The Influence of Maternal Race and Ethnicity on the Risk of Postpartum Hemorrhage
Abstract Number: S-7
Abstract Type: Original Research
Introduction: Racial/ethnic disparities in obstetric outcomes in the United States are some of the most profound in medicine. They have been documented in preterm delivery, hypertensive disorders, intrauterine growth restriction, intrauterine fetal demise, and maternal mortality; in general, blacks are disproportionally affected. Racial/ethnic disparities in postpartum hemorrhage (PPH) have been less intensively studied. In this investigation, we examined racial/ethnic disparities in the risk of the most common form of PPH, uterine atony. We employed a residual direct effect methodology to determine the extent these disparities can be explained by patient demographics/ comorbidities and hospital level factors.
Methods: This analysis was based on the Nationwide Inpatient Sample, a cohort of 2,488,974 birth hospitalizations from between 2005 and 2008. The frequencies of atonic PPH and atonic PPH requiring blood transfusion during this 4-year period were determined. We developed multivariable logistic regression models to estimate the relative risk of these outcomes in maternal racial/ethnic groups by sequentially adding potential mediators including maternal socio-demographic characteristics, obstetric factors, patient comorbidities, and hospital-level characteristics.
Results: The overall frequency of atonic PPH was 2.14%; atonic PPH resulting in transfusion occurred in 0.24% of deliveries. Hispanic ethnicity and Asian/Pacific Islander race were associated with a statistically significant increased risk of atonic PPH (odds ratio (OR) 1.25, 95% confidence interval [1.23-1.28] and 1.28 [1.24-1.33], respectively, with whites as referent). After adjustment for all potential mediators the direct effect estimate was little changed from the unadjusted measure (adjusted OR for Hispanics 1.22 [1.19, 1.24] and for Asian/Pacific Islanders, 1.31 [1.26, 1.36]). Similar results were observed for Hispanics and Asian/Pacific Islanders for atonic PPH with transfusion. The risk of atonic PPH in Blacks was similar to whites (OR 0.99 [0.97-1.02]); this effect estimate was also little changed with adjustment for mediators. Black race was associated with a significantly increased risk of atonic PPH with transfusion compared to whites (OR 1.59 [1.47-1.71] and there was significant attenuation of this effect with adjustment for mediators (aOR 1.18 [1.09-1.28]).
Conclusion: Hispanic ethnicity and Asian/Pacific Islander race are significant risk factors for atonic PPH independent of measured potential mediators; biological differences may play a role. Blacks are at increased risk of atonic PPH resulting in transfusion; patient and hospital characteristics mediate much of this increased risk.