Sickness scoring and response to treatment as predictors of outcome from critical illness

Abstract

A physiological sickness scoring system (SS), based on the APACHE II score, has been used to assess outcome from critical illness in 128 patients admitted to a general intensive care unit. Physiological data were collected on each patient from admission until death or discharge from the unit, and survival was recorded as survival to home. The admission SS correctly classified 80.6% of survivors, and 70.4% of non-survivors. Predictive power did not improve with time using the daily SS. However, when the proportional change in SS over time was included in the analysis, predictive power improved; and at day 4, 87.1% of survivors and 75% of non-survivors were correctly classified. At intermediate levels of sickness severity (admission score of 13-18), a reduction in SS of 30% by day 4 reduced the risk of death by 32%; at higher levels (greater than 18) a similar reduction in SS was associated with a 47% reduction in death-risk. Failure to obtain a reduction in score by day 4 was associated with increased risk of death. Survivors consistently showed a greater fall in SS by day 4 than non-survivors. The APACHE score and its modifications provide an accurate, unitary measure of physiological disturbance. Correction of abnormal physiology, and the measurement of responsiveness to therapy are important components in the prediction of outcome from critical illness.