APACHE IIComprehensive ICU Severity Matrix
Physiological Vitals
Laboratory I
Laboratory II & Status
Severity Processor
Provide full physiological and laboratory data to calculate the complete APACHE II severity index.
Guidelines & Evidence
Verified
Last Review: 2026-07-17
When to Use
When to Use
Standardized scoring of physiological compromise in acute pancreatitis
To predict development of severe acute pancreatitis (SAP) and necrosis
To aid in triage for ICU level care within the first 24-48 hours of admission
Used in clinical research to compare baseline severity between cohorts
Serial measurements (admission, 24h, 48h) to track disease trajectory
Patient Population
Adults with acute pancreatitis. Originally an ICU general mortality score (1985), it remains the most rigorously validated physiological instrument in pancreatology. The APACHE II threshold of >=8 was included in the 1992 Atlanta Symposium definition of severe acute pancreatitis.
When Not to Rely on This Score Alone
First 12 hours - physiological parameters may still be fluid-responsive; the score is more stable at 24 and 48 hours
Simple triage from ED - scores like BISAP or SIRS are faster and often sufficient for initial ward vs. ICU decisions
Predicting infected necrosis - the score is a better marker of early organ failure than of late local complications
Elderly patients - the age component can artificially elevate scores independent of disease severity
How it Works
Complexity and Inputs
APACHE II incorporates 12 physiological variables (Temp, MAP, HR, RR, Oxygenation, pH, Na, K, Cr, Hct, WBC, GCS), age, and chronic health status. It is not point-based for easy bedside calculation and typically requires a calculator or automated system. The Glasgow Coma Scale and Creatinine components carry disproportionate weight in pancreatitis scoring.
Thresholds for Pancreatitis
01
Score < 8: Low predicted risk of severe disease; mortality < 4%. Excellent negative predictive value for severe AP.
02
Score >= 8: Classical threshold for Severe Acute Pancreatitis (SAP) per Atlanta Symposium. Mortality 11-18%. AUC 0.83-0.86 for predicting severe AP at admission.
03
Score > 13: High-risk phenotype; mortality rates approach 40-50%. Strongly associated with persistent organ failure.
04
Score >= 18 at 48h: Sensitivity 46%, Specificity 98% for mortality (Khan 2002).
48-Hour vs Admission Performance
| Admission AUC (severe AP) | 0.861 (Simoes 2011) - 0.834 (Kumar 2017) |
| 48h AUC (severe AP) | 0.892 (Simoes 2011) - best of all scores compared |
| Admission >=8 OR for severity | 11.07 (95% CI 4.14-29.61) |
| 48h >=8 OR for severity | 19.03 (95% CI 7.00-51.70) |
| Admission accuracy (mortality) | 75% (Khan 2002) |
| 48h accuracy (mortality) | 93% (Khan 2002) |
| AUC for pancreatic necrosis | 0.855 (Kumar 2017) |
| AUC for organ failure | 0.831 (Kumar 2017) |
| AUC for ICU admission | 0.885 (Kumar 2017) |
Dynamic Assessment
The most important indicator in pancreatitis is the dynamic change in score. Khan et al. (2002) found that deteriorating APACHE II scores over 48 hours were significantly associated with fatal outcome (P = .03). Non-survivors showed a median increase of 8 points, while survivors showed a median decrease of 1 point. The combined score (admission + 48h sum) was strongly associated with pancreatic necrosis (P = .001) and organ failure (P < .001).
Clinical Pearls
Predictive Performance vs Other Scores
In head-to-head comparisons, APACHE II (AUC 0.83-0.89) performs favorably against Ranson (AUC 0.75-0.88), BISAP (AUC 0.68-0.82), and Glasgow-Imrie (AUC 0.81). Kumar et al. (2017) found APACHE II was significantly superior to BISAP for predicting severe AP (P = .02) and comparable to the modified CT Severity Index (P = .13). Simoes et al. (2011) reported APACHE II at 48h had the highest AUC (0.892) of all scores evaluated in their cohort of 126 patients.
Sensitivity and NPV Profile
| Necrosis sensitivity (>=8) | 93.3% (Kumar 2017) |
| Necrosis NPV (>=8) | 96.2% (Kumar 2017) |
| Organ failure sensitivity (>=8) | 92.9% (Kumar 2017) |
| Organ failure NPV (>=8) | 96.2% (Kumar 2017) |
| ICU admission sensitivity (>=8) | 92.3% (Kumar 2017) |
| ICU admission NPV (>=8) | 95.8% (Kumar 2017) |
| Mortality sens at admission (>=10) | 31% (Khan 2002) |
| Mortality sens at 48h (>=18) | 46% (Khan 2002) |
| Mortality spec at 48h (>=18) | 98% (Khan 2002) |
APACHE-O (Obesity Modification)
The APACHE-O score adds BMI to the standard APACHE II, based on evidence that obesity (BMI > 30) increases both the risk and mortality of severe AP (OR 2.3 for severe AP). However, Lee and Cho (2022) report that APACHE-O did not demonstrate improved accuracy over standard APACHE II in validation studies.
