Early Clinical Predictors for Adverse Maternal Outcome from Obstetric Hemorrhage

The development of early, accurate predictors for adverse maternal outcomes associated with obstetric hemorrhage may be useful as a trigger for initiating timely referral and resuscitation activities. Hemodynamic compensatory mechanisms mask abnormal values of traditional clinical indicators such as heart rate (HR) and blood pressure (BP) until women become severely compromised.   The purpose of this analysis was to evaluate the performance of alternative indicators of hemodynamic stability in predicting adverse maternal outcome for women entering hypovolemic shock secondary to obstetric hemorrhage.

Data from six studies of the non-pneumatic anti-shock garment (NASG) were compiled for 7,804 participants.  We evaluated the performance of several clinical signs at study entry: HR, systolic BP, pulse pressure, shock index (SI), mean arterial pressure, and modified shock index (MSI) in predicting an extreme adverse outcome, defined as death or severe maternal morbidity.  We estimated Receiver Operating Characteristic (ROC) curves for each combination of clinical sign and outcome, using semi-parametric standardization and generated area under the curve (AUC) estimates and their 95% confidence intervals. 

MSI and SI at study entry had the highest AUC values, at 0.630 (0.597-0.663) and 0.611 (0.580-0.642), respectively, which were significantly higher than the AUC of the other clinical indicators.  We considered a number of cut-point values as threshold values.

Our results suggest that shock index and modified shock index may be more accurate clinical predictors of adverse maternal outcome associated with obstetric hemorrhage.  These may have an important role within a clinical prediction system for indicating early need for resuscitation efforts.