Current shortages of primary care physicians (PCPs) are expected to worsen in the coming years.1–3 Recent findings from the Health Resources and Services Administration suggest that the demand for primary care will result in a projected shortage of approximately 20,400 PCPs.4 These projections take on increased significance as up to 32 million people gain financial access to health insurance coverage under the Patient Protection and Affordable Care Act (ACA).5–7
Many experts have embraced the idea of increasing the supply of PCPs and expanding the roles and numbers of primary care nurse practitioners (PCNPs) providing primary care services as one way of addressing these projected shortages and consequent threats to access to care.8
Although concerns about the supply of PCPs were recognized prior to the passage of the ACA, the act provided new impetus and limited funding to increase the supply of these clinicians and improve access to care in rural and underserved areas.6,9–11 Similar to provisions aimed at increasing the supply of PCPs and improving their geographic distribution, the ACA authorized the expansion of the PCNP workforce through increased opportunities for education and interprofessional collaboration; however, many of these provisions have not been funded by Congress.8,12
Other policy levers acting on the health care system have the potential to affect the supply of and demand for primary care clinicians. Efforts are under way to redesign the way primary care is delivered. Many of these efforts focus on moving toward a more collaborative care delivery system with the goal of ensuring that care is provided in a way that maximizes the skills and training of all clinicians.13–16 In addition, there is ongoing discussion around standardizing state nursing scope-of-practice laws. Finally, surveys suggest that patients are receptive to having PCNPs take on a greater role in their care.17–20
Addressing the primary care shortage is likely to require more than improving payment rates for primary care services, increasing the number of training and residency positions, and expanding the contributions of NPs and other primary care workers. Surveys of PCPs find an increasingly dissatisfied workforce, and studies of residents suggest that many in primary care training programs are planning for a career in a subspecialty.21–23
These and many other issues affecting the primary care workforce were documented in special issues of Academic Medicine24 and Health Affairs25 in late 2013. However, missing from the more than 50 articles and essays published in those two journals was any consideration of primary care professionals’ willingness to recommend careers in primary care. Earlier studies by us and others have demonstrated a relationship between (1) perceptions of workforce supply and career and job satisfaction and (2) the willingness to recommend and pursue careers in the health professions.26–36
In the present study, we carried out a national survey of PCPs and PCNPs to obtain data that would allow us to examine PCPs’ and PCNPs’ willingness to recommend a career as a PCP or a PCNP. Specifically, we examined associations between PCPs’ and PCNPs’ willingness to recommend a career in primary care and their attitudes about job and career satisfaction, perceptions of collaborative working relationships, and the potential effects of an increased PCNP supply on physicians’ incomes and employment.
We mailed the survey to an initial sample of 1,914 clinicians from November 23, 2011 to April 9, 2012. We chose those clinicians by randomly selecting 957 PCPs from the American Medical Association Masterfile (maintained by Medical Marketing Services), a comprehensive listing of all licensed physicians in the United States. Similarly, we randomly selected 957 PCNPs from the NP Masterfile provided by Medical Marketing Services. Eligibility for the survey was restricted to clinicians who were licensed NPs or physicians trained in a primary care specialty, actively working in primary care practice, and providing direct patient care services. (See Supplemental Digital Table 1 at http://links.lww.com/ACADMED/A251 for a list of eligible specialties.)
Recognizing that specialty information in the sample frames might not be accurate, sampled clinicians were instructed to mail back the instrument if they were not working in a primary care setting. Harris Interactive, Inc., managed the data collection on our behalf.
We developed the survey instruments, one for PCPs and one for PCNPs, using expert review and subject pretesting of measures, and drawing from multiple prior health workforce surveys that we had developed. Domains included scope of work, perceptions of labor supply and NP practice, and personal and practice characteristics of PCPs and PCNPs. (For the survey instruments, see Supplemental Digital Appendices 1 and 2 http://links.lww.com/ACADMED/A251.)
