Additive EuroSCORE Calculator (1999)

The 1999 points model, with the 2003 logistic EuroSCORE computed live from the same seventeen facts — because the two agree at low risk and diverge enormously at high risk, which is why the logistic model replaced it.

Last updated 8 Oct 2026

Additive EuroSCORE

17 items → 0–46 points

Age (years)Two published readings, and this record implements the common one. Table 2’s wording is “per 5 years or part thereof over 60 years — 1”, which read literally gives 0 points at exactly 60 and 1 at 61. Every implementation found, including the Merck and MSD Manuals calculators, instead bands it: under 60 = 0, 60–64 = 1, 65–69 = 2, and so on to 95–100 = 8. The banded reading is used here. It also reproduces the stated maximum of 45 points. The 2003 logistic model in the second result row uses a third coding again — a continuous term that is 1 at 59 and under and rises by one per year — so the two numbers on this page read the same age three different ways.

Female sex

Chronic pulmonary diseaseA treatment criterion, not a spirometric one — and the definition EuroSCORE II silently inherits, since the 2012 paper gives none.

Extracardiac arteriopathyClaudication, carotid occlusion or stenosis above 50 per cent, or intervention on the abdominal aorta, limb arteries or carotids.

Neurological dysfunctionNot a stroke history as such — Revised Cardiac Risk Index scores that instead — but the functional consequence of one.

Previous cardiac surgery

Serum creatinine above 200 µmol/LAbout 2.3 mg/dL. The published table prints the unit as “mmol/l”, a thousandfold typo — a creatinine of 200 mmol/L is not a number that occurs, and every reproduction reads it as µmol/L, as this record does. See CKD-EPI 2021 for why one creatinine threshold means different filtration in different bodies.

Active endocarditisWhether the case is endocarditis at all is the Duke criteria‘ question.

Critical preoperative stateA closed list, given in full in the table below. It is not a clinical judgement that the patient is unwell.

Unstable anginaA drug-chart criterion — intravenous nitrates until the anaesthetic room — and narrower than the clinical syndrome. EuroSCORE II replaced it with CCS class 4 angina, which is not the same item.

Left ventricular dysfunctionTwo levels split at 30 per cent. EuroSCORE II uses four levels split at 50, 30 and 20, so a patient at 25 per cent is in the bottom category here and in the third of four there.

Recent myocardial infarction

Pulmonary hypertensionAbove 60 mmHg — a much higher bar than EuroSCORE II, which starts scoring at 31. See also pulmonary vascular resistance.

Emergency operationOne binary, where EuroSCORE II has four urgency levels including salvage.

Other than isolated CABGOne point value for everything from a single valve to a triple procedure — the item EuroSCORE II split into four weights.

Surgery on the thoracic aorta

Post-infarct septal ruptureThe heaviest single item in the model, and the one EuroSCORE II dropped altogether — it was too rare in the 2010 cohort to fit.

8pointsExample

72-year-old woman for coronary bypass plus aortic valve replacement; chronic lung disease on inhaled steroids; ejection fraction 40 per cent; nothing else

Two arithmetics over one set of facts

additive = age points + Σ item points · 0–2 low, 3–5 medium, 6+ high
logistic = e^y / (1 + e^y), y = −4.789594 + 0.0666354 × x_age + Σ β_ix_i, x_age = 1 at age 59 or under, +1 per year

why there are two numbers: the 2003 logistic EuroSCORE uses the same seventeen risk factors as the 1999 additive model, refitted as a regression on the same database. So one set of answers produces both, and the difference between them is purely the arithmetic — not different inputs, not a different cohort, not a different outcome. That is what makes the comparison worth printing
the 2003 coefficients: constant −4.789594; age 0.0666354 per unit; female 0.3304052; creatinine above 200 µmol/L 0.6521653; extracardiac arteriopathy 0.6558917; pulmonary disease 0.4931341; neurological dysfunction 0.841626; previous cardiac surgery 1.002625; recent infarction 0.5460218; LVEF 30–50% 0.4191643; LVEF under 30% 1.094443; systolic PA above 60 mmHg 0.7676924; active endocarditis 1.101265; unstable angina 0.5677075; emergency 0.7127953; critical preoperative state 0.9058132; ventricular septal rupture 1.462009; other than isolated coronary surgery 0.5420364; thoracic aortic surgery 1.159787
the divergence, at three risk levels: a 62-year-old man for isolated bypass with nothing else: 1 point against 1.07 per cent — a gap of 0.07 percentage points. A 72-year-old woman for bypass plus valve with lung disease and an ejection fraction of 40: 8 points against 11.19 per cent, a gap of 3.19. A 78-year-old redo with an ejection fraction under 30, creatinine above 200, pulmonary hypertension, arteriopathy and lung disease: 20 points against 88.91 per cent, a gap of 68.9. The two are closest at low risk and the gap grows without limit, because a sum of points is linear and a logistic function saturates
and neither is right at the top: an 88.91 per cent predicted mortality is not a credible figure for an operable patient, and the EuroSCORE project’s own site says the logistic model “significantly overpredicted mortality” — in the 2010 cohort it predicted 7.57 per cent against 3.9 per cent observed. The additive model errs the other way in the same patients. EuroSCORE II was fitted to replace both
derivation and discrimination: developed on 13 302 patients randomly split from the EuroSCORE database of 19 030 and tested on the remainder, with an area under the ROC curve of 0.79 in development and 0.76 in validation. The validation set’s size is printed as 1 497 in Table 1 and 1 479 in the abstract; the first is consistent with the paper’s own totals
which operation this is for: adult cardiac surgery. This model is for CARDIAC surgery. The RCRI and ARISCAT are for non-cardiac surgery and were derived in cohorts from which cardiac operations were excluded; the two sets of numbers are not interchangeable in either direction.

