modified apache ii score for predicting mortality of variceal bleeding

ABSTRACT

Prediction of outcome of based on personal experience is subjective. There is a need for a method of prediction to suit our current state of practice. To evaluate a modified APACHE II system in predicting mortality in acute variceal bleeding. Prospective study conducted from June 2005 through September 2006 in the Ibn Sina Hospital Bleeding Centre ICU. Clinical and laboratory data for 102 patients were allotted a modified APACHE II score, a child-Pugh class and fed to SPSS to calculate predicted mortality. The mean [ +/- SD] age 49.2[ +/- 16.1] years. Male were 79[77.5%]. Patients with liver cirrhosis and periportal fibrosis were 75 [73.5%] and 27 [26.5%]] respectively. The mean predicted mortality was 22.8% while the actual mortality was 26.5% [27] patients. The mean modified APACHE II score among the deceased group was 19.04. Rate of deaths among liver cirrhosis patients was 12 [44.4%] with mean predicted mortality of 27.3%. Deaths among periportal fibrosis occur at mean modified APACHE II score 17.7 and mean predicted mortality of 20%. Seventeen patients had ICU stay >/= 4 days. Ten out of them died [58.8%]. Modified APACHE II score is effective in predicting outcome of patients with variceal bleeding. Score of >/= 15 points and long ICU stay are associated with high mortality