A comparison of severity of illness scoring systems for critically ill obstetric patients

Study objective: To evaluate the predictive ability of three scoring systems, acute physiology and chronic health evaluation (APACHE II), simplified acute physiology score (SAPS II), and mortality probability models (MPM II) in critically ill obstetric patients compared to a control group of nonobstetric female patients of similar age group (range, 17 to 41 years). Design: A retrospective medical chart review of obstetric and nonobstetric female patients between 17 and 41 years of age. Setting: Two university hospitals. Patients: Ninety-three obstetric patients and 96 nonobstetric female patients were identified from 12,740 consecutive ICU admissions. Results: The actual mortality of the obstetric and the nonobstetric group was 10.8% (95% confidence interval [CI], 5.3 to 19.0%) and 12.5% (95% CI, 6.6 to 21.0%), respectively. The observed mortality was not statistically different from the mortality predicted by APACHE II, SAPS II, and MPM II (14.7%, 7.8%, and 9.1% for the obstetric group and 10.9%, 9.0%, and 9.9% for the nonobstetric group). Predictive accuracy was assessed by the c-index, which is equivalent to the area under the receiver operator characteristic (ROC) curve. There were no significant differences in the c-index for APACHE II, SAPS II, and MPM II within or between the obstetric group ([mean±SE], 0.93±0.02, 0.90±0.04, and 0.91±0.04, respectively) and the nonobstetric group (0.97±0.02, 0.95±0.03, and 0.96±0.02 respectively). Conclusions: We conclude that APACHE II, SAPS II, and MPM II assess the ICU outcome of critically ill obstetric patients as accurately as nonobstetric critically ill female patients of similar age group.