RÉSUMÉ
BACKGROUND: Critical care in obstetrics has received much attention in recent times. Despite progress in medical field and improvement in health facilities provided, maternal mortality is still very high in most of the developing countries. AIM: To study and analyze records of patients requiring intensive care in obstetrics and to assess utility of simplified acute physiology score (SAPS II) for predicting maternal mortality. SETTING: A multidisciplinary intensive care unit (ICU) at a tertiary care center. DESIGN: Retrospective review. MATERIALS AND METHODS: Fifty-seven consecutive obstetric patients' records requiring ICU admissions were studied for clinical picture, diagnosis, complications, morbidity and mortality over a period of 21/2 years - from 1st May 2002 to 31st Oct. 2004. SAPS II score was calculated according to the different variables for predicting mortality. STATISTICAL ANALYSIS: SAPS II scores were regressed on mortality status using logistic regression analysis. The predictability was assessed by goodness-of-fit test and receiver operated characteristic curve. RESULTS: Maternal mortality in obstetric ICU admissions was 1.15/1,000 deliveries, amounting to 40.35% of obstetric ICU admissions. The mean SAP II score was significantly higher (40.04 +/- 12.97 vs. 22.6 +/- 7.31) in those patients who died compared to survivors (P < 0.001%). CONCLUSIONS: The SAPS II accurately predicted mortality in obstetric patients admitted to ICU. Computation of the score as a routine in ICU may help in identifying those at high risk of mortality and then to reduce this risk.