Written and maintained by CASRAI Editorial Board
Last updated
The Acute Physiology and Chronic Health Evaluation II (APACHE II) score is a bedside instrument that converts a patient’s admission physiology into a single number used to estimate risk of hospital death and to compare severity of illness across ICU patients. Introduced by Knaus and colleagues in 1985 (Critical Care Medicine 13(10):818–829), it remains one of the most widely used ICU severity scores four decades later, despite two newer versions existing.
This page covers what the score is built from, how it is calculated, what it is actually used for in a patient-safety or quality-improvement context, and how it differs from SOFA — a different instrument that gets confused with APACHE II because both are calculated in the ICU from lab and vital-sign data. For the organ-dysfunction-tracking instrument, see CASRAI’s guide to the SOFA score and qSOFA score; for the wider deterioration-detection and severity-scoring landscape, see the Patient Safety & Infection Prevention hub.
The three components
The APACHE II score is the sum of three separately calculated parts. The total ranges from 0 to 71, though in practice it is rare for a patient to accumulate more than about 55 points.
1. Acute Physiology Score (APS) — 12 variables
The APS is built from the worst value recorded for each of 12 physiologic variables in the first 24 hours after ICU admission:
- Rectal (core) temperature
- Mean arterial pressure
- Heart rate
- Respiratory rate
- Oxygenation (A–aDO2 or PaO2, depending on FiO2)
- Arterial pH
- Serum sodium
- Serum potassium
- Serum creatinine (doubled if the patient is in acute renal failure)
- Hematocrit
- White blood cell count
- Glasgow Coma Scale (subtracted from 15 to give the points contribution)
Each variable is scored 0–4 points depending on how far it deviates from a normal range in either direction — the scoring is symmetric, so both a very low and a very high value for the same variable can score the same points. “Worst value” means whichever single reading in the first 24 hours is furthest from normal, not an average or an admission value.
2. Age Points
Age contributes 0–6 points on a stepped scale, with older patients scoring more points independent of their physiology — this is a fixed demographic adjustment, not a clinical measurement.
3. Chronic Health Points
Chronic Health Points (0, 2, or 5) are added only if the patient has a documented history of severe organ-system insufficiency or is immunocompromised, and only apply differently depending on whether the ICU admission was for an elective postoperative reason or a nonoperative/emergency reason. A patient with none of the qualifying chronic conditions scores 0 here regardless of age or acute physiology.
The three components sum to the total APACHE II score:
APACHE II = Acute Physiology Score + Age Points + Chronic Health Points
What APACHE II is actually used for
Two distinct uses get conflated, and separating them matters for how a patient-safety or quality department should treat the number:
- Individual risk estimation. The original 1985 publication paired the raw score with a disease-category-specific coefficient to generate a predicted hospital mortality probability for an individual patient. In practice, most ICUs today use APACHE II as a severity indicator and triage/prognostic input alongside clinical judgment, not as a standalone mortality calculator for bedside decisions about an individual patient — a specific predicted-mortality percentage should be treated cautiously without the full disease-category coefficient set behind it.
- Case-mix adjustment and ICU-to-ICU benchmarking. This is the use case most relevant to a hospital quality or patient-safety office: because APACHE II is calculated the same way everywhere, it gives ICUs a common severity yardstick. Comparing raw mortality rates between two ICUs (or the same ICU over time) is misleading if one unit admits sicker patients — APACHE II-based case-mix adjustment lets a standardized mortality ratio (observed deaths divided by APACHE II-predicted deaths) be calculated so that ICUs, or the same unit across quarters, can be compared on a severity-adjusted basis rather than raw counts. See CASRAI’s guide to the standardized mortality ratio for how that adjustment is typically constructed.
Because it is recalculable from routinely collected admission data, APACHE II is also commonly embedded in ICU quality dashboards and used as an eligibility/stratification variable in critical-care research — a trial enrolling septic-shock patients, for example, may report baseline APACHE II scores by arm to demonstrate the randomization balanced severity of illness.
APACHE II vs. SOFA: two different instruments, don’t conflate them
APACHE II and the SOFA score are both calculated from ICU lab and vital-sign data, which is why they get confused, but they measure different things and answer different questions:
| APACHE II | SOFA | |
|---|---|---|
| What it measures | Overall physiologic derangement + age + chronic health, at a single point in time | Degree of dysfunction in 6 specific organ systems |
| When it’s calculated | Once, from the worst values in the first 24 hours of ICU admission | Repeatedly — daily or more often, to track a trajectory |
| Primary purpose | Admission severity / mortality-risk estimation and case-mix adjustment across a whole ICU population | Tracking evolving organ dysfunction in an individual patient over the stay; the Sepsis-3 definition of sepsis is built on an acute rise in SOFA |
| Score range | 0–71 | 0–24 |
| Typical unit of comparison | ICU vs. ICU, or the same ICU over time (benchmarking) | Patient’s own trajectory day to day |
In short: APACHE II asks “how sick was this patient on arrival, relative to the general ICU population?” SOFA asks “how is this specific patient’s organ function changing over the course of their stay?” A department can, and often does, use both — APACHE II for admission-level severity/case-mix reporting, SOFA for tracking a specific patient’s course and for sepsis-related definitions.
Newer versions exist — and why APACHE II is still the one most widely used
APACHE II is not the only version. APACHE III followed in 1991, adding physiologic variables and refining disease-category weighting, and APACHE IV followed in 2006, adding further variables (including mechanical ventilation status and the effect of sedation on the Glasgow Coma Scale component) and disease-specific subgroup modeling. Both later versions are more granular and, in validation studies, generally more discriminating predictors of mortality than APACHE II. In practice, though, APACHE II remains the version most commonly cited and calculated in day-to-day ICU use and in published research, largely because of its comparative simplicity and the long track record of external validation behind it — the newer versions’ added complexity has limited how widely they were adopted relative to APACHE II. A hospital selecting a severity-scoring tool for a new case-mix or benchmarking initiative should confirm which version any comparison dataset or published benchmark actually used, since APACHE II, III, and IV scores are not directly interchangeable.
Frequently asked questions
What does an APACHE II score of 25 mean?
A higher score reflects greater physiologic derangement, older age, and/or more chronic health burden at ICU admission, and is associated with higher predicted mortality risk in the original disease-category-adjusted model. A specific number in isolation should not be read as a precise individual mortality percentage without the underlying disease-category coefficient; it is more reliably used as a relative severity indicator for triage context and for case-mix/benchmarking purposes.
When is APACHE II calculated?
Once per ICU admission, using the worst value recorded for each of the 12 physiologic variables during the first 24 hours after admission — it is not typically recalculated daily the way SOFA is.
Does a high APACHE II score by itself justify withdrawing care?
No. APACHE II was designed and validated as a population-level severity-adjustment and research-stratification tool, not as an individual bedside decision instrument for withdrawal of care. Clinical decisions about an individual patient require clinical judgment and the full clinical picture, not a single admission score in isolation.
Is APACHE II the same as SOFA?
No — see the comparison table above. APACHE II is a single admission-severity score across the whole patient (12 variables plus age and chronic health); SOFA is a repeatable, six-organ-system dysfunction score tracked over the ICU stay.
Are APACHE III and APACHE IV replacing APACHE II?
Not in practice. Both are more complex, more data-intensive, and generally more discriminating in validation studies, but APACHE II remains the version most commonly used and reported, particularly outside dedicated APACHE-licensed software environments.
