A Health Affairs Study Tracked DPC and Concierge Practice Sites From 2018 to 2023. The Number Grew 83 Percent. The Physician Share of the Workforce Did Not.

Analysis by DPC Insider — Research Desk
#research#workforce#advanced-practice#practice-growth#health-affairs#concierge-medicine
Sources: Growth In Number Of Practices And Clinicians Participating In Concierge And Direct Primary Care, 2018–23 | Health Affairs · Growth In Number Of Practices And Clinicians Participating In Concierge And Direct Primary Care, 2018–23 | Johns Hopkins Bloomberg School of Public Health · Concierge and direct primary care practices surge nationwide | Physicians Practice · Jane M Zhu MD, MPP, MSHP | OHSU People · Nurse Practitioners Occupational Outlook Handbook | U.S. Bureau of Labor Statistics · Direct Primary Care Arrangements 50-State Survey | McDermott Will & Schulte · State Practice Environment | American Association of Nurse Practitioners

The numbers behind DPC’s growth are not in dispute. What is harder to track — and rarely measured at a national level — is who is actually doing the clinical work inside those practices, and whether the answer has changed.

A Health Affairs study published in December 2025 adds the data layer the field has been missing. Jane M. Zhu, a physician-researcher at Oregon Health & Science University, along with co-authors Trisha Marsh, Daniel Polsky, Aine Huntington, and Zirui Song, built a national sample of concierge and direct primary care practices by linking public and proprietary data — an approach the authors describe as novel, applied to a sector where, as they note, evidence on scope and workforce characteristics has been scant — and their estimates were likely a lower bound. The result is the clearest picture yet of how the DPC and concierge workforce has evolved.

From 2018 to 2023, the number of practice sites grew by 83.1 percent. The number of clinicians participating in them grew by 78.4 percent.

Those headline numbers confirm what anyone tracking the field has observed. The growth is real and it is documented at scale. What the study also found is that the composition of who works in these practices has shifted in a specific direction.

A Different Workforce Than Five Years Ago

In 2018, 67.3 percent of clinicians in concierge and direct primary care practices were physicians. By 2023, that share had fallen to 59.7 percent. Advanced practice clinicians — primarily nurse practitioners and physician assistants — made up the difference.

The absolute number of physicians in these practices grew. A falling share does not mean fewer physicians in DPC. It means advanced practice clinicians entered faster.

That trajectory is consistent with what is happening across primary care more broadly. The Bureau of Labor Statistics projects nurse practitioner employment to grow roughly 41 percent from 2025 to 2035 — well above the average across all occupations. Physician assistant employment is projected to grow 21 percent over the same period. If APPs are joining the primary care workforce at rates significantly exceeding physicians, DPC would be unusual if that pattern didn’t appear inside it.

The more interesting question is whether DPC is attracting APPs for the same reasons it attracts physicians. The model appeals to physicians who want out of the insurance billing apparatus: smaller panels, longer appointments, a more direct relationship with patients, and greater clinical autonomy than traditional employed settings typically offer. Those motivations do not require a physician. A nurse practitioner with full practice authority — which the American Association of Nurse Practitioners recognizes in 27 states and the District of Columbia — can deliver primary care under the same direct-pay structure, for largely the same reasons, in states where DPC statutes do not restrict the agreement to physicians.

Where State Law Does Not Match the Reality

The regulatory picture complicates this. Some state DPC exemption laws — the statutes that carve DPC agreements out of insurance regulation, shielding them from being regulated as insurance products — use physician-specific language. A nurse practitioner or physician assistant running a DPC practice in those states may sit in ambiguous territory: delivering care under a structure the statute was not designed to accommodate.

McDermott Will & Schulte’s 50-state survey of DPC arrangements documents this variability. The problem is not new. But the Health Affairs workforce data gives it urgency. If the APP share of the DPC workforce continues to grow — and the five-year trajectory in this study suggests it might — state legislatures that defined DPC around physician provision will increasingly be describing a narrower version of what’s actually on the ground.

The Identity Question the Numbers Surface

DPC’s origin narrative is substantially a physician story. The model emerged in part from a reaction to what fee-for-service medicine had done to primary care practice — the burned-out physicians, the compressed appointment times, the documentation burden that consumed time that used to go to patient care. The physicians who built DPC built it around a specific grievance and a specific alternative.

That narrative stays intact even as the workforce inside the model changes. But it describes a smaller fraction of what’s actually happening. APPs in DPC practices deliver real primary care. Patients in those practices have access to smaller panels, more direct communication, and longer appointments — the structural features that distinguish DPC regardless of the credential behind the exam room door. The clinical argument for the model does not depend on the title of the person delivering it.

What does depend on that title is the conversation about professional autonomy — the part of the DPC argument that is specifically about what practicing medicine can look like when the insurance billing system is removed from the equation. Whether that argument has the same resonance when directed at a nurse practitioner as when directed at a physician who trained under a different set of expectations is not a question the data resolves.

The Open Question

The Health Affairs study is a workforce study. It counts practice sites and clinician types. It does not measure patient outcomes in APP-led DPC practices versus physician-led ones — and the study itself notes that longitudinal evaluations of these models remain lacking. Whether the care delivered differs, and whether patients can tell, is not something the study asked.

What the data establishes is structural: the field has grown substantially since 2018, and the share of the people delivering care who are physicians has declined. Both of those things are now documented in a published, national-sample study.

What they mean for the patient choosing a DPC practice, the state legislature writing a new exemption bill, or the DPC community deciding who the model is for — the study raises all of that and answers none of it.