Variation in Preterm Birth Rates Across Prenatal Care Sites in New York

Objective: To investigate variation in preterm birth rates by the site at which prenatal care was received. Design: Cross-sectional cohort study. Setting: New York State. Participants: Claims and encounter data on singleton live births that were covered by New York Medicaid (N = 154,377). Methods: We analyzed data from New York Medicaid and the American Community Survey. We established sites of prenatal care using geocoded billing addresses for prenatal visits. We calculated descriptive statistics and conducted logistic regression analyses to determine variation in crude and risk-adjusted preterm birth rates by prenatal care site. Results: The crude preterm birth rates averaged 7.8% (range = 2.0%–18.7%) by prenatal care site. The adjusted preterm birth rate was 8.0% (range = 2.8%–18.5%) by prenatal care site. Risk-adjusted preterm birth site-level rates at the 90th percentile were 2.7 times higher than those in the 10th percentile. The variation in risk-adjusted preterm birth site-level rates was not fully explained by birth volume, rural site location, or racial and ethnic composition of the patients who received prenatal care at the site. Conclusion: Wide variation in risk-adjusted preterm birth rates across prenatal care sites exists, and factors beyond known individual demographics and medical factors contribute to the variation. Further research is warranted to identify why receiving care at some prenatal sites is associated with higher risk of preterm birth than receiving care at others.