Predictors of Outcome in Obstetric Patients Admitted to Intensive Care Unit (ICU). Utility of Sequential Organ Failure Assessment (SOFA) Score

Abstract

PURPOSE: Obstetric admissions to ICU is an important measure of maternal morbidity, but has not been investigated in detail. Identifying criteria and scoring systems may help develop guidelines for better management of patients, better utilization of existing resources and reduce cost burdens. This study evaluated clinical maternal characteristics and SOFA score in predicting outcome of obstetric patients admitted to ICU METHODS: 38 obstetric patients with severe illness admitted to ICU of a tertiary care hospital between October 2010 and April 2011 were followed till death/ discharge from ICU in ongoing prospective consecutive follow-up study. Their clinical features, laboratory parameters, disease course and SOFA scores were evaluated RESULTS: 10/38 (26.3%) died. The non survivors were young (20-30yrs), multiparous women from lower socio-economic status with poor antenatal supervision. The indications for admission were: eclampsia/ pre-eclampsia(50%), hemorrhage/ coagulopathy(20%), sepsis (10%) and associated medical conditions (20%). Majority (60%) had a SOFA score between 9-12. Mean SOFA score in nonsurvivors was 12.6 vs 5.5 in survivors (p< 0.001). A critical cut-off score of 9 had 90% sensitivity and 96% specificity in predicting mortality in obstetric patients (AUC= 0.975). 9 out of 9 women admitted for elective ventilation and 11 out of 13 admitted with hemorrhage and/or coagulopathy survived CONCLUSIONS: SOFA score >9 at admission predicts a very high risk of mortality in obstetric patients admitted to ICU. Survival prospects are better in women admitted for elective ventilation and those with hemorrhage or coagulopathy CLINICAL IMPLICATIONS: SOFA score has a good applicability in predicting mortality in obstetric population and should be used in ICU. Universal antenatal care and efforts to reduce eclampsia/precclampsia can bring down need for ICU care in pregnant population.