Mortality Risk Prediction With an Updated Acute Physiology and Chronic Health Evaluation II Score in Critically Ill Obstetric Patients

Abstract

Background: Acute Physiology and Chronic Health Evaluation II (APACHE II) score has shown low prognostic ability to predict death in the obstetric population. The objective of this study was to evaluate whether an updated form of the APACHE II score would perform better in predicting mortality in critically ill obstetric patients. Methods: A retrospective cohort study of pregnant and postpartum women (up to 42 days after delivery) who were admitted to the intensive care unit (ICU) was carried out at an ICU at Rafael Calvo Maternity Hospital, a large obstetric hospital in Colombia. Data were collected on consecutive obstetric patients admitted to the ICU between 2006 and 2011. A discrimination and calibration analysis was completed on the original APACHE II score and an updated APACHE II score. Results: Data were collected on a total of 726 obstetric patients. The area under the receiver–operating characteristic curve was 0.86 (95% confidence interval [95% CI], 0.80-0.93) for both APACHE II and the updated APACHE II scores. Mortality ratio for the original APACHE II was 0.30 (95% CI, 0.19-0.41) and 0.85 (95% CI, 0.56-1.24) for the updated APACHE II. Conclusions: The APACHE II overestimates mortality in the sample population. The updated APACHE II model predicts mortality more accurately in the obstetric population. This formula may be useful in adapting the existing APACHE II to current mortality risk in obstetric critical care populations.