AUB-HAS2 Cardiovascular Risk Index Calculator
AUB-HAS2 Cardiovascular Risk Index Calculator
Six facts, one point each, scored before non-cardiac surgery. Derived at the American University of Beirut and validated in 1.17 million NSQIP operations, where it discriminated better than the Revised Cardiac Risk Index in every surgical subgroup.
AUB-HAS2 index
6 criteria → 0–6Known coronary artery disease; breathless climbing one flight of stairs; age 68; haemoglobin 134 g/L; elective laparoscopic cholecystectomy
Scoring
0 to 1 low · 2 to 3 intermediate · more than 3 high
- which question this answers
- This score answers one question: what does an elective operation cost this patient. It does not assess a patient who has presented with possible myocardial ischaemia, which is what GRACE and the HEART score are for. The outcome here is death, myocardial infarction or stroke within 30 days of non-cardiac surgery — wider than the RCRI‘s, which excluded death and did not count stroke
- discrimination, and against the RCRI
- area under the curve 0.90 in the Beirut derivation cohort and 0.82 in validation. In the NSQIP validation of 1,167,278 operations from 2008 to 2012 it reached 0.818 against the RCRI’s 0.716 (P<0.001), and was higher in every subgroup — from 0.71 in vascular and thoracic surgery to above 0.80 in general, orthopaedic and plastic surgery, and by site from 0.73 for spine to 0.83 for cholecystectomy
- why the comparison is not quite fair
- the authors attribute their advantage to broader data elements, a wider derivation population and a more comprehensive outcome — and that last point cuts both ways. The RCRI was derived against a narrower outcome in 1999 and is being scored here against a wider one, so some of the 0.818 against 0.716 gap is a difference in what is being predicted rather than in how well. The 2022 ESC guideline declined to recommend any single score and said none can be disqualified on current evidence
- the cohorts, and what they are not
- derivation was a single tertiary centre in Beirut; validation was a North American surgical registry. Neither is a random sample of surgery anywhere, and the two papers give the NSQIP denominator as 1,167,414 and 1,167,278 — a difference of 136 operations, printed here because the two figures are both published and this page does not reconcile sources silently
Worked example
Known coronary artery disease; breathless climbing one flight of stairs; age 68; haemoglobin 134 g/L; elective laparoscopic cholecystectomy
History of heart disease 1 + symptoms (dyspnoea) 1 = 2 points
2 points is the intermediate stratum, where the Beirut derivation cohort recorded 2.0 per cent for death, myocardial infarction or stroke at 30 days
The same patient on the RCRI scores 1 — class II — because that index has no symptom item and a cholecystectomy is not high-risk surgery in Lee's narrow definition
So one index puts this patient in its second stratum of four and the other in its second of three. The numbers are not comparable and the two should not be averaged
Add an emergency presentation and a haemoglobin of 108 g/L and the total reaches 4, the high stratum, where NSQIP found 10 per cent or more in every specialty
2.0 per cent is a frequency among 3,284 patients in one Beirut hospital. It is not the chance that this patient has an event
The six criteria, and how they differ from the RCRI’s
| AUB-HAS2 criterion | Scores 1 when | The RCRI’s equivalent |
|---|---|---|
| History of heart disease | Heart disease of any kind | Narrower — ischaemic heart disease only, by a specified definition |
| Symptoms of heart disease | Angina or dyspnoea | No equivalent at all. The RCRI has no symptom item |
| Age | 75 years or over | No equivalent. Age does not appear in the RCRI |
| Anaemia | Haemoglobin below 12 g/dL (120 g/L) | No equivalent. The RCRI’s only laboratory item is creatinine |
| Vascular surgery | Vascular surgery of any kind | Partly — suprainguinal vascular only, inside the high-risk surgery item |
| Emergency surgery | Emergency presentation | No equivalent. Emergencies were excluded from the RCRI’s derivation |
What each score predicted in the derivation cohort
| Score | Stratum | 30-day death, MI or stroke | NSQIP validation group size |
|---|---|---|---|
| 0 | Low | 0% | 583,161 |
| 1 | Low | 0.5% | 361,973 |
| 2 | Intermediate | 2.0% | 149,057 |
| 3 | Intermediate | 5.6% | 56,622 |
| More than 3 | High | 15.7% | 16,465 |
A newer index, a wider outcome, and why the comparison is awkward
AUB-HAS2 was derived in 2019 from 3,284 patients prospectively enrolled before non-cardiac surgery at the American University of Beirut Medical Center, where death, myocardial infarction or stroke within 30 days occurred 38 times — 1.2 per cent. Six facts predicted it independently: a history of heart disease, symptoms of heart disease, age 75 or over, a haemoglobin below 12 g/dL, vascular surgery and emergency surgery. Each is worth one point, so the score is simply how many apply, and the strata are 0 to 1, 2 to 3, and more than 3.
