SAPS II Score Calculator

SAPS II explained: 24-hour worst values, admission type, physiology, GCS, chronic conditions, and how the total score maps to estimated hospital mortality. For education and risk stratification in critical care.

SAPS II (Simplified Acute Physiology Score II)

Demographics, admission, chronic health

Age (years)

Type of ICU admission

Uses the original SAPS II surgical vs medical admission definitions (Table 2).

Chronic diseases (original Table 2)

AIDSHematologic malignancyMetastatic cancer

Physiology — worst value in first 24 h of ICU

Heart rate (bpm)

Systolic BP (mmHg)

Temperature — rectal or bladder (°C)

Mechanically ventilated or CPAP — use PaO₂/FiO₂ for the oxygenation component

FiO₂ (0.21–1)

PaO₂ (mmHg)

Urine output (mL / ~24 h)

BUN (mg/dL)

WBC (10⁹/L)

Potassium (mEq/L)

Sodium (mEq/L)

Bicarbonate (mEq/L)

Total bilirubin (mg/dL)

Glasgow Coma Score (3–15)

Disclaimer: SAPS II supports severity adjustment and research benchmarking. It is not a substitute for clinical judgment. Mortality estimates reflect 1990s-era calibration and may not match modern ICU populations.

SAPS II — Formula and references

What SAPS II is

The Simplified Acute Physiology Score II (SAPS II) summarizes 12 physiology variables, age, and Glasgow Coma Score, plus optional chronic health points (AIDS, hematologic malignancy, metastatic cancer). It was developed from European/North American ICU data. Physiology uses the worst value in the first 24 hours after ICU admission; admission type (scheduled surgical, medical, or unscheduled surgical) is determined at ICU entry per the paper.

Total score and mortality

SAPS II = sum of age, admission type, HR, systolic BP, temperature, oxygenation (ventilated vs not), urine output, BUN, WBC, K⁺, Na⁺, HCO₃⁻, bilirubin, GCS, and chronic disease points (if present).

When the patient is mechanically ventilated (or otherwise on supplemental O₂ where PaO₂/FiO₂ is used), oxygenation points use the ratio in mmHg. Otherwise use PaO₂ alone.

Original logistic model for hospital death before ICU discharge:

Logit = −7.7631 + 0.07337 × SAPS II + 0.9971 × ln(SAPS II + 1)

P(death) = e^Logit / (1 + e^Logit)

Calibration differs by era and case mix; treat the percentage as a reference estimate, not a guaranteed individual probability.

Important usage notes

Use the worst value in the first 24 h of ICU stay (original paper definition), not a single “admission snapshot.”
BUN is serum urea nitrogen in mg/dL; chronic points reflect pre-ICU conditions per the original definitions.
Urine output should reflect roughly 24-hour balance where possible; oliguria thresholds drive high points.

Reference

Le Gall JR, Lemeshow S, Saulnier F. A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study. JAMA. 1993;270(24):2957–2963. doi:10.1001/jama.1993.03510240069035.

What SAPS II measures

The Simplified Acute Physiology Score II (SAPS II) is one of the major second-generation ICU severity tools. It compresses how sick a patient is during the early ICU course into a single integer that can be compared across diagnoses, used for severity adjustment in research, and translated—via a fixed logistic equation from the original development cohort—into an estimated probability of dying before hospital discharge.

Unlike scores built around a primary diagnosis, SAPS II is deliberately generic: the same variables and weights apply whether the patient has sepsis, a stroke, respiratory failure, or postoperative complications. That generality is useful for broad ICU registries and outcomes comparisons, but it also means SAPS II does not encode diagnosis-specific pathophysiology the way disease-focused rules might.

Where the data come from (conceptually)

SAPS II was derived from a large, multinational sample of consecutive adult ICU admissions—the kind of population that reflected real-world mixed medical and surgical ICUs at the time of the study. Patients were excluded when they fell outside the intended scope (for example, younger adults, specialized cohorts such as burns or certain cardiac surgery pathways in the original definitions). The model was split into development and validation sets; reporting included calibration and discrimination metrics so readers could judge how well raw scores tracked with observed hospital mortality.

The calculator on this site applies the same published coefficient structure for the component points and the same form of logistic transformation for mortality, so you can reproduce “textbook” totals and reference-line risk estimates when inputs match the original variable definitions.