We mailed the surveyed individuals four times to obtain data. We sent the first mailing via priority mail and included a cover letter, the survey instrument, a $35 incentive check, and a postage-paid return envelope. The second and third mailings were complete packets (absent the incentive) sent by first-class mail. To achieve our response rate target, we sent the fourth mailing with a $60 prepaid check as an incentive.
We compared PCP respondents to non-respondents and found small differences between the two groups on number of years in practice, sex, and region. We created weights to account for these differences and used these weights in all analyses (weights ranged from 0.42986 to 2.53572). We compared responding PCNPs to those that did not respond and found small statistically significant differences on sex and region. The sampling frame did not contain data on years in practice. We created weights to account for these differences and used them in all analyses. Weights for PCNPs ranged from 0.91367 to 1.49120.
We used the entire sample of respondents for our analyses of all attitudinal measures and personal and practice characteristics. The sampling error for the entire sample is ± 3.1%. We examined descriptive personal and practice characteristics (see Table 1) and multiple outcomes concerning perceptions of supply and demand, including perceptions of shortages of primary care clinicians nationally and in the local community, career and work recommendations, and the impact of increasing the supply of PCNPs in the United States. Exact wordings of questions are shown in Supplemental Digital Appendices 1 and 2, which can be found at http://links.lww.com/ACADMED/A251.
We created a categorical scope-of-practice variable using the ZIP Codes provided by respondents. This variable segmented our respondents into those practicing in less restrictive, moderately restrictive, and most restrictive states based on state laws governing entry into the profession and requirements for physician collaboration, reimbursement, and prescribing at the time the survey was taken. The exact specifications of this variable are shown in Supplemental Digital Tables 2 and 3, found at http://links.lww.com/ACADMED/A251.
Our analysis was descriptive with a primary focus on the comparison of attitudes and experience of PCPs and PCNPs in primary care settings. We examined univariate and bivariate relationships, comparing NPs and PCPs using two-sample t tests for continuous variables, and chi-square tests for categorical variables on measures posed to both groups.
Of the 1,914 surveys mailed, 972 eligible clinicians (505 PCPs, 467 PCNPs) filled out and returned the survey, 8 eligible clinicians refused to return the survey, 201 clinicians proved to be not eligible, and 733 clinicians were of unknown eligibility (602 did not return the survey and could not be contacted; 131 could not be contacted at all—no forwarding address). We used procedures established by the American Association for Public Opinion Research37 to estimate the proportion of eligible clinicians among the 733 who were of unknown eligibility. In estimating that proportion, we assumed that the same proportion of the clinicians of unknown eligibility were eligible as we found to be eligible in the group of clinicians whom we successfully contacted; that proportion was 0.83. Using this approach, we estimated that there were approximately 1,589 eligible clinicians and thus a response rate of 61.2%.
Detailed characteristics of the respondents are shown in Table 1. As the table indicates, on average PCNPs were older and more likely to be white and female than PCPs (P ≤ .01 for all comparisons). In addition, PCNPs reported fewer years in practice and lower incomes than PCPs (P ≤ .01 for all comparisons). Among PCNPs, 318 (68%) were located in highly restrictive scope of practice states, 66 (14%) in moderately restrictive states, and 82 (17%) in states with no restrictions on scope of practice. The distribution across states was similar for PCPs: 354 (70%) in highly restrictive states, 68 (13%) in moderate ones, and 83 (16%) in states with no restrictions. (The total of these percentages is not quite 100% because of rounding.)
Perceptions of workforce supply and demand
The survey asked both groups of clinicians about their perceptions of the supply of PCPs and PCNPs in the United States and their local communities. As shown in Table 2, more than 80% of clinicians in both groups said they believed there is a national shortage of PCPs. PCPs were less likely than PCNPs to report national shortages of PCNPs: 263 PCPs (52%) versus 364 PCNPs (78%), P ≤ .001. When asked about the community in which they practice, approximately one-half of PCNPs and PCPs reported a shortage of PCPs (P < .04). PCNPs were significantly more likely than PCPs to indicate that the supply of PCNPs in their community was somewhat or much less than the demand: 167 PCPs (33%) versus 266 PCNPs (57%), P ≤ .001.
