Background: Evidence demonstrating the association between social determinants of health and infant outcomes do not support causal conclusions. I aimed to characterize trends in ICD social code documentation, examine the association between maternal social risk factors and infant outcomes, and collect qualitative data to interpret quantitative trends.
Methods: Quantitative data was collected from a retrospective study of live-born births in California, 2007-2020. ICD codes were used to classify maternal homelessness, inadequate housing, food insecurity, and unemployment. Infant outcomes included prematurity, low birthweight, small for gestational age, NICU admission, emergency department admission, rehospitalization and death. I estimated risk ratios using log-linear Poisson regression adjusted for maternal race, payer, and education. Nine individual interviews with obstetric providers explored experiences screening and documenting social risks. Qualitative data was analyzed using an iterative and inductive thematic analysis guided by grounded theory.
Results: 6,089,327 pregnant person-infant pairs were included in analyses, 4,002 (65.7 per 100,000 live births) of whom had a documented health-related social need (HRSN). Infants with a maternal HRSN had a higher risk of prematurity (aRR 2.7, 95%CI 2.5, 2.9), low birthweight (aRR 2.7, 95%CI 2.5, 2.9), SGA (aRR 1.5, 95%CI 1.4, 1.6), NICU admission (aRR 3.5, 95%CI 3.2, 3.8), and infant death (aRR 3.0, 95%CI 2.5, 5.9). Screening and documentation practices were influence by institution-specific workflows and the presence of ancillary staff.
Conclusions: HRSNs are poorly documented and likely underestimate HRSN prevalence. Among those for whom documentation exists, infants with a maternal HRSN have an increased risk of morbidity and mortality.
