ORIGINAL RESEARCH
Paula M. Kett, PhD, MPH, RN; Samantha W. Pollack, MHS; Lars E. Peterson, MD, PhD; Davis G. Patterson, PhD
Corresponding Author: Paula M. Kett, PhD, MPH, RN; Center for Health Workforce Studies, Department of Family Medicine, University of Washington School of Medicine.
Email: pmk@uw.edu
DOI: 10.3122/jabfm.2025.250418R1
Keywords: Education, Family Medicine, Family Physicians, Health Workforce, Longitudinal Studies, Maternal Health Services, Obstetrics, Pregnancy
Dates: Submitted: 10-29-2025; Accepted: 04-13-2026
Status: In Press.
BACKGROUND: More information is needed regarding why family physicians continue or cease to provide OB care, as the number of family physicians providing OB services has decreased over time. This study aims to address this gap through describing factors associated with OB service provision over time.
METHODS: For this longitudinal study, we used 2008-2022 American Board of Family Medicine data on family physicians’ demographic and practice characteristics linked to Area Health Resource File community characteristic data. We constructed a binary variable – provision of OB services (delivering babies, providing prenatal care, and/or providing newborn care) – as our main outcome of interest, as well as variables representing “always providing”, “starting” and “stopping” OB services based on participants’ first and last entries in the data. We used the output from multivariable logistic regression analyses to calculate the predicted probability of providing OB care according to a range of individual-, practice-, and community-level factors.
RESULTS: The final sample totaled 44,929 individuals with 106,001 observations. The provision of OB services by family physicians decreased from 62.0% in 2008 to 15.0% in 2022. Analyses demonstrated a significantly higher probability of ever providing OB care if the family physician practices in an academic health center (49.4%) or safety-net practice (27.3%), compared to hospital-based practices (15.9%), practices in the Midwest (25.5%) or the West (22.1%) compared to the Northeast (14.7%), is in a rurally-located practice (26.0%) and/or has a CNM in their practice (28.0% versus 17.7% without). We observed similar associations with respect to factors associated with “always providing” OB care or “starting” OB services after entering practice.
CONCLUSION: The longitudinal design of this study allowed for examination of factors associated with sustaining OB services over time and offered a comprehensive understanding of family physicians who provide this care. Factors consistently predictive of OB service provision include practicing in a rural area and in an academic medical center or safety-net practice, and having a CNM co-located in the practice. More research is needed to determine facilitating factors that enable targeted work towards supporting provision of these services among family physicians.