Its claim to attention is the validation. The authors applied it to 1,167,278 operations in the American College of Surgeons NSQIP registry from 2008 to 2012 and reported an area under the curve of 0.818 against the Revised Cardiac Risk Index’s 0.716, higher in every surgical subgroup and at every site examined. The weakest subgroups were vascular and thoracic surgery at 0.71 — which is where preoperative cardiac risk is most often asked about, and where the RCRI also performs worst.
The comparison deserves a caveat the headline figure does not carry. The authors themselves attribute part of their advantage to a more comprehensive outcome: AUB-HAS2 was built to predict death, infarction and stroke, while Lee’s index was built in 1999 to predict major cardiac complications and explicitly excluded death. Scoring a 1999 index against a 2019 outcome and finding it worse is partly a statement about the outcome. The 2022 ESC guideline looked at the field and declined to recommend any single score, noting that none can be disqualified on current evidence.
Two things about the published strata are worth stating plainly. The derivation cohort recorded no events at a score of zero, and that zero is a consequence of 38 events in 3,284 patients rather than a demonstration that the risk is nil — the much larger NSQIP validation found a score of zero running below 0.5 per cent in most specialties, which is small and is not zero. And the index collapses 4, 5 and 6 into one top stratum, so like the RCRI it has no resolution where the risk is highest. Derivation in a single tertiary centre in Beirut and validation in a North American registry also means neither cohort is a random sample of surgery anywhere. A score’s output is a cohort frequency, not this patient’s probability: a stratum in which 9 per cent had an event describes that stratum, not which 9 per cent. This page reports what a stratum predicted in a named study. It recommends no action.
Frequently asked questions
What does AUB-HAS2 stand for?
History of heart disease, symptoms of Heart disease (angina or dyspnoea), Age 75 or over, anaemia with a Serum haemoglobin below 12 g/dL, vascular Surgery and emergency Surgery — two H items, one A and three S. AUB is the American University of Beirut, where it was derived. Each criterion scores one point and the score is the number present.
Is AUB-HAS2 better than the RCRI?
It discriminated better in the NSQIP validation — 0.818 against 0.716, and higher in every subgroup — but the comparison is not clean. AUB-HAS2 was designed to predict 30-day death, infarction and stroke, while the RCRI was designed in 1999 to predict major cardiac complications and excluded death. Some of the gap is a difference in outcome rather than in quality. The 2022 ESC guideline recommends neither over the other.
Does a score of zero mean there is no cardiovascular risk?
No. The Beirut derivation cohort recorded no events at a score of zero, but it recorded only 38 events among 3,284 patients in total, so that figure rests on absence of observation rather than on evidence of absence. In the 1.17-million-operation NSQIP validation a score of zero ran below 0.5 per cent in most specialties, which is low and is not nil.
Why does AUB-HAS2 include anaemia and the RCRI does not?
Because the two indices were derived from different candidate variables in different eras and populations. AUB-HAS2’s haemoglobin threshold is 12 g/dL — 120 g/L — for both sexes, which is far higher than ARISCAT’s 10 g/dL for pulmonary risk, so the same patient can score for anaemia on one index and not on another.
Can I use this score for a patient presenting with chest pain?
No. Every patient in both cohorts was being assessed before an operation. A patient who has presented with possible acute coronary syndrome is the GRACE, HEART and TIMI scores’ question, and those were derived in exactly that population. The two families of score are not interchangeable in either direction.
Related calculators
References
- Dakik HA, Chehab O, Eldirani M, et al. A new index for pre-operative cardiovascular evaluation. J Am Coll Cardiol. 2019;73(24):3067–3078.
- 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.
- Halvorsen S, Mehilli J, Cassese S, et al. 2022 ESC Guidelines on cardiovascular assessment and management of patients undergoing non-cardiac surgery. Eur Heart J. 2022;43(39):3826–3924.
- Lee TH, Marcantonio ER, Mangione CM, et al. Derivation and prospective validation of a simple index for prediction of cardiac risk of major noncardiac surgery. Circulation. 1999;100(10):1043–1049.
Not medical advice. For healthcare professionals and education. Reference intervals vary by laboratory and assay — always use your own laboratory's. Never base a dose or a treatment decision on this page alone. Full disclaimer at calcengines.com/disclaimer/