Ability of shock index and heart rate to predict the percentage of body blood volume lost after vaginal delivery as an indicator of severity: results from a prospective cohort study

BACKGROUND: Postpartum hemorrhage (PPH) is the leading cause of maternal mortality worldwide, but it mainly affects women from low- and middle-income countries. Despite being a treatable condition, the high number of maternal deaths resulting from PPH is outstanding for at least 25 years. Late diagnosis and difficulties in identifying women who will develop severe postpartum bleeding can, in part, explain the high incidence of PPH. Over the past few years, researchers have focused on identifying a simple, accessible and low-cost diagnostic tool that could be applied to avoid maternal deaths. In particular, it has been suggested that vital signs and shock index (SI) could be useful. The objective of this study was to evaluate whether vital signs are correlated with the percentage of body blood volume (BBVp) lost after vaginal delivery. METHODS: A prospective cohort study was performed at the Women’s Hospital of UNICAMP, Brazil. The inclusion criteria were women delivering vaginally who did not suffer from hypertension, hyper- or hypothyroidism, cardiac disease, infections or coagulopathy. Blood loss was measured over 24 hours using a calibrated drape and by weighing compresses, gauzes and pads. Vital signs were measured up to 24 hours after delivery. We evaluated the BBVp lost, and generated a Receiver operating characteristics (ROC) curve with area under the curve (AUC) analysis to determine the cut-off values for vital signs to determine the likelihood of postpartum bleeding above the 90(th) percentile within 24 hours of delivery. RESULTS: A total of 270 women were included. The mean blood loss within 24 hours of vaginal delivery was 570.66 ± 360.04 mL. In the first 40 minutes, 73% of the total blood loss over the 24-hour period had occurred, and within 2 hours, 91% of women had bled 90% of the total blood loss. Changes in SI and heart rate (HR) were statistically significant in predicting postpartum bleeding (P ≤ 0.05). Higher values for likelihood ratio (LR) to identify BBVp loss above the 90(th) percentile within 2 hours were a SI above 1.04 at 41-60 minutes after birth (LR = +11.84) and a HR above 105.2 bpm at 21-40 minutes after birth (LR = +4.96). Both measures showed high specificity but low sensitivity. CONCLUSION: Values of SI and HR are statistically significant in predicting postpartum bleeding with high specificity but low sensitivity. The cut-off points were 1.04 for SI and 105 bpm for HR.