Abstract

Purpose: To determine the accuracy of the admission APACHE II score for prediction of in-hospital mortality in acute pancreatitis (AP) using population-based data. In addition, to determine thresholds for increased mortality by APACHE II score. Methods: Patient data was generated from the Cardinal Health Research Database, a large population dataset that has supported publicly reported hospital performance in Pennsylvania for 20 years. All cases from Jan 2004 - Dec 2004 identified by principal diagnosis ICD9-CM 577.0 (AP) were included. Admission APACHE II scores were calculated for all patients. Missing values were entered as normal. Performance of admission APACHE II was measured using area under the ROC curve (AUC). Comparison of APACHE II score with observed mortality was performed to determine appropriate cutoffs for high vs. low-risk patient groups. Results: There were 18,256 cases from 212 hospitals with 225 deaths (1.2%). Mean age was 55.0 and 49.3% were males. Mean admission APACHE II score was 7.0 among all patients. Using our method, admission APACHE II AUC = .828. Observed mortality by Apache II score is depicted in Table 1.TABLE 1: Caption not available.Conclusion: Our application of the APACHE II scoring system produced an accurate tool for early prediction of mortality in patients with AP. Mortality in our AP population did not significantly increase until an APACHE II score >10.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call