Clinical Pearls
APACHE II score at 48h is more predictive than at 0h - always reassess
A low APACHE II (< 8) has an excellent NPV (96%) for severe disease - useful as a rule-out tool
Deteriorating score over 48h is an ominous sign (median +8 in non-survivors)
Integrate with Contrast-Enhanced CT (CECT) after 72h if score remains elevated to look for necrosis
APACHE II is comparable to modified CTSI for predicting severity (p = 0.13)
Next Steps
Management Pathways
01
Score >= 8: Aggressive fluid resuscitation (guided by haemodynamics), early enteral nutrition, and ICU/Step-down monitoring.
02
Persistent elevation at 48h: Khan et al. (2002) showed 48h score predicts outcome with 93% accuracy. Consider CECT to evaluate for pancreatic necrosis. Rule out abdominal compartment syndrome.
03
Deteriorating score at 48h: Median +8 in non-survivors (Khan 2002). Escalate monitoring, consider transfer to tertiary centre with interventional radiology and surgical capability.
04
Low score (< 8) with improving trend: Safe for ward-level care. High NPV (96%) effectively rules out severe disease progression.
Complementary Prognostic Tools
BISAP Score
Ranson Criteria
Balthazar Score (CT Severity)
Marshall Score (Organ Failure)
Modified CT Severity Index (MCTSI)
Key Evidence: Khan et al. 2002
In 125 patients with severe acute pancreatitis (85% SAP per Atlanta criteria), the 48-hour and combined (admission + 48h) APACHE II scores accurately predicted outcome in 93% of patients vs 75% by admission score alone. Patients with fatal outcomes showed a median increase of 8 points over 48 hours; survivors showed a median decrease of 1 point.
The Evidence
The Original Score
APACHE II: a severity of disease classification system.
Knaus WA et al. • Critical Care Medicine. 1985;13(10):818-29. The foundational ICU physiological score.
View SourceAPACHE-II score for assessment and monitoring of acute pancreatitis.
Larvin M et al. • Lancet. 1989;2(8656):201-5. Early validation specifically within the window of acute pancreatitis.
View Source48-Hour Superiority (Khan 2002)
Improved prediction of outcome in patients with severe acute pancreatitis by the APACHE II score at 48 hours after hospital admission compared with the APACHE II score at admission.
Khan AA et al. • Archives of Surgery. 2002;137(10):1136-40. 125 patients. 48h and combined scores predicted outcome with 93% accuracy vs 75% for admission score. PMID: 12361419.
View SourceComparison with BISAP, Ranson, MCTSI (Kumar 2017)
A comparison of APACHE II, BISAP, Ranson's score and modified CTSI in predicting the severity of acute pancreatitis based on the 2012 revised Atlanta Classification.
Kumar AH et al. • Gastroenterology Report. 2017;6(2):127-31. 50 patients. APACHE II AUC 0.834 for severe AP, 0.855 for necrosis, 0.831 for organ failure. Superior to BISAP (P=.02). PMCID: PMC5952961.
View SourceComprehensive Review (Lee and Cho 2022)
Predicting severity of acute pancreatitis.
Lee DW et al. • Medicina. 2022;58(6):787. Comprehensive review of scoring systems. APACHE II AUC 0.80-0.89. Mortality <4% for score <8, 11-18% for score >=8.
View SourcePortuguese Cohort Validation (Simoes 2011)
Predicting acute pancreatitis severity: comparison of prognostic scores.
Simoes M et al. • Gastroenterology Research. 2011;4(5):216-22. 126 patients. APACHE II at 48h had highest AUC (0.892). Admission OR 11.07; 48h OR 19.03 for severe AP.
View SourceOrigins & History
U.S. Critical Care History
Developed by William Knaus at George Washington University and published in 1985. It was the first scoring system to demonstrate that physiological derangement could be quantified into a reliable mortality probability across diverse disease states. The Atlanta Symposium (1992) formally incorporated APACHE II >= 8 as one of the criteria defining severe acute pancreatitis, cementing its role in pancreatology.
Paradigm Shift in Usage
Khan et al. (2002) proposed a paradigm shift: rather than using APACHE II to triage between mild and severe AP at admission, its greatest value lies in identifying the subgroup of patients with established severe pancreatitis who will develop poor outcomes. Serial measurement (admission, 24h, 48h) captures the dynamic nature of the disease - a deteriorating trajectory is more ominous than any single snapshot.
Last Comprehensive Review: 2026-07-17