What is the accuracy of Apache (Acute Physiology and Chronic Health Evaluation) scoring in predicting mortality in critical care units?

Medical Advisory BoardAll articles are reviewed for accuracy by our Medical Advisory Board
Educational purpose only • Exercise caution as content is pending human review
Article Review Status
Submitted
Under Review
Approved

Last updated: September 28, 2025View editorial policy

Personalize

Help us tailor your experience

Which best describes you? Your choice helps us use language that's most understandable for you.

Accuracy of APACHE Scoring Systems in Predicting Mortality in Critical Care Units

The APACHE II scoring system demonstrates high accuracy for predicting mortality in critically ill patients with a sensitivity of 83.3% and specificity of 91%, though regional calibration is often required due to variations in pre-ICU care patterns across different healthcare systems. 1

Overview of APACHE Scoring Systems

APACHE (Acute Physiology and Chronic Health Evaluation) scoring systems are widely used severity classification tools that evaluate physiological measurements, age, and chronic health status to predict mortality risk in ICU patients:

  • APACHE II:

    • Consists of 12 physiological variables, age points, and chronic health evaluation
    • Total score ranges from 0-71 (higher scores correlate with increased mortality)
    • Cut-off value of 15 provides optimal diagnostic accuracy with sensitivity of 85.3% and specificity of 77.4% 2
  • APACHE IV:

    • Most recent iteration with improved calibration
    • Excellent discrimination (area under ROC curve = 0.88) 3
    • Good calibration (Hosmer-Lemeshow C statistic = 16.9, p = .08)
    • For 90% of 116 ICU admission diagnoses, the ratio of observed to predicted mortality was not significantly different from 1.0 3

Accuracy and Performance

Discrimination and Calibration

  • APACHE II demonstrates high accuracy with area under curve of 0.88 1
  • APACHE IV shows excellent discrimination for mortality prediction (AUC 0.99) in some studies 4
  • Comparison studies show accuracy rates of predicting mortality at 81%, 79%, and 81% for APACHE II, APACHE IV, and SAPS III, respectively 5

Regional Variations and Need for Recalibration

  • The original US APACHE II model showed variable ability to accurately predict risk of death when applied to UK patients, leading to a 'local' UK recalibration 6
  • Pre-ICU care patterns significantly affect model performance:
    • Inadequate resuscitation may result in greater physiological deterioration
    • Excellent pre-ICU resuscitation may result in fewer physiologic derangements on ICU admission 6

Factors Affecting Scoring System Performance

Pre-ICU Care Variations

  • Different admission practices between countries (e.g., 85% of patients admitted directly to ICU from ED in US vs. only 33% in UK) 6
  • Delays in transfer to ICU associated with higher mortality even after case-mix adjustment 6
  • "Lead time bias" - one study found that incorporating pre-ICU data resulted in substantially higher predicted mortality than standard calculations 6

Post-Admission Care Differences

  • Variability in treatments across regions (e.g., fluid resuscitation varied by upwards of 30% among nations) 6
  • Discharge practices vary significantly by region, affecting interpretation of outcome data 6

Comparative Performance with Other Scoring Systems

  • For acute pancreatitis, APACHE II ≥8 is a significant cutoff value for predicting severe disease 6
  • BISAP score (Blood urea nitrogen, Impaired mental status, SIRS, Age >60, Pleural effusion) provides similar accuracy to APACHE II for predicting death but is simpler to apply 6, 1
  • In some studies, APACHE II and SAPS III show slightly higher accuracy than APACHE IV (81%, 81%, and 79% respectively) 5

Clinical Application Considerations

Strengths

  • Robust mortality prediction across various critical illnesses
  • Useful for benchmarking ICU performance
  • Helps in risk stratification and resource allocation

Limitations

  • Relatively cumbersome, requiring collection of multiple parameters 1
  • Accuracy deteriorates over time, requiring periodic updates (APACHE IV was developed to address limitations of earlier versions) 3
  • Regional variations necessitate local calibration for optimal performance 6

Conclusion

APACHE scoring systems provide valuable mortality prediction in critical care settings, but their performance is influenced by regional care patterns, pre-ICU management, and discharge practices. For optimal accuracy, these scores should be periodically updated and calibrated to local populations.

Professional Medical Disclaimer

This information is intended for healthcare professionals. Any medical decision-making should rely on clinical judgment and independently verified information. The content provided herein does not replace professional discretion and should be considered supplementary to established clinical guidelines. Healthcare providers should verify all information against primary literature and current practice standards before application in patient care. Dr.Oracle assumes no liability for clinical decisions based on this content.

Have a follow-up question?

Our Medical A.I. is used by practicing medical doctors at top research institutions around the world. Ask any follow up question and get world-class guideline-backed answers instantly.