APACHE II Score
- Introduced by
- William A. Knaus, et al.
- Year
- 1985
- Type
- ICU severity and mortality prediction
- Family
- Clinical scoring
What does it do?
The Acute Physiology and Chronic Health Evaluation (APACHE) II score, introduced by Knaus et al. in 1985, is a 71-point severity of illness classification system for critically ill patients.
It combines acute physiological parameters, age, and chronic health status to predict intensive care unit (ICU) mortality, facilitating patient risk stratification and research standardization.
The intuition behind the method
APACHE II operationalizes the concept that ICU mortality risk depends on the degree of acute organ dysfunction, patient age, and underlying chronic conditions. By integrating 12 physiological variables (vital signs, laboratory values, neurological status), age, and chronic health history, the score captures a patient's overall vulnerability. Higher scores correlate strongly with mortality risk, enabling clinicians to identify high-risk patients, adjust expectations, and facilitate informed discussions with families. APACHE II also standardizes severity assessment across ICU populations for quality improvement and outcome research.
History
Knaus et al. developed APACHE II at George Washington University in 1985 as an update to the original 1981 APACHE system. The addition of an age component and revised physiological point assignments improved discrimination. APACHE II became the de facto standard for ICU severity assessment throughout the 1990s. Later refinements (APACHE III, 1991; APACHE IV, 2006) were developed but APACHE II remains widely used due to simplicity and familiarity.
When to use it
APACHE II is primarily used in ICU settings for severity assessment, quality adjustment in mortality comparison studies, and research on critical illness. It is most reliable in mixed ICU populations but may over- or under-predict mortality in specific subgroups (e.g., trauma, cardiac surgery). Serial APACHE II scores can track clinical trajectory, though daily application is uncommon in routine care.
Limitations and considerations
- Requires collection of multiple data points within 24 hours, demanding intensive data management
- Does not account for treatment limitations (e.g., DNR orders) that influence actual mortality
- Calibration varies by ICU, patient population, and time period; models from one center may not apply to another
- Physiological parameters are often redundant, and the score may not capture unique disease-specific factors (sepsis biology, cardiac disease)
- Does not predict long-term outcomes such as functional status or quality of life
Source conditions do not replace independent scientific review or validation by an analysis engine.
Strengths
- Comprehensive assessment integrating multiple physiological systems, age, and comorbidity
- Strong discrimination and calibration for hospital mortality prediction across diverse ICU populations
- Enables fair comparison of mortality rates between ICUs by adjusting for case severity
- Supports research standardization and clinical trial stratification
- Well-validated over decades with extensive literature on its use and limitations
Applications
- Hospital and regional ICU performance monitoring by comparing observed vs. expected mortality, adjusted for APACHE II severity
- Clinical trial inclusion and stratification to ensure comparable baseline severity between treatment arms
- Prognostic information for family discussions: communication of mortality risk based on APACHE II in context of patient values
- Sepsis and ARDS outcome research, predicting organ support duration and ICU length of stay
- Health services research on resource utilization, examining whether high APACHE II correlates with intensive interventions and costs
How it works
How it works
Over the first 24 hours of ICU admission, clinicians collect the worst (or most deranged) values for 12 physiological parameters: temperature, mean arterial pressure, heart rate, respiratory rate, oxygenation, pH, sodium, potassium, creatinine, hematocrit, white blood cell count, and Glasgow Coma Scale score. Patient age and chronic health conditions are also documented.
Each physiological parameter is assigned 0–4 points based on deviation from normal ranges; points are summed (0–60). Age points (0–6) are added; chronic health points (0–5) are added if the patient has severe organ dysfunction or immunocompromised state. Total APACHE II ranges from 0 to 71.
APACHE II < 10 predicts ~1–5% hospital mortality; 10–19 predicts ~5–15% mortality; 20–29 predicts ~15–40% mortality; 30–39 predicts ~40–80% mortality; ≥ 40 predicts > 80% mortality. Mortality prediction requires calibration to the local population and must be combined with clinical judgment.
Common pitfalls
- Using APACHE II to predict individual patient outcomes or make life-sustaining treatment decisions without considering clinical context, goals of care, and family preferences.
- Failing to validate the local APACHE II–mortality relationship; applying published coefficients to a different ICU population introduces bias.
- Collecting worst values retrospectively instead of prospectively during the first 24 hours, introducing selection bias.
- Ignoring the score's population-level predictive power; a high APACHE II does not mean a specific patient will die.
Frequently asked questions
Should I use APACHE II to decide whether to withdraw life support?
No. APACHE II is a population-level prognostic tool, not an individual decision-making instrument. A high APACHE II indicates increased risk but does not determine a patient's outcome or preferences. Always incorporate clinical judgment, patient autonomy, family wishes, and reversibility of the underlying condition.
Why does APACHE II differ from SOFA?
APACHE II assesses severity at ICU admission (snapshot); SOFA assesses organ dysfunction daily (dynamic). APACHE II includes age and chronic health; SOFA does not. SOFA is simpler and better for tracking improvement/deterioration. Both are valid; choice depends on clinical context.
How do I score APACHE II if the patient is comatose or sedated?
Use the Glasgow Coma Scale component directly; do not substitute clinical judgment. For sedated patients, the GCS will be low by design. Document sedation status. Some centers use APACHE III or modified APACHE II for sedated patients.
Can APACHE II predict which patients will need prolonged mechanical ventilation?
APACHE II predicts mortality but only moderately predicts ICU or mechanical ventilation duration. APACHE II combined with age, comorbidity, and reason for intubation provides better prediction. Consider disease-specific tools for more precise outcome prediction.
Related methods
Sources
Knaus, W. A., Draper, E. A., Wagner, D. P., & Zimmerman, J. E. (1985). APACHE II: a severity of disease classification system. Critical Care Medicine, 13(10), 818-829.
Open DOI referenceCopied.Zimmerman, J. E., Kramer, A. A., McNair, D. S., & Malila, F. M. (1996). Variations in resource utilization among intensive care units in the United States. Critical Care Medicine, 24(8), 1261-1268.
Open publication referenceCopied.