Worked example

72-year-old woman for coronary bypass plus aortic valve replacement; chronic lung disease on inhaled steroids; ejection fraction 40 per cent; nothing else
Age 72 falls in the 70–74 band: 3 points
Female 1 + chronic pulmonary disease 1 + moderate LV dysfunction 1 + other than isolated CABG 2 = 5
Total 8 points, which is the high-risk group — 11.2 per cent observed mortality in the 1999 database
The same seventeen answers through the 2003 logistic equation: age term max(1, 72 − 58) = 14, so 0.0666354 × 14 = 0.9328956, plus 0.3304052 + 0.4931341 + 0.4191643 + 0.5420364 = 2.7176356; y = −4.789594 + 2.7176356 = −2.0719584, giving 11.19 per cent
A gap of +3.19 percentage points between a points sum read as a percentage and the regression fitted to the same facts — and EuroSCORE II gives 6.15 per cent for this patient, lower than either
Set the age to 62 and remove the valve, the lung disease and the ventricular dysfunction: 1 point against 1.07 per cent, a gap of 0.07. The two models agree almost exactly at low risk, which is why the additive one survived for four years
Add a previous cardiac operation, poor ventricular function, a creatinine above 200 and pulmonary hypertension to the worked patient and the gap passes 60 percentage points. Neither number should be read as this patient's chance of dying

The seventeen items, as published in 1999

Seventeen items, eighteen rows, because ventricular function has two. Everything except age is binary, which is what makes the model scorable from memory and what limits it: three 1-point findings outrank one 3-point finding.
Item	Definition as published	Points
Age	Per 5 years or part thereof over 60 years	1
Female	—	1
Chronic pulmonary disease	Long-term use of bronchodilators or steroids for lung disease	1
Extracardiac arteriopathy	Claudication, carotid occlusion or more than 50% stenosis, previous or planned intervention on the abdominal aorta, limb arteries or carotids	2
Neurological dysfunction	Disease severely affecting ambulation or day-to-day functioning	2
Previous cardiac surgery	Requiring opening of the pericardium	3
Serum creatinine	Above 200 µmol/L preoperatively (printed as mmol/l in the paper)	2
Active endocarditis	Still under antibiotic treatment for endocarditis at the time of surgery	3
Critical preoperative state	Ventricular tachycardia or fibrillation or aborted sudden death; cardiac massage; ventilation before the anaesthetic room; inotropes; balloon counterpulsation; or acute renal failure with anuria or oliguria under 10 mL/h	3
Unstable angina	Rest angina requiring intravenous nitrates until arrival in the anaesthetic room	2
LV dysfunction, moderate	LVEF 30–50%	1
LV dysfunction, poor	LVEF under 30%	3
Recent myocardial infarction	Within 90 days	2
Pulmonary hypertension	Systolic PA pressure above 60 mmHg	2
Emergency	Carried out on referral before the beginning of the next working day	2
Other than isolated CABG	Major cardiac procedure other than or in addition to CABG	2
Surgery on thoracic aorta	For disorder of ascending, arch or descending aorta	3
Post-infarct septal rupture	—	4

The same patient through all three EuroSCOREs

Computed on this page’s own arithmetic from the two published coefficient sets. The EuroSCORE II column is filled only for the worked patient, because EuroSCORE II is not a refit of these seventeen items — it takes a creatinine clearance, an NYHA class and a four-level urgency the 1999 model never collected, so the other two rows cannot be mapped onto it without inventing inputs.
Patient	Additive points	Logistic 2003	EuroSCORE II	Additive-to-logistic gap
62, male, isolated bypass, nothing else	1	1.07%	—	+0.07 pp
72, female, bypass + valve, lung disease, EF 40%	8	11.19%	6.15%	+3.19 pp
78, female, redo + valve, EF under 30%, creatinine above 200, PA above 60, arteriopathy, lung disease	20	88.91%	—	+68.91 pp

Why a points sum had to be replaced by a regression

The additive EuroSCORE was published in 1999 and was, for its moment, an excellent piece of design: seventeen facts available at any preoperative clinic desk, integer weights from 1 to 4, and a total meant to be read directly as an approximate percentage risk of death. It was developed on 13 302 patients split from a 19 030-patient European database and tested on the rest, with an area under the ROC curve of 0.79 and 0.76. Its three risk groups — 0 to 2, 3 to 5, and 6 or more — had observed mortalities of 0.8, 3.0 and 11.2 per cent.

