Birth hospital and racial and ethnic differences in severe maternal morbidity in the state of California
Mahasin S Mujahid,Peiyi Kan,Stephanie A Leonard,Elleni M Hailu,Elizabeth Wall-Wieler,Barbara Abrams,Elliott Main,Jochen Profit,Suzan L Carmichael
DOI: https://doi.org/10.1016/j.ajog.2020.08.017
Abstract:Background: Birth hospital has recently emerged as a potential key contributor to disparities in severe maternal morbidity, but investigations on its contribution to racial and ethnic differences remain limited. Objective: We leveraged statewide data from California to examine whether birth hospital explained racial and ethnic differences in severe maternal morbidity. Study design: This cohort study used data on all births at ≥20 weeks gestation in California (2007-2012). Severe maternal morbidity during birth hospitalization was measured using the Centers for Disease Control and Prevention index of having at least 1 of the 21 diagnoses and procedures (eg, eclampsia, blood transfusion, hysterectomy). Mixed-effects logistic regression models (ie, women nested within hospitals) were used to compare racial and ethnic differences in severe maternal morbidity before and after adjustment for maternal sociodemographic and pregnancy-related factors, comorbidities, and hospital characteristics. We also estimated the risk-standardized severe maternal morbidity rates for each hospital (N=245) and the percentage reduction in severe maternal morbidity if each group of racially and ethnically minoritized women gave birth at the same distribution of hospitals as non-Hispanic white women. Results: Of the 3,020,525 women who gave birth, 39,192 (1.3%) had severe maternal morbidity (2.1% Black; 1.3% US-born Hispanic; 1.3% foreign-born Hispanic; 1.3% Asian and Pacific Islander; 1.1% white; 1.6% American Indian and Alaska Native, and Mixed-race referred to as Other). Risk-standardized rates of severe maternal morbidity ranged from 0.3 to 4.0 per 100 births across hospitals. After adjusting for covariates, the odds of severe maternal morbidity were greater among nonwhite women than white women in a given hospital (Black: odds ratio, 1.25; 95% confidence interval, 1.19-1.31); US-born Hispanic: odds ratio, 1.25; 95% confidence interval, 1.20-1.29; foreign-born Hispanic: odds ratio, 1.17; 95% confidence interval, 1.11-1.24; Asian and Pacific Islander: odds ratio, 1.26; 95% confidence interval, 1.21-1.32; Other: odds ratio, 1.31; 95% confidence interval, 1.15-1.50). Among the studied hospital factors, only teaching status was associated with severe maternal morbidity in fully adjusted models. Although 33% of white women delivered in hospitals with the highest tertile of severe maternal morbidity rates compared with 53% of Black women, birth hospital only accounted for 7.8% of the differences in severe maternal morbidity comparing Black and white women and accounted for 16.1% to 24.2% of the differences for all other racial and ethnic groups. Conclusion: In California, excess odds of severe maternal morbidity among racially and ethnically minoritized women were not fully explained by birth hospital. Structural causes of racial and ethnic disparities in severe maternal morbidity may vary by region, which warrants further examination to inform effective policies.