hospital, without low-risk or emergency patients [1]. The AUB-HAS2 cardiovascular risk index was developed to address these weaknesses by including symptoms of active cardiac disease, age stratification and anaemia as variables, and by using a broader outcome measure including death and stroke.
Calculating the AUB-HAS2 cardiovascular risk index
The index consists of six binary variables, each worth 1 point:
where:
- = History of heart disease (previous myocardial infarction, coronary angioplasty, cardiac surgery, heart failure, atrial fibrillation or moderate/severe valvular disease confirmed by echocardiography)
- = Symptoms of angina or dyspnoea
- = Age ≥ 75 years
- = Anaemia (haemoglobin < 12 g/dL)
- = Vascular surgery
- = Emergency surgery
The score ranges from 0 to 6. The calculator classifies 0 as low risk, 1 to 2 as intermediate risk and 3 or higher as high risk.
The index was derived by Dakik and colleagues at the American University of Beirut and validated in a large retrospective cohort of 1,167,278 non-cardiac surgical procedures recorded in the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) between 2008 and 2012 [1]. Data after 2012 were not included because NSQIP then stopped recording patients' cardiac history, a central variable in the index. The derivation cohort included both elective and emergency procedures and low-risk patients, unlike the RCRI cohort. The outcome measure was a composite of death, myocardial infarction or stroke within 30 days of surgery.
Interpretation in practice
| Score | Risk class | 30-day event rate (death, myocardial infarction or stroke) | Clinical action |
|---|---|---|---|
| 0 | Low | <0.5% in most procedure types [1]; 0.3% in a CAD cohort [2] | Further cardiological investigation or specific postoperative monitoring is rarely warranted. The patient's functional capacity and the indication for surgery govern. |
| 1–2 | Intermediate | Approximately 2–5% depending on the procedure [1,2] | Individual assessment. At a score of 1–2 with a low-risk procedure, surgery can usually proceed with routine monitoring. With vascular surgery or other high-risk procedures, consider whether preoperative optimisation or enhanced postoperative monitoring is warranted. |
| ≥3 | High | ≥10% in most procedures; up to >30% in aortic aneurysm repair and colonic surgery [1]; 21.7% at a score >3 in a CAD cohort [2] | Preoperative cardiological assessment throughout, optimisation of evidence-based medical therapy and enhanced postoperative monitoring. |
An important feature is that more than half of all patients in the NSQIP cohort scored 0, apart from vascular surgery patients, who by definition score at least 1 [1]. The main value of the index therefore lies in identifying quickly and reliably the large proportion of patients whose risk is so low that further investigation adds nothing, rather than in fine-tuning the risk estimate for the intermediate group.
Validation and performance
In the large NSQIP validation cohort, AUB-HAS2 showed an AUC of 0.818 for the composite 30-day outcome, compared with 0.716 for RCRI (p < 0.001) [1]. The superiority held across all surgical specialties and site-specific procedures. The AUC was highest in general surgery (0.83) and orthopaedics (0.81), which together made up 73% of the population, and lowest in vascular surgery (0.71) and thoracic surgery (0.71). Even in vascular surgery, however, there was a clear risk gradient, from 3.1% at a score of 1 to 37.8% at a score >3 in open aortic aneurysm repair.
A Chinese multicentre study at two tertiary hospitals in Zhejiang validated the index specifically in 10,294 patients with documented coronary artery disease undergoing non-cardiac surgery between 2013 and 2024 [2]. The c-statistic was 0.765 for AUB-HAS2 versus 0.689 for RCRI (p < 0.001), with better calibration on the calibration plots. The event rate rose progressively: 0.3%, 2.3%, 5.2%, 10.1% and 21.7% for scores of 0, 1, 2, 3 and >3. The improvement over RCRI was not, however, statistically significant in three subgroups: patients with diabetes, those given anaesthesia other than general, and certain low-risk procedure types (orthopaedics, neurosurgery, gynaecology, ophthalmology, dentistry and ENT) [2].
