Search for a command to run...
Abstract Background: Women with von Willebrand disease (VWD) are at increased risk for postpartum hemorrhage (PPH), yet data on bleeding outcomes and hemostatic management in the United States remain limited. While prophylactic therapy (e.g., desmopressin or von Willebrand factor [VWF] concentrates) is recommended for individuals with low VWF levels, practice varies nationally, and the effectiveness of these strategies in routine care remains unclear. Moreover, systemic inequities, such as underdiagnosis and unequal access to specialty care, may also influence outcomes and treatment patterns. We aimed to characterize PPH risk, hospital utilization, and maternal outcomes by VWD status and treatment type, while examining the impact of sociodemographic context on claims-based risk modeling. Methods: We conducted a retrospective cohort study using a U.S. commercial insurance claims database (2011–2021) to evaluate the association between VWD and PPH risk. Delivery encounters were identified using inpatient claims, and VWD diagnosis was based on ICD codes. Logistic regression assessed odds of PPH and secondary outcomes, including length of stay above the median, 30-day ED visit or readmission, and in-hospital mortality, comparing women with and without VWD. Analyses were performed in both the full cohort and a 1:4 propensity-matched cohort matched on age band, delivery year/month, obesity, pregnancy type, and geozip. Among women with VWD, outcomes were also evaluated by treatment type: no therapy, desmopressin, or VWF concentrates. Results: Among 1,498,768 deliveries, 1,514 involved women with VWD. Compared to those without VWD, these patients had significantly higher odds of PPH (OR 1.70, 95% CI 1.38–2.10), prolonged length of stay (OR 1.58), and 30-day ED visit or readmission (OR 1.20). Mortality was rare and not significantly different. Among women with VWD, treatment with VWF concentrates was associated with higher odds of PPH compared to no treatment (OR 3.17, 95% CI 1.28–7.87), suggesting potential confounding by indication. Desmopressin showed a non-significant trend toward increased risk (OR 3.32, 95% CI 0.93–11.82). No significant differences were observed in readmission or mortality across treatment groups. In exploratory models including geozip-level racial and ethnic composition, structural context influenced outcomes. Women residing in <25% White geozips had lower odds of PPH (OR 0.89) but higher odds of maternal mortality (OR 3.97) and prolonged hospitalization (OR 1.89). These variables were more predictive of adverse outcomes than VWD itself in some models, leading to their exclusion from final regressions. Conclusions: Women with VWD face elevated risks of PPH and hospitalization in the peripartum period. These findings confirm prior literature while quantifying risks in a contemporary U.S. population. Among treated patients, those receiving VWF concentrates had the highest odds of PPH, likely reflecting triage of individuals perceived to be at elevated hemorrhagic risk. This underscores the challenge of disentangling treatment effect from disease severity in observational data. While desmopressin showed a similar trend, its interpretation is limited by confounding and lack of laboratory data, such as VWF levels or DDAVP response testing. Exploratory models showed that social context and geography, particularly the racial composition of a patient's community, were strongly associated with maternal outcomes. These patterns suggest structural factors may contribute to disparities in bleeding outcomes, potentially reflecting variation in diagnosis, prophylaxis, or recognition of hemorrhage. That patients in predominantly non-White geozips had lower odds of PPH but higher odds of mortality and extended hospitalization raises concern for underdiagnosis or delayed care, though further study is needed to confirm these mechanisms. Race and ethnicity were removed from final models to reduce confounding, but these patterns underscore the need for richer, integrated data sources to evaluate the role of structural racism and unequal access to specialized care. Our findings support prospective studies to define the effectiveness of peripartum treatments in VWD and to disentangle treatment effects from underlying risk. Data systems that integrate laboratory results, clinical decision-making, and social determinants of health will be critical to advancing both precision and equity in maternal outcomes.