Willingness to recommend a career
The survey asked, “Given what you know about the state of health care, would you advise a qualified high school or college student to pursue a career as a PCP/PCNP?” A total of 282 PCPs (56%) would recommend their own career, as would 410 PCNPs (88%). However, both PCPs and PCNPs were significantly more likely to recommend a career as a PCNP than as a PCP: 333 PCPs (66%), 410 NPs (88%).
We explored the association between respondents’ perception of a workforce shortage of either PCPs or PCNPs and their willingness to recommend a career in those professions. As shown in Figure 1, perceptions of a shortage of PCPs either nationally or in their own communities did not affect willingness to recommend the career among either PCNPs or PCPs. In contrast, PCPs reporting a shortage of PCNPs were significantly more likely than those who did not perceive a shortage of these clinicians to recommend a career as a PCNP: 172 (71%) versus 155 (64%), P ≤ .05. Perceptions of a shortage of PCNPs did not affect PCNPs’ willingness to recommend their own career.
Job and career satisfaction
We asked respondents to report on satisfaction with both their career choice and their current employment (see Figure 2) and explored the relationship between career and job satisfaction with career recommendations. Majorities of both PCPs and PCNPs reported being very or somewhat satisfied with their current employment (88% of PCNPs, 83% of PCPs; P = .068). However, the two groups differed significantly in terms of satisfaction with career. Three hundred forty-one PCNPs (73%) reported that they were very satisfied with their career, significantly more than the 232 PCPs (46%) who reported being very satisfied with their career as a primary care clinician (P ≤ .001; see Figure 2). Career satisfaction was associated with willingness to make career recommendations among PCPs. Among PCPs who were less than very satisfied with their careers, 101 (37%) would definitely or probably recommend a career as a PCP, and 172 (63%) would recommend a career as a PCNP. A similar pattern was found among PCPs who were less than very satisfied with their current employment (data not shown). Among PCNPs, approximately equal proportions of those who were less than very satisfied with their careers or their current employment would recommend either career (data not shown).
Ratings of opportunities to influence the workplace and working relationships between clinicians
In addition to career and job satisfaction measures, we included items to assess clinicians’ views of their influence on decisions made in the workplace, as well as their interprofessional working relationships. We found that PCPs were more likely than PCNPs to hold positive views about opportunities to influence decisions affecting these two dimensions of the workplace environment. More PCPs than PCNPs perceived that opportunities to affect the organization of their workplace—273 PCPs (54%) versus 135 PCNPs (29%), P ≤ .001—and influence decisions about patient care—338 PCPs (67%) versus 205 PCNPs (44%), P ≤ .001— were excellent or very good. The two clinician groups did not differ significantly on ratings of the quality of working relationships between PCNPs and PCPs in their workplace.
We next explored the relationship between workplace ratings and willingness to recommend their career. As shown in Figure 3, for those PCNPs rating their workplace as fair or poor on “opportunities to influence workplace organization,” 90 (46%) would definitely recommend a career as a PCNP and 55 (28%) would give the same recommendation regarding a career as a PCP, compared with 27 (19%) and 28 (20%) of PCPs, respectively (P ≤ .001 for both comparisons). Findings on willingness to recommend a career as a primary care clinician showed a similar pattern among those clinicians who rated their practice site as fair or poor on opportunities to influence patient care. Here, 50 PCNPs (41%) would recommend their own career and 29 (26%) would recommend a career as a PCP compared with 14 (20%) and 12 (17%) PCPs, respectively (P ≤ .01 for all comparisons).
Impact of an increasing supply of PCNPs
The survey included a series of items on the likely impact of increasing the number of PCNPs in primary care practice in the United States. PCNPs and PCPs held different beliefs about (1) the effect of an increased supply of PCNPs on their own income—288 PCPs (57%) believed their income would decrease, compared with 103 PCNPs (22%), P ≤ .001—and (2) the effect of the replacement of PCPs by PCNPs—374 PCPs (74%) believed that a greater number of PCNPs would lead to the replacement of PCPs with these clinicians, compared with 233 PCNPs (50%), P ≤ .001. There were no significant differences between either type of clinician in their willingness to make career recommendations and their beliefs about the effects of an increased number of PCNPs on physician replacement or income (data not shown).
