Retrospective Use of Patients' Characteristics to Assess Variation in Prenatal Care Utilization.

 We used patients' medical and psychosocial risk factors to explore prenatal care utilization and health outcomes to inform prenatal care tailoring.  This retrospective cohort study assessed patients who gave birth at an academic institution from January 1 to December 31, 2018, using electronic health record (EHR) data. Patients were categorized into four phenotypes based on medical/psychosocial risk factors available in the EHR: Completely low risk; High psychosocial risk only; High medical risk only; and Completely high risk. We examined patient characteristics, visit utilization, nonvisit utilization (e.g., phone calls), and outcomes (e.g., preterm birth, preeclampsia) across groups.  Of 4,681 patients, the majority were age 18 to 35 (3,697, 79.0%), White (3,326, 70.9%), multiparous (3,263, 69.7%), and Completely high risk (2,752, 58.8%). More Black and Hispanic patients had psychosocial risk factors than White patients. Patients with psychosocial risk factors had fewer prenatal visits (10, interquartile range [IQR]: 8-12) than those without (11, IQR: 9-12). Patients with psychosocial risk factors experienced less time in prenatal care, more phone calls, and fewer EHR messages across the same medical risk group. Rates of preterm birth and gestational hypertension were incrementally higher with additional medical/psychosocial risk factors.  Data readily available in the EHR can assess the compounding influence of medical/psychosocial risk factor on patients' care utilization and outcomes. · Medical and psychosocial needs in pregnancy can inform patient phenotypes and are associated with prenatal care use and outcomes.. · Patient phenotypes are associated with prenatal care use and outcomes.. · Patients with high psychosocial risk spent less time in prenatal care and had more phone calls in pregnancy.. · Tailored prenatal care models may proactively address differences in patient's needs..

Informations de copyright

Thieme. All rights reserved.

Déclaration de conflit d'intérêts

A.F.P. is a consultant for Maven. The remaining authors report no conflicts of interest related to the subject matter of this manuscript.