Maternal Conditions and Cesarean Delivery in a High-Burden Southeast Louisiana Birth Cohort: Associations and Limitations of Administrative Delivery Data

Background: Adverse perinatal outcomes impose a substantial burden on maternal health services, particularly in regions with a high prevalence of maternal medical conditions and marked health inequities. Population-specific evidence on how common maternal conditions relate to delivery outcomes remains limited for high-burden Southeast Louisiana. Methods: We conducted a retrospective cohort study of 7161 delivery encounters at an urban New Orleans hospital between 2020 and 2024, using electronic health record data from the LSU Health Clinical Data Warehouse. Multivariable logistic regression estimated adjusted associations of three non-overlapping exposure groups—gestational hypertension/preeclampsia (O13/O14), diabetes in pregnancy (predominantly gestational), and coded obesity—with cesarean delivery, ascertained from the recorded delivery-mode code and adjusted for age, race, ethnicity, and marital status; a composite of cesarean or preterm-labor coding was examined only as an exploratory secondary outcome. Because each encounter carried a single defined-condition code, the exposures were mutually exclusive, and their estimates describe separate exposure groups rather than the relative contributions of co-occurring conditions. Results: The cesarean rate ascertained from the delivery-mode code (11.9%) was well below the national benchmark (approximately 32%), indicating substantial administrative under-capture; such non-differential misclassification of the primary outcome would be expected to bias associations toward the null and limit the reliability of the effect estimates. Gestational hypertension/preeclampsia (adjusted odds ratio [aOR] 1.83, 95% confidence interval [CI] 1.53–2.19), diabetes (aOR 1.94, 95% CI 1.51–2.49), and coded obesity (aOR 1.41, 95% CI 1.05–1.88) were each associated with higher odds of cesarean delivery after adjustment for the available covariates. Conclusions: Because important clinical confounders (e.g., parity, prior cesarean delivery, multiple pregnancy, gestational age) were unavailable and the outcomes were not validated against clinical records, these associations should be interpreted as adjusted rather than independent. They are hypothesis-generating observations from administrative data rather than evidence for a risk-stratification tool, and they illustrate both the potential and the limits of routinely collected delivery data for monitoring maternal health services in high-burden settings.