How this calculator works
Twelve acute physiology variables contribute up to 4 points each, but the variables are scored from the worst value in the first 24 hours of intensive care, not from the current value alone. The Glasgow Coma Scale contributes 15 minus the recorded total. Age adds 0 to 6 points. A history of severe organ insufficiency or immunocompromise adds 2 points after elective surgery and 5 points otherwise. Creatinine points double in acute renal failure. The total is a severity description, not an individual prognosis, and the original mortality estimates are not reproduced here. The result is an estimate for information only and is not medical advice.
Formula
APACHE II = acute physiology score + age points + chronic health points. Acute physiology score = the twelve variable points, including 15 minus the Glasgow Coma Scale total. Age points: 0 under 45, 2 at 45 to 54, 3 at 55 to 64, 5 at 65 to 74 and 6 at 75 or older.
Worked example
A 70-year-old with temperature 39.5 C, mean arterial pressure 65 mmHg, heart rate 125, respiratory rate 30, PaO2 65 mmHg, pH 7.30, sodium 155, potassium 6.2, creatinine 2.5 mg/dL, haematocrit 50%, white cells 20 x 10^3/uL and Glasgow Coma Scale 12 scores 26 acute physiology points. Age adds 5 and there are no chronic health points, giving 31.
Advertisement
Frequently asked questions
Which measurements belong in the score?
Use the worst value for each variable during the first 24 hours in intensive care. Mixing times or using the best value understates the score.
How does acute renal failure change the creatinine points?
The published rule doubles the creatinine points in acute renal failure. Chronic renal failure does not trigger the doubling in the original table.
Why does a lower Glasgow Coma Scale add more points?
The nervous system row is scored as 15 minus the Glasgow Coma Scale total, so a lower score contributes more points.
Is the total a predicted mortality?
No. This page returns the score only. Published mortality estimates depend on the diagnosis group and the validation cohort, and are not reproduced.