AUTHORS
Aaron N. Cheng, MD, Kamyar Arasteh, PhD2, Andrea H. Roe, M.D., MPH3, Corinna Schultz, MD4, Stephanie Guarino, MD MSHP5, Sophie Lanzkron, MD, MHS6
AFFILIATIONS
1 Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania, United States
2 Nemours Children's Hospital, Wilmington, Delaware, United States
3 Penn Medicine, Philadelphia, Pennsylvania, United States
4 Nemours Children's Hospital, Wilmington, Delaware, United States
5 Nemours Children's Hospital, Wilmington, Delaware, United States
6 Thomas Jefferson University, Philadelphia, Pennsylvania, United States
ABSTRACT
Introduction
Pregnancy in sickle cell disease (SCD) is associated with substantially higher risk of severe maternal morbidity (SMM) compared to the general population. While specialized multidisciplinary care is recommended for individuals with SCD, the impact of healthcare access and delivery location on maternal outcomes has not yet been formally investigated. In Pennsylvania (PA), access to maternity services and specialized SCD care varies considerably by geography, raising the concern that structural barriers to care may contribute to adverse pregnancy outcomes. We evaluated the association between measures of healthcare access and SMM among pregnant individuals with SCD.
Methods
We performed a retrospective study using PA inpatient discharge data (available through the PA Health Care Cost Containment Council) to identify delivery hospitalizations, using a previously published ICD code-based algorithm (Kulina et al, Matern Child Health J, 2008). The primary outcome was SMM as defined by the Centers for Disease Control and Prevention, excluding transfusion and vaso-occlusive crisis (VOC) from the composite outcome. VOC during delivery hospitalization was evaluated separately as a secondary outcome.
We performed mixed-effects logistic regression models to estimate the association between SCD and SMM, while adjusting for maternal age, delivery year, geographic region, Medicaid status, and admission source. Random effects accounted for clustering by hospital and by individual patient. Furthermore, we evaluated whether delivery at a National Alliance of Sickle Cell Centers (NASCC)-recognized hospital, distance from the patient’s residence to the delivery hospital, and residence within a maternity care desert (defined as counties without labor and delivery services, according to the Center for Rural PA) were independently associated with SMM among individuals with SCD.
Results
From 1990 to 2024, the overall incidence of SMM increased across delivery hospitalizations in PA. Compared with non-Black controls, Black individuals without SCD had higher adjusted odds of SMM (adjusted odds ratio [aOR] 1.37, 95% CI: 1.32-1.42, p<0.001), while individuals with SCD experienced markedly higher odds of SMM (aOR 10.47, 95% CI: 8.07-13.58, p<0.001). Among delivery hospitalizations in individuals with SCD, delivery at an NASCC-recognized center was not significantly associated with SMM (aOR 0.82, 95% CI: 0.52-1.27, p=0.373). However, in those with SCD, increasing travel distance between the patient’s home to the delivery hospital was independently associated with higher odds of SMM (aOR 1.15 per 10-mile increase, 95% CI: 1.0-1.31, p=0.049). Among all patients, residence within a maternity care desert was associated with increased odds of SMM (aOR 1.07, 95% CI: 1.02-1.12, p=0.004); however, analysis restricted to SCD did not identify a statistically significant association. Overall, 874 of 2,754 delivery hospitalizations (31.7%) among individuals with SCD were complicated by VOC. Maternal mortality occurred in 5 of 2,754 delivery hospitalizations among people with SCD (18 per 10,000) compared to 530 of 5,023,986 delivery hospitalizations among individuals without SCD (1.05 per 10,000), corresponding to a crude risk ratio of 17.21 (95% CI: 7.14-41.49, Fisher’s exact p<0.001).
Conclusion
In this statewide analysis including over 30 years of hospital discharge data, individuals with SCD experienced over ten-fold higher adjusted odds of SMM during delivery hospitalization compared with non-Black controls. For those with SCD, longer travel distance to the delivery hospital was independently associated with worse maternal outcomes. Among all patients, residence in a maternity care desert was associated with SMM. Delivery at centers currently recognized by NASCC was not associated with reduced SMM; however, this should be interpreted cautiously because NASCC recognition was established only during the final years of the study period. The association may also be confounded by referral of higher-risk pregnancies to specialized centers. Together, these results demonstrate that healthcare access is a potentially modifiable determinant of maternal outcomes. As access to specialized hematologic and obstetric care varies considerably across the United States, this analytic framework may be applied to guide targeted health system interventions. Our results support the development of innovative care models to ensure broad access to specialty expertise.
DISCLOSURES
Aaron N. Cheng, MD: Agios Pharmaceuticals - Consultancy (Includes expert testimony)