Timing: the first ICU day

SAPS II is anchored to the first twenty-four hours after ICU admission, not a single pre-ICU snapshot. For each physiologic and laboratory input, you are meant to use the value associated with the worst degree of abnormality within that window—the measurement that would assign the highest points for that variable if more than one value is available. This “worst-in-window” rule is easy to state but labor-intensive at the bedside; high-quality scoring for audit or research typically relies on structured data collection and explicit local policies for missing data.

Operationally, teams should not recompute SAPS II serially throughout the ICU stay as a trending score; the score is defined once around the index period. If a patient is discharged from the ICU and later readmitted, a new twenty-four-hour window begins, and a new score may be calculated under the same rules.

The seventeen ingredients (conceptual map)

The classic description is “twelve physiological variables, age, type of admission, and three underlying disease indicators,” plus Glasgow Coma Scale as the neurologic anchor among the physiologic set. Together they produce a bounded integer score that correlates with short-term mortality risk in the original cohorts.

Admission type distinguishes scheduled surgical, medical, and unscheduled (emergency) surgical pathways. This captures baseline risk related to how and why the patient entered the ICU independent of the minute-to-minute vitals.
Age enters as grouped bands: older age strata receive more points because, all else equal, the same degree of physiologic instability carries higher risk in advanced age.
Circulation and rhythm proxies include heart rate and systolic blood pressure—simple to chart but powerful because hypotension and extreme tachycardia mark perfusion stress.
Temperature (core, rectal or bladder in the original framing) flags high-grade fever or hyperthermia patterns that are penalized in the table.
Oxygenation branches depending on whether the patient is receiving mechanical ventilation or CPAP, where PaO₂/FiO₂ ratio is used, versus spontaneous breathing, where PaO₂ alone drives the subscore. That split mirrors how clinicians escalate support when gas exchange fails.
Renal trajectory is represented by twenty-four-hour urine volume and blood urea nitrogen (BUN), aligning with acute kidney injury patterns and catabolic stress.
Inflammation and marrow appear through white blood cell count extremes (both severe leukopenia and marked leukocytosis can be penalized).
Electrolytes and acid–base use sodium, potassium, and bicarbonate bands to reward homeostasis and punish dangerous derangements.
Hepatobiliary stress uses total bilirubin cutoffs that reflect cholestasis and liver dysfunction relevant to multi-organ failure.
Glasgow Coma Scale encodes level of consciousness; lower scores add substantial points because obtundation often signals severe intracranial, toxic, or systemic illness.
Chronic health flags mark AIDS, hematologic malignancy, and metastatic cancer when present as background conditions that independently worsen prognosis.

From points to a mortality percentage

After summing component points, SAPS II can be plugged into the published logistic model: a linear combination of the total score and the natural logarithm of (score + 1) yields a log-odds, which is then passed through the logistic function to give a percentage. That percentage is a group-level calibration artifact from the era and case mix used to fit the model, not a guaranteed personal probability.

Modern ICUs differ in baseline risk, coding, early mobility, sedation practice, sepsis bundles, and mechanical ventilation strategies. Even when the arithmetic of SAPS II is perfect, the predicted mortality can be miscalibrated high or low at your hospital. In practice, contemporary systems often prefer refreshed models (for example, SAPS 3) or institution-specific recalibration for administrative benchmarking, while still teaching SAPS II because it remains embedded in historical literature and some tertiary datasets.

Using this calculator responsibly

Treat the output as an adjunct for communication, education, and risk stratification—not a sole basis for withholding or intensifying therapy.
Align laboratory units and temperature sites with the definitions your quality program uses; mixed conventions are a common source of silent error.
Document how missing data were handled if you rely on SAPS II for registry submission; imputation rules should be explicit.
When comparing SAPS II across hospitals or years, remember that identical scores can imply different true risks if admission practices and exclusion criteria differ.

Relationship to other ICU scores

SAPS II sits alongside APACHE II and successive SAPS versions as part of the same conceptual family: reduce rich clinical data to a severity index that correlates with mortality. Each system makes different tradeoffs among variable count, collection burden, time window, and calibration updates. SAPS II remains comparatively compact and historically influential; SAPS 3 reflects a later generation of admission-time modeling with different boxes and regional customization options. Choosing among them is less about which is “most correct” in abstract and more about which matches your data governance, outcome definition, and external benchmark.