The Science of Healthcare Delivery
Over the past 15 years, policy reports and news headlines have predicted dire physician shortages.1-4 But if workforce projections included a realistic assessment of the roles of NPs and physician assistants (PAs), these high-profile concerns over the supply of physicians might seem considerably less dramatic.
Healthcare workforce projections carry crucial policy implications. Provider shortages can shortchange patients, and overproduction imposes costs on society. Investments in medical education carry high price tags, often paid for by the public.5 Overproduction potentially is devastating to new trainees, who might not be able to secure employment—a pressing concern not only for physicians but also for the NP and PA professions, which are experiencing unprecedented growth.6,7 And supplier-induced demand can raise the overall cost of healthcare if increased availability of services causes patients to use healthcare more than might be justified by their health problems.6,8
The Association of American Medical Colleges (AAMC) commissions the most publicized physician supply and demand projections on an annual basis. Its 2019 update continued to predict large shortages of physicians: between 46,900 and 121,900 by 2032.9 The most recent iterations of the AAMC report improved over the initial ones by factoring NPs and PAs into its projections. The projected scenarios that include NPs and PAs anticipate a narrower gap between physician supply and demand.
The microsimulation techniques used in the AAMC projections allow use of a range of assumptions to produce a corresponding range of projections—a reasonable approach to incorporating uncertainties in the healthcare environment. The updated AAMC report models the effect of adding NPs and PAs to the provider mix using a moderate and high estimate of the proportion of a physician's labor that an NP or PA can replace. The AAMC uses estimates of 25% in primary care, 15% in medical subspecialties, and 10% in surgical subspecialties; the high estimates are twice these portions.
Unfortunately, even the “high” substitution ratios that the AAMC chose to use for NPs and PAs are unrealistically low. The report includes no references that support these ratios, and the AAMC did not respond to my request for any literature it used to select the ratios for the forecasting models. AAMC administrators defend these ratios by suggesting that NPs and PAs often are used to perform clinical tasks that would otherwise not have been provided to patients. If this is true, they argue, PAs and NPs are not just replacing physician labor for some tasks, but also adding to the basket of services provided.10 Following this logic, the substitution ratio of NPs and PAs, compared with physicians, should be low. For example, if a PA or NP is available, they might be used for extensive patient education or for additional follow-up visits that the physician would not have otherwise performed. Although anecdotal reports exist of these uses of PAs, published research suggests that this is not the norm. For example, in a comparison of patients cared for with and without a PA involved, using data from the federal nationally representative Medical Expenditure Panel Survey, adjusting for patient complexity and analyzing over a 1-year period of care, patients who had PA involvement did not have higher use of office visits.11 A number of additional analyses from national samples show that although patients seen by physicians tend to be slightly older and less healthy than patients seen by NPs and PAs, the patient populations and patient care tasks of PAs and NPs are generally similar to those of physicians.12-20
Although the AAMC projections employ a substitution ratio of 25% for PAs and NPs in their “status quo” projections, the weight of evidence points to a reasonable primary care substitution ratio at or above 75%.21-25 Of particular note, in 2011, the Secretary of Health and Human Services convened an interprofessional committee comprising 29 experts to review available evidence and advise the secretary on guidelines for determining health professions shortage areas (HPSAs). One of the central questions posed to this committee was the appropriate method of accounting for NP and PA contributions to primary care. The committee deliberated for 36 days over 14 months and concluded that a reasonable substitution ratio for the purpose of determining HPSA status was 75%. To use a higher ratio, the committee commented, might disadvantage areas that rely heavily on NPs, PAs, and nurse-midwives because scope-of-practice limitations prevent them from supplying the full range of services in some states.23
Moreover, back-of-the-envelope calculations provide an intuitive argument. If the productivity of NPs and PAs was as low as indicated in the AAMC's “high” scenario, it would not be profitable to hire them. Strong employer demand for NPs and PAs and wage increases that persistently exceed inflation suggest that this is not the case.26
The latest iteration of AAMC's projection model accounting for NPs and PAs, even at unrealistically low ratios, demonstrates their large potential effect on physician shortage estimates. For example, Exhibit 1 in the 2019 report indicates that, once NPs and PAs are factored in, the supply and demand projection lines for physicians converge about 2030, indicating that supply will meet demand. For the primary care projections that factor in NPs and PAs as substituting for one-half of a physician, the report's third exhibit shows the physician supply and demand lines converging in 2024. If projections used more realistic substitution ratios for NPs and PAs, the much-publicized future shortfalls of physicians would shrink dramatically or perhaps even disappear. Health workforce planning should be based on the best evidence available in order to avoid either shortfalls or surpluses—both carry high costs to society.