A retrospective analysis of NSQIP data compared performance in emergency and elective surgery and found comparable discrimination in both groups, with an AUC around 0.80 in each, even though the event rate was five times higher in emergency surgery (7.0% versus 1.4%) [3].
In 32,337 patients with end-stage renal disease in the NSQIP database, AUB-HAS2 was superior to RCRI in predicting both mortality and the composite cardiovascular outcome, with mortality rising from 1.8% to 23% across the score steps [4].
In a smaller German prospective study of 199 patients aged ≥65 years undergoing elective non-cardiac surgery of intermediate or high surgical risk, AUB-HAS2 was the only one of the simple risk indices to predict postoperative morbidity significantly (AUC 0.646), while RCRI did not reach significance [5]. Adding NT-proBNP improved the predictive value of the index (AUC 0.703), suggesting that biomarkers may complement the index in selected cases.
Limitations
AUB-HAS2 has so far been validated exclusively in retrospective cohorts, most of them based on NSQIP data from North American hospitals. The only prospective validation is a short report without detailed statistics [6]. The multicentre study from China is to date the only external validation outside the NSQIP framework [2].
NSQIP data from 2008 to 2012 reflect an earlier surgical era. Modern minimally invasive techniques may have changed procedure-specific risk, particularly for intrathoracic procedures, which the RCRI classifies as high risk but where the complication rate today may be lower [2].
The index does not capture functional capacity, a central parameter in the current AHA/ACC 2024 guidelines for preoperative evaluation [7]. A patient with good functional capacity and a score of 2 may have a lower actual risk than the score suggests, and vice versa. The index should therefore be seen as a screening tool that complements, rather than replaces, clinical assessment of functional capacity.
The variable "history of heart disease" is broad and gathers ischaemic heart disease, heart failure, atrial fibrillation and valvular disease under a single point. This simplifies use but means that a patient with a PCI five years earlier and no symptoms scores the same as a patient with active heart failure, despite very different risks.
The anaemia threshold is set at Hb < 12 g/dL irrespective of sex, which may overclassify women, for whom lower thresholds are physiologically justified.
References
- Dakik HA, Sbaity E, Msheik A, et al. AUB-HAS2 Cardiovascular Risk Index: Performance in Surgical Subpopulations and Comparison to the Revised Cardiac Risk Index. J Am Heart Assoc. 2020;9(10):e016228. PMID: 32390481
- Li X, Wang C, Jiang H, et al. Performance of the AUB-HAS2 cardiovascular risk index in coronary artery disease: a multicenter retrospective cohort study. Ann Med. 2026;58(1):2664250. PMID: 42029725
- Sbaity E, Tamim H, Zalaquett NG, et al. Comparison of the performance of the AUB-HAS2 Cardiovascular Risk Index in emergency vs elective surgeries. J Cardiol. 2024;84(1):55–58. PMID: 38382579
- Patel R, DiPastina K, Bhat V, et al. Comparing Revised Cardiac Risk Index and American University of Beirut HAS2 in End-Stage Renal Disease Patients Undergoing Noncardiac Surgery: A Retrospective Analysis of the National Surgical Quality Improvement Program Database. Nephron. 2025;149(11):653–660. PMID: 40505641
- Schmidt G, Frieling N, Schneck E, et al. Comparison of preoperative NT-proBNP and simple cardiac risk scores for predicting postoperative morbidity after non-cardiac surgery with intermediate or high surgical risk. Perioper Med (Lond). 2024;13(1):44. PMID: 38760848
- Dakik HA, Eldirani M, Kaspar C, et al. Prospective validation of the AUB-HAS2 cardiovascular risk index. Eur Heart J Qual Care Clin Outcomes. 2022;8(1):96–97. PMID: 33017006
- Mangano N, Ganesan V, Shibly Y, et al. Update in Perioperative Ischemic Workup: Integrating 2024 AHA/ACC Guidelines and Contemporary Evidence. J Cardiovasc Dev Dis. 2026;13(7):309. PMID: 42505886
