Objective To investigate the acute physiology and chronic health evaluation (APACHE)-II for evaluating the severity of the illness in patients with cardiovascular disease and the ability to predict prognosis.

Methods Selected 1439 patients, who were admitted into the ICU for treatment in Fuwai Hospital from 2003 to 2008, calculated each patient's APACHE-II score and the risk of death (R value). In accordance with the APACHE-II scores (<10, 10–20, > 20) and the type of ICU common diseases (congenital heart disease; coronary heart disease; valvular heart disease; cardiomyopathy; pulmonary heart disease), all the patients were divided into groups to analyse the consistency of actual mortality and predicted mortality and the relationship between actual mortality and APACHE-II scores. Compared the overall scores and each section of the non-survival group and survival group of APACHE-II scores.

Results 1439 cases of patients, the actual death case is 122, male 77, female 45. Predicted mortality and actual mortality both are related to APACHE-II scores, they increased along with APACHE-II scores, the mortality of >20 group is 45.5%. <10 group APACHE-II score, non-survival group (7.38±1.389) and survival group (7.44±1.464) were not significantly different (p>0.05). In 10 to 20 group, >20 group and the overall, APACHE-II score of non-survival was significantly higher than the survival group (p<0.05). APACHE-II method to the mortality rate is expected to determine the individual prognosis, area under the ROC curve is 0.689, low diagnostic value. Compared actual mortality and predicted mortality, divided by score, only >20 group predicted mortality is in the actual mortality rate 95% CI, 10 to 20 group, >20 group, there are significant differences between the predicted mortality and actual mortality. Divided by the disease, only the forecasting mortality of pulmonary heart disease group is in the actual mortality 95% CI. In remaining groups, there is significant difference.

Conclusion APACHE-II scores is related to the disease critical levels, the higher the score the higher the risk of death, a significant increase in mortality in >20 groups, clinicians should arouse great attention. In the high score (>20 points), APACHE-II predicted mortality and actual mortality are similar. The prognosis of this part patient with cardiovascular disease have a certain value. But on the whole, APACHE-II predicted mortality for cardiovascular disease is not very well. Particularly in low scores (<10 points, 10–20 points), predicted mortality and actual mortality were significantly different, and in <10 group APACHE-II scores of the non-survival group and survival group were not significantly different, which the APACHE-II score is not related to death in <10 group. In conclusion, APACHE-II scores indeed, to some extent, has a rough evaluation of the critical level of cardiovascular disease, especially in the high score section, but for mortality prediction, the value of their diagnosis is low.