Growing concerns over current and projected shortages of PCPs and reduced access to primary care have prompted many workforce policy experts to focus not only on increasing the supply and geographic distribution of PCPs but also on expanding the PCNP workforce and removing regulations restricting their practice. Results of our survey raise some surprising issues for policy makers and others who are interested in transitioning the current primary care workforce into one that is not only adequately sized but also differently configured and aligned to respond to the primary care health needs of Americans.
In other areas of workforce research,38–43 levels of autonomy and opportunities to contribute to the decision-making process in the workplace have been shown to predict work satisfaction and retention, and our findings suggest that this holds true among primary care clinicians. However, even in settings in which they perceive a lack of opportunity to influence workplace organization and patient care, PCNPs were still more willing than PCPs to recommend their career to qualified college students, suggesting possibly that the intrinsic rewards of being a PCNP may offset other negative effects. Prior studies have underscored the important relationship between professional perceptions of workforce supply, career and job satisfaction, and willingness to recommend a career; however, in light of our survey findings, what are health workforce planners and policy makers to conclude when, on the one hand, the PCPs we studied were more likely to believe that the shortage of primary care clinicians is primarily due to a shortage of physicians but, on the other hand, were more likely to recommend a career in primary care as a PCNP? Moreover, PCPs’ greater willingness to recommend a career as a PCNP over a career as a PCP is difficult to reconcile, given their beliefs about the impact of increasing the supply of primary care PCNPs on their own incomes.
It is possible that PCPs’ greater willingness to recommend a career as a primary care PCNP over a career in their own profession could reflect their pessimism about the future of primary care medicine. Dissatisfaction with factors not assessed in this survey—lower payments and incomes relative to specialists, long work hours, increasing bureaucracy and compliance oversight, devaluation of primary care among the academic medical community, the additional years of education, and high debt levels following the completion of medical education, particularly in relation to their salary as compared with physicians in other specialties—could weigh heavily enough to offset PCPs’ misgivings about primary care PCNPs and thus explain their greater willingness to recommend that qualified students pursue careers as PCNPs.
There are several limitations to our work. First, our sample source for PCNPs, as compared with our source for physicians, had a higher rate of inaccurate contact information and did not contain data on clinical activities associated with direct patient care. Second, in relatively small samples such as these, it is not possible to control for all of the personal and clinical practice characteristics that may differ across and within groups of clinicians.
In addition, it is difficult to adequately adjust for variation in state nursing scope-of-practice laws. Our scope-of-practice variable was not associated with any of our outcomes and was subsequently removed from the analysis. We cannot determine whether the lack of significant findings on this variable was due to a true lack of difference or, instead, the relatively small size of the PCNP sample. In addition, there is considerable variation in scope-of-practice regulations within the highly and moderately restrictive states, making it difficult to create mutually exclusive groups of states based on restrictiveness.44 Further research on similarities and differences among these state’s scope-of-practice regulations could help to refine this grouping variable, thus making it more useful for policy analysis.
Finally, as in any survey, there is concern about response bias. Although our response rate is high for a clinician survey, it is possible that nonresponders were systematically different from responders, resulting in nonresponse bias. However, the demographics of our PCNP respondents closely resemble those found in the National Sample Survey of PCNPs conducted by the Health Resource and Services Administration.45
Our findings suggest that funding new programs to increase the supply of PCPs may not have the intended long-term effect if these newly trained physicians come to see primary care as a career they cannot recommend to others. Simply bolstering a dissatisfied workforce with additional training and residency slots is unlikely to solve this problem. Redesigning the work of the PCP through new models of care designed to increase efficiency, quality, and patient centeredness could increase job satisfaction and help make this career