Abstract Introduction: Pregnancies complicated by sickle cell disease (SCD) are associated with significant maternal and fetal risks. While prior studies have identified these risks, many are from single centers or older cohorts. A contemporary, large-scale analysis is needed to accurately quantify the burden of adverse outcomes and healthcare utilization in this population, which is crucial for guiding clinical practice and health policy. This study aimed to evaluate a comprehensive set of maternal morbidity, delivery, and economic outcomes associated with SCD in a nationally representative sample of inpatient delivery admissions. Methods: We performed a retrospective cohort study using data from the 2017-2022 National Inpatient Sample (NIS), the largest all-payer inpatient database in the United States. Our cohort included all delivery hospitalizations for women aged 15-50, identified using a combination of Diagnosis-Related Group (DRG) and International Classification of Diseases, Tenth Revision (ICD-10) codes for delivery. Hospitalizations for multiple gestations (e.g., O30, O31) and abortive outcomes (e.g., O02, O03) were excluded due to complications associated with multiple gestations and abortions. The primary exposure, SCD, was identified by the presence of any ICD-10 diagnosis code for D57. To create a comparable control group, we employed 1:4 nearest-neighbor propensity score matching without replacement. Propensity scores, representing the probability of having SCD, were estimated using a logistic regression model that included maternal age category, race, primary insurance payer, and pre-existing comorbidities (obesity, diabetes, chronic hypertension), as well as hospital characteristics (teaching status, bed size, and geographic region). A caliper of 0.05 standard deviations of the logit of the propensity score was used to ensure high-quality matches, and only patients within the region of common support were included. Post-matching covariate balance was rigorously assessed using standardized mean differences, with a value 0.1 considered indicative of good balance. Primary outcomes included preeclampsia/eclampsia (O14, O15), cesarean delivery (O82), severe maternal morbidity (SMM, based on a composite of diagnoses including eclampsia, shock, and others), and blood transfusion. Secondary outcomes were length of stay (LOS) and total hospital charges. All outcomes were compared between the matched SCD and non-SCD cohorts using complex survey-weighted regression models to account for the NIS sampling design. Survey-weighted logistic regression was used for binary outcomes, yielding Odds Ratios (ORs). For the skewed continuous outcomes of LOS and charges, we used generalized linear models with a gamma distribution and log link to generate Rate Ratios (RRs). Results: The initial cohort comprised 4,184,870 delivery hospitalizations, of which 29,551 (0.7%) were to women with SCD. After propensity score matching, the final cohort included 29,551 SCD deliveries and 118,204 matched non-SCD deliveries. The matching process achieved excellent balance across all measured covariates. In adjusted analyses, pregnancies with SCD were associated with a significantly higher odds of preeclampsia/eclampsia (OR 1.13, 95% Confidence Interval CI 1.08-1.19; p0.001), cesarean delivery (OR 1.10, 95% CI 1.07-1.13; p0.001), SMM (OR 1.18, 95% CI 1.14-1.21; p0.001), and blood transfusion (OR 1.95, 95% CI 1.80-2.11; p0.001). Furthermore, SCD was associated with significantly longer hospital stays (RR 1.08, 95% CI 1.07-1.10; p0.001) and higher total hospital charges (RR 1.06, 95% CI 1.04-1.08; p0.001). Conclusions: In this propensity score-matched national analysis, sickle cell disease remains an independent and significant risk factor for a wide range of adverse maternal outcomes and increased healthcare utilization during delivery hospitalization. These data confirm the substantial burden of SCD in pregnancy and highlight the critical need for specialized, multidisciplinary care, vigilant monitoring, and effective pre-conception counseling to optimize outcomes for this vulnerable population.
Richard Mansour (Mon,) studied this question.