π₯ SAPS II Calculator
ICU severity scoring with predicted mortality.
For informational purposes only, not medical advice. Consult a healthcare professional.
The Simplified Acute Physiology Score II, from Le Gall and colleagues 1993 European study, predicts hospital mortality for ICU patients from 15 variables measured in the first 24 hours. Points run 0 to 163, converted to a mortality probability by a logistic equation. A score of 30 predicts roughly 10 percent mortality; 50 predicts about 50 percent; 70 predicts near 90 percent. ICUs use it for benchmarking and research stratification, not for individual triage. This calculator reproduces the score and the mortality equation for education and reference.
How to Use This Calculator
Select the band for each of the 15 physiologic variables plus age, chronic disease, and admission type, using the worst values from the first 24 hours in ICU.
Use the worst value of each variable during the first 24 hours after ICU admission, per the original SAPS II method.
Work through all 15 selects. The score accumulates automatically with the published point weights.
Read the total and the predicted hospital mortality from the validated logistic equation.
How the Math Works
Each variable contributes its published points: age up to 18, heart rate up to 11, systolic BP up to 13, temperature, oxygenation if ventilated up to 11, urine output up to 11, BUN up to 10, WBC up to 12, potassium, sodium, bicarbonate, bilirubin up to 9, GCS up to 26, chronic disease up to 17, and admission type up to 8. The mortality equation is logit equals -7.7631 plus 0.0737 times score plus 0.9971 times ln(score+1), converted to probability. The defaults here, a 65-year-old medical admission with normal values, score 18 and predict about 5 percent mortality. GCS dominates the points because neurologic failure predicts death so strongly.
Worked Example
A patient totaling 52 points has a predicted hospital mortality of 50.7%%. The calculator shows this alongside the score because the points alone mean little without the conversion: 52 points sits near the middle of the mortality curve, where each additional handful of points moves the probability steeply. The same equation gives about 10 percent at 30 points and near 90 at 70.
Reading Your Result
Under 20 points predicts single-digit mortality and describes a relatively stable ICU patient. The 30 to 50 band is the steep part of the curve where mortality climbs from roughly 10 to 50 percent. Above 60, predicted mortality exceeds 70 percent and goals-of-care conversations usually parallel the medical management. The score benchmarks populations and stratifies research; at the bedside it informs but never dictates, because calibration drifts across eras and case mixes.
| SAPS II | Predicted mortality | Context |
|---|---|---|
| 20 | ~4%% | Relatively stable |
| 30 | ~10%% | Moderate severity |
| 50 | ~50%% | Steep curve zone |
| 70 | ~90%% | Very high severity |
Limitations
SAPS II predicts for populations, and applying it to individuals invites error; calibration has drifted since 1993 as ICU care improved. It uses first-24-hour worst values, so it cannot guide minute-to-minute decisions. Lead-time bias affects scores computed after variable lengths of pre-ICU care. It excludes cardiac surgery, burns, and young children, and should never drive triage or withdrawal decisions alone.
Related Calculators
Related tools: qSOFA sepsis score sepsis risk screener revised trauma score calculator.
For informational purposes only, not medical advice. Consult a healthcare professional.
Frequently Asked Questions
What is SAPS II?
The Simplified Acute Physiology Score II: 15 physiologic and demographic variables from the first 24 ICU hours, weighted 0 to 163 points and converted to predicted hospital mortality by a logistic equation. Published by Le Gall et al. in 1993 from European ICUs. The standard severity score for ICU benchmarking.
How is SAPS II different from APACHE?
Both predict ICU mortality from admission physiology. APACHE uses more variables with diagnostic weights and has gone through more revisions; SAPS II is simpler with 15 variables and no diagnostic coefficients. Many European ICUs favor SAPS II, many American ones APACHE.
What SAPS II predicts 50%% mortality?
About 50 to 52 points. The curve is S-shaped: roughly 10%% at 30 points, 25%% at 40, 50%% at 52, 75%% near 60, and 90%% around 70. The steep middle is why small point changes matter most there.
Can SAPS II decide who gets the ICU bed?
No. It is a benchmarking and research tool, not a triage instrument, and using it to deny care would be both invalid and unethical. Triage uses clinical judgment, reversibility, and patient preferences. The score describes populations, not fates.
Why does GCS score so many points?
Because neurologic failure was the strongest mortality predictor in the derivation data: GCS under 6 contributes 26 points, more than any other single variable. The weights are empirical from the original logistic regression.
What are the worst-value conventions?
Each variable uses its most abnormal value during the first 24 hours after ICU admission. This standardizes the score across patients with different measurement timing. Values before ICU admission do not count.
Does SAPS II work for children?
No. Pediatric physiology and norms differ, and pediatric ICUs use PRISM or PIM scores instead. SAPS II is validated for adults 18 and older.
How has SAPS II held up over time?
Reasonably for ranking severity, less well for absolute calibration: modern ICU mortality runs lower than 1993 predictions at the same score. SAPS 3 was developed partly to address this. Recalibration to local data improves accuracy.
What is the maximum SAPS II?
163 points, predicting near-certain mortality. In practice scores above 100 are uncommon and the equation saturates. The realistic working range in most ICUs is 10 to 80.
Should families be told the SAPS II?
The number can inform honest prognostic conversations, but quoting a percentage as a personal prediction overstates its precision. Experienced clinicians translate it into words about the overall picture, uncertainty included, rather than reciting the equation output.