(164) Examining the role of neighborhood level indices in predicting spontaneous preterm birth

Objective: Cervical length is an established but limited predictor of spontaneous preterm birth (sPTB). We examined whether measures of the neighborhood environment, which reflect structural determinants of pregnancy outcomes, improve prediction of sPTB as compared to CL alone.

Study Design: This was a retrospective cohort study of 1000 patients and each of their gestations with: 1) delivery at a tertiary, urban academic hospital between January 2013-August 2023 and 2) at least one transvaginal CL at 16w0d-24w6d gestational age (GA). The primary outcome was sPTB, defined from chart review as delivery prior to 37w0d GA. CL was abstracted from ultrasound images. Medical history, GA at delivery, progesterone therapy, and cerclage placement were abstracted from the electronic medical record. We ascertained neighborhood structural and social determinants of health (SDOH) from the Index of Concentration at the Extremes (indicator of racial-economic segregation) and Child Opportunity Index (neighborhood resources), based on patient address and 2020 census tract. We used logistic regression to predict the probability of sPTB in models with (1) CL only, (2) CL plus neighborhood indices, and (3) CL, neighborhood, and individual-level characteristics We compared model discrimination using the area under the receiver operating characteristics curve (AUC). Analyses were performed in R version 4.4 (24 April 2024).

Results: The 1000 patients yielded 1038 pregnancies that met inclusion criteria. 21.4% of pregnancies had a prior sPTB (Table 1). 13.3% of index pregnancies ended in sPTB. Decreased CL was associated with increased odds of sPTB in all models. A model with CL only achieved an AUC of 0.56 and 0.57 with addition of neighborhood indices (Table 2). Including other individual covariates increased performance to an AUC of 0.72.

Conclusion: Inclusion of the ICE and COI neighborhood indices did not improve sPTB prediction among individuals who underwent CL screening. Research should explore additional, diverse SDOH indicators alongside individual clinical characteristics to identify those at highest risk of sPTB.