Why nurse practitioners and physician assistants may be the future of primary care, and why they are already thriving in concierge medicine

Before evaluating who should deliver primary care, strip the question to its fundamentals. Primary care exists to solve a small number of basic problems: continuity of trust between a patient and a clinician, timely access when something is wrong, accurate diagnosis and management of common and chronic conditions, and coordination when specialty care is needed.

Those functions depend on training depth, clinical judgment, time and communication. A physician brings a longer and broader scientific and residency pathway. An NP or PA brings, depending on the individual, extensive direct patient-care experience, a generalist or team-based clinical education, and, in many states, authority to diagnose and treat. The shortage of primary care physicians is fundamentally a supply problem. How the system responds to it, and how well, depends on how clinicians are organized.

The most persistent objection to NP- and PA-led primary care is quality, and it deserves a direct, balanced answer.

The best-known randomized trial, published in JAMA in 2000 by Mundinger and colleagues, assigned more than 1,300 patients to an NP or a physician for primary care. Health status and utilization were comparable, and NP patients with hypertension had lower diastolic blood pressure.⁷ The trial has recognized limits: it is now more than 25 years old, follow-up was relatively short, and NPs practiced with the same authority and physician backup available in the study setting.

Later reviews point in the same direction. Stanik-Hutt and colleagues analyzed 37 studies and found NP care comparable or better across 11 aggregated quality measures.⁸ A 2015 systematic review of randomized trials by Swan and colleagues found NP care comparable to physician care, with some studies showing better cholesterol and blood pressure control, higher patient satisfaction and longer consultations.⁹ For PAs and NPs together, studies in community health centers and in early diabetes management found practice patterns and outcomes comparable to physicians.^10,11

There is credible contrary evidence. In an analysis of Medicare ACO data, Hattiesburg Clinic in Mississippi found that patients whose primary clinician was an NP or PA had higher per-member spending ($43 per month unadjusted, $119 risk-adjusted), more emergency visits, and that physicians outperformed on nine of ten quality measures.¹² The authors described the work as observational rather than a formal study, and it reflects one system in which NPs and PAs carried their own panels; the clinic attributed much of the cost difference to more testing and specialist referrals.^12,13

Read together, the literature supports a clear but bounded conclusion. For routine diagnosis, chronic disease management, prevention and patient education, most studies find NP and PA care comparable to physician care, frequently with higher satisfaction. Outcomes appear to depend heavily on structure: how panels are assigned, how escalation works and how clinical leadership is organized.

A 2025 Health Affairs study by Zhu and colleagues, using linked national data, found that concierge and DPC practice sites grew 83.1% between 2018 and 2023, and the number of clinicians in them grew 78.4%, from 3,935 to 7,021.^20,21 Over the same period, physicians fell from 67.3% to 59.7% of that workforce, while NPs and PAs rose from 32.7% to 40.3%.²² By CMT's calculation from those published figures, that is roughly 2,830 NPs and PAs in 2023, more than double the number in 2018. Discussing the research, Dr. Zhu noted that frontline clinicians across primary care, including in academic and hospital-owned practices, are increasingly advanced practice clinicians.²³

Two practice designs appear most often. In the segmented or hybrid model, a physician converting to concierge retains an NP or PA to continue caring for the larger non-membership panel, preserving existing relationships during the transition; conversion firms describe this as a standard option.²⁴ In a growing number of states, NPs and PAs also own and lead membership practices outright.

Setting also matters for the clinicians themselves. PAs in office-based private practice report the lowest burnout of any setting, 30%, compared with 32.6% for PAs overall and 41.5% in community health centers.⁴ Membership practices, with smaller panels and fewer handoffs, are designed around the conditions associated with that difference.

None of this minimizes physicians. The comparable-outcomes evidence concerns routine and chronic-condition visits across large populations. It does not extend to undifferentiated or high-acuity diagnostic complexity, multi-system disease, procedural care, or the pattern recognition physicians develop through residency and subspecialty training. The studies above were largely conducted in scope-appropriate, collaborative settings, and the strongest contrary evidence came from a setting where that structure was weakest.

The reading most consistent with CMT's Learn, Build, Lead framework is that physicians who lead sustainable, patient-centered practices increasingly do so as the senior clinical and business leader of a team. A hybrid concierge model with a segmented panel, a membership practice with NP or PA colleagues handling preventive and chronic-stable visits, or a mentorship role for advanced practice clinicians all allow physicians to meet rising demand without absorbing an unsustainable panel. That is a leadership decision, not a concession.

Physicians who plan around this shift, rather than treating it as a threat, are the ones most likely to build the sustainable, patient-centered practices this platform exists to help them build.

Design

Narrative evidence review of national workforce projections, licensure and certification data, randomized and observational quality studies, and patient experience research, with claim-level verification against the cited sources.

Derived figures

The estimate of roughly 2,830 NPs and PAs in concierge and DPC practices in 2023 was calculated by CMT by multiplying the published clinician total by the published advanced practice share in Zhu et al. (2025). The source data combine concierge and DPC practices and cannot be separated.

Limitations

Many quality studies are observational, and randomized evidence is dated. Study settings vary in supervision and team structure. Patient satisfaction instruments differ. Visit-length figures for membership practices are industry estimates. Workforce projections differ by model, and state scope-of-practice laws change frequently.

Research support

Literature searches for this brief were supported by AI-assisted research tools. Every statistic is tied to the cited source, and derived figures are labeled. Concierge Medicine Today does not accept vendor sponsorship in exchange for coverage.