For every maternal death, >100 women experience severe maternal morbidity, which is a life-threatening diagnosis, or undergo a life-saving procedure during their delivery hospitalization. Similar to racial/ethnic disparities in maternal death, black women are more likely to experience severe maternal morbidity than white women. Site of care has received attention as a mechanism to explain disparities in other areas of medicine. Data indicate that black women receive care in a concentrated set of hospitals and that these hospitals appear to provide lower quality of care. Whether racial differences in the site of delivery contribute to observed black-white disparities in severe maternal morbidity rates is unknown.

The purpose of this study was to determine whether hospitals with high proportions of black deliveries have higher severe maternal morbidity and whether such differences contribute to overall black-white disparities in severe maternal morbidity.

We used a published algorithm to identify cases of severe maternal morbidity during deliveries in the Nationwide Inpatient Sample of the Healthcare Cost and Utilization Project for 2010 and 2011. We ranked hospitals by their proportion of black deliveries into high black-serving (top 5%), medium black-serving (5% to 25% range), and low black-serving hospitals. We analyzed the risks of severe maternal morbidity for black and white women by hospital black-serving status using logistic regressions that were adjusted for patient characteristics, comorbidities, hospital characteristics, and within-hospital clustering. We then derived adjusted rates from these models.

Seventy-four percent of black deliveries occurred at high and medium black-serving hospitals. Overall, severe maternal morbidity occurred more frequently among black than white women (25.8 vs 11.8 per 1000 deliveries, respectively; P < .001); after adjustment for the distribution of patient characteristics and comorbidities, this differential declined but remained elevated (18.8 vs 13.3 per 1000 deliveries, respectively; P < .001). Women who delivered in high and medium black-serving hospitals had elevated rates of severe maternal morbidity rates compared with those in low black-serving hospitals in unadjusted (29.4 and 19.4 vs 12.2 per 1000 deliveries, respectively; P < .001) and adjusted analyses (17.3 and 16.5 vs 13.5 per 1000 deliveries, respectively; P < .001). Black women who delivered at high black-serving hospitals had the highest risk of poor outcomes.

Most black deliveries occur in a concentrated set of hospitals, and these hospitals have higher severe maternal morbidity rates. Targeting quality improvement efforts at these hospitals may improve care for all deliveries and disproportionately impact care for black women.