Maternity Care Access and Infant Mortality.

Importance: Infant mortality, ie, death within the first year of life, serves as a critical health indicator. Objective: To evaluate the association of maternal residence in counties with no or limited access to maternity care with infant mortality overall and by maternal race and ethnicity and timing of death. Design, Setting, and Participants: This cross-sectional study used cohort-linked birth/infant death records for live births to US residents from 2017 to 2021, collected from the National Center for Health Statistics. Statistical analyses were conducted from July 2024 to June 2025. Exposures: Maternity care access by maternal county residence. Care access was categorized as none, low, moderate, or full based on availability of obstetric hospitals and birth centers, ratio of obstetric clinicians to births, and proportion of uninsured women aged 19 to 54 years. Main Outcome and Measures: The association between infant mortality and access to maternity care was assessed using multivariable log-binomial regression. Stratified analyses were conducted by timing of death (neonatal and postneonatal) and maternal race and ethnicity to assess for differences by subgroups. Results: A total of 18 682 916 live births were included (maternal age: 5 458 056 [29.2%] aged 30-34 years; maternal race and ethnicity: 4 426 077 [23.7%] Hispanic, 2 711 614 [14.5%] non-Hispanic Black, and 9 600 056 [51.4%] non-Hispanic White). Infant mortality rates (deaths per 1000 live births) increased as access decreased, with the highest rates in no-access counties (6.5 deaths per 10 000 live births) and the lowest in full-access counties (5.2 deaths per 1000 live births). The adjusted model found that infants in no-access counties had higher mortality risk compared with those in full-access counties (adjusted risk ratio, 1.14; 95% CI, 1.10-1.17; P < .001). When stratified by maternal race and ethnicity, non-Hispanic White infants in no-access counties had higher risk compared with non-Hispanic White infants in full-access counties (adjusted risk ratio, 1.20; 95% CI, 1.16-1.25). No significant differences were identified for other racial and ethnic groups. When stratified by timing of death, risk in the neonatal and postneonatal period was higher for infants in no-access counties compared with those in full-access counties (neonatal: adjusted risk ratio, 1.15; 95% CI, 1.11-1.19; postneonatal: adjusted risk ratio, 1.12; 95% CI, 1.06-1.17). Conclusions and Relevance: In this population-based cross-sectional study, infants in no-access counties had higher mortality risk than those with full access, persisting regardless of time of death. When examined by race and ethnicity, differences in mortality risk between living in a full access and no access county were observed among White infants only, indicating that unmeasured barriers may limit the protective effect of access for some racial and ethnic groups.