APACHE II Score Calculator

APACHE II expresses how sick a critically ill patient is as a single integer built from three parts: an Acute Physiology Score from twelve variables, age points from 0 to 6, and chronic health points from 0 to 5, summing to a theoretical maximum of 71 although totals above 40 are rare. The physiologic variables are temperature, mean arterial pressure, heart rate, respiratory rate, oxygenation (alveolar-arterial gradient when FiO2 is at least 0.5, otherwise PaO2), arterial pH, serum sodium, potassium and creatinine, hematocrit, white cell count, and a neurological term calculated as 15 minus the Glasgow Coma Scale. Each is scored 0 to 4 according to how far it deviates from a physiologic reference range in either direction, which is why both hypothermia and fever, both bradycardia and tachycardia, attract points. Creatinine points double in acute renal failure. Values are the worst recorded in the first 24 hours after ICU admission, and any variable not measured is assumed normal. Approximate hospital mortality runs from about 4 percent at scores of 0 to 4 up to around 85 percent above 35. Several limitations follow from the design. Treatment delivered during that first 24 hours alters the very values being scored, so effective early resuscitation lowers the score and flatters the outcome. Patients transferred from another ICU carry lead-time bias. The full published mortality equation also requires a diagnostic category coefficient and an emergency surgery term that a score alone omits. Most importantly, it was built for comparing groups, not for prognosticating in an individual or informing withdrawal of treatment.

APACHE originated with William A. Knaus and colleagues at George Washington University Medical Center, with the first version published in Critical Care Medicine in 1981. The motivating question was administrative rather than clinical: intensive care units reported widely differing mortality rates, and there was no way to tell whether a unit with poor outcomes was treating sicker patients or delivering worse care. Any fair comparison required a severity measure to adjust for case mix. That first version used 34 physiologic variables and proved too laborious for routine data collection. APACHE II, published by Knaus, Elizabeth Draper, Douglas Wagner and Jack Zimmerman in Critical Care Medicine in 1985, cut the variable list to twelve using data from 5,815 ICU admissions across 13 United States hospitals, and became one of the most heavily cited papers in critical care medicine. Successors followed but did not displace it. APACHE III appeared in Chest in 1991, derived from 17,440 admissions at 40 hospitals with 17 variables, and APACHE IV was published by Zimmerman and colleagues in 2006 using data from 2002 and 2003. Both are proprietary and require licensing, which is the practical reason APACHE II remains the version most often reported in the literature despite reflecting ICU practice from the early 1980s. It sits alongside SAPS II, published by Jean-Roger Le Gall and coworkers in 1993, and the SOFA score introduced by Jean-Louis Vincent in 1996, which tracks organ dysfunction over time rather than severity at a single admission point.