It has one structural flaw, and it is a flaw of shape rather than of calibration. Clinical risk factors do not add. Two findings that each triple the odds of death do not between them add six percentage points wherever you started; they multiply the odds, and the same multiplication means something very different at a baseline of 1 per cent and of 20 per cent. A sum of points is linear and mortality saturates, so the two can agree over only part of the range. The authors knew it, and in 2003 refitted the identical seventeen items as a logistic regression — the second number on this page.

What the comparison shows, live, is that the two are nearly indistinguishable at low risk and separate without limit as risk rises. In the published comparison that settled the question — 4 091 patients, observed mortality 3.6 per cent — the additive score predicted 5.4 per cent and the logistic model 8.3 per cent, and the additive score sat above observed mortality in every risk decile except the highest, where it fell below. So the honest statement of the problem is two-sided: the additive model under-predicts the sickest patients, and the logistic model over-predicts everybody. The project’s own site says the logistic EuroSCORE “significantly overpredicted mortality”, and in the 2010 cohort it predicted 7.57 per cent against 3.9 per cent observed — which is what produced EuroSCORE II.

Keep this page for two purposes and no others: reading an older record or paper that quotes an additive EuroSCORE, and seeing what a non-additive risk actually looks like when you force it into a sum. 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. Every figure here comes from a named cohort, and cohorts differ in case mix, era, outcome definition and treatment; where your own institution’s protocol differs, it takes precedence. This page reports what a stratum predicted in a named study. It recommends no action.

Frequently asked questions

Should I use the additive or the logistic EuroSCORE?

Neither, where EuroSCORE II is available: it was fitted in 2010 precisely because both 1999-era models mispredicted on contemporary patients. Between the two older ones the published answer is the logistic model, which keeps a stable observed-to-expected ratio across risk deciles where the additive model does not.

Can I read the additive score as a percentage?

That was the design, and at a total of 0 to 5 it is roughly true. Above about 6 points it stops being true in a direction that matters: the published comparison found the additive score below observed mortality in the highest risk decile while sitting above it in all the others. The second result row on this page shows how far the regression fitted to the same facts has moved by then.

Why does the age item have two different readings?

Because the published wording and the published implementations disagree. “Per 5 years or part thereof over 60 years” read strictly gives nothing at exactly 60; the Merck and MSD Manuals calculators give 1 point for 60 to 64. This record follows the calculators, which is what the instrument is in practice, and the difference is one point for a patient aged exactly 60.

Is a score of 20 really a 90 per cent chance of dying?

No, and that is the point of printing it. The 2003 logistic equation genuinely returns 88.91 per cent for the high-risk patient tabulated above, and no surgical series has ever observed anything like that in operable patients. The logistic EuroSCORE’s over-prediction at the top of its range is documented by its own authors and is why EuroSCORE II was fitted.

What happened to post-infarct septal rupture in EuroSCORE II?

It was dropped. It carried the heaviest single weight in 1999 — 4 points, and a coefficient of 1.462009 in the logistic version — but was too rare in the 2010 cohort to estimate, so the newer model has no term for it at all. A patient having surgery for one is therefore a patient on whom EuroSCORE II is silent about the single most striking thing about them.

Related calculators

References

Nashef SAM, Roques F, Michel P, Gauducheau E, Lemeshow S, Salamon R, and the EuroSCORE study group. European system for cardiac operative risk evaluation (EuroSCORE). Eur J Cardiothorac Surg. 1999;16:9–13.
Roques F, Michel P, Goldstone AR, Nashef SAM. The logistic EuroSCORE. Eur Heart J. 2003;24(1):1–2. Table 1 gives the constant and all eighteen coefficients.
Roques F, Nashef SAM, Michel P, et al. Risk factors and outcome in European cardiac surgery: analysis of the EuroSCORE multinational database of 19 030 patients. Eur J Cardiothorac Surg. 1999;15:816–823.
Jin R, Grunkemeier GL. Does the logistic EuroSCORE offer an advantage over the additive model? Interact Cardiovasc Thorac Surg. 2006;5(1):15–17.
The Merck and MSD Manuals’ additive EuroSCORE calculators, read as the independent corroboration of the 1999 point table and as the source of the banded age reading (under 60 = 0 through 95–100 = 8) and of the stated maximum of 45 points.
The EuroSCORE project’s own site, read for the model’s history, for the statement that the 2003 logistic model “significantly overpredicted mortality”, for the curation by Royal Papworth Hospital and for the absence of any terms of use, copyright notice or licence statement.

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/