4. Cooper RA, Getzen TE, McKee HJ, Laud P. Economic and demographic trends signal an impending physician shortage
. Health Aff (Millwood)
5. US Government Accountability Office. Physician
workforce: HHS needs better information to comprehensively evaluate graduate medical education funding. GAO-18-240. http://www.gao.gov/assets/700/690581.pdf
. Accessed July 1, 2019.
6. Lopes MA, Almeida ÁS, Almada-Lobo B. Handling healthcare workforce
planning with care: where do we stand. Hum Resour Health
7. Salsberg E, Quigley L. Are we facing a physician assistant
8. Labelle R, Stoddart G, Rice T. A re-examination of the meaning and importance of supplier-induced demand. J Health Econ
10. Richardson G, Maynard A, Cullum N, Kindig D. Skill mix changes: substitution or service development. Health Policy
11. Morgan PA, Shah ND, Kaufman JS, Albanese MA. Impact of physician assistant
care on office visit resource use in the United States. Health Serv Res
. 2008;43(5 Pt 2):1906–1922.
12. Hooker RS, McCaig LF. Use of physician
assistants and nurse practitioners in primary care, 1995-1999. Health Aff (Millwood)
13. Cipher DJ, Hooker RS, Guerra P. Prescribing trends by nurse practitioners and physician
assistants in the United States. J Am Acad Nurse Pract
15. Morgan P, Everett C, Hing E. Nurse practitioners, physician
assistants, and physicians in community health centers, 2006-2010. Healthc (Amst)
16. Stange K. How does provider supply and regulation influence health care markets? Evidence from nurse practitioners and physician
assistants. J Health Econ
17. Morgan P, Everett CM, Smith VA, et al Factors associated with having a physician
, nurse practitioner, or physician assistant
as primary care provider for veterans with diabetes mellitus. Inquiry
18. Virani SS, Akeroyd JM, Ramsey DJ, et al Comparative effectiveness of outpatient cardiovascular disease and diabetes care delivery between advanced practice providers and physician
providers in primary care: implications for care under the Affordable Care Act. Am Heart J
19. Virani SS, Akeroyd JM, Ramsey DJ, et al Health care resource utilization for outpatient cardiovascular disease and diabetes care delivery among advanced practice providers and physician
providers in primary care. Popul Health Manag
20. Jackson GL, Smith VA, Edelman D, et al Intermediate diabetes outcomes in patients managed by physicians, nurse practitioners, or physician
assistants: a cohort study. Ann Intern Med
21. Larson EH, Hart LG, Ballweg R. National estimates of physician assistant
productivity. J Allied Health
22. Morgan P, Everett CM, Hing E. Time spent with patients by physicians, nurse practitioners, and physician
assistants in community health centers, 2006-2010. Healthc (Amst)
23. Negotiated Rulemaking Committee on the Designation of Medically Underserved Populations and Health Professional Shortage
Areas. Negotiated Rulemaking Committee on the Designation of Medically Underserved Populations and Health Professional Shortage
Areas Report: Appendices and Addenda. http://www.hrsa.gov/advisorycommittees/shortage/nrmcfinalreport.pdf
. Accessed July 1, 2019.
24. Doescher MP, Andrilla CH, Skillman SM, et al The contribution of physicians, physician
assistants, and nurse practitioners toward rural primary care: findings from a 13-state survey. Med Care
26. Quella A, Brock DM, Hooker R. Physician assistant
wages and employment, 2000-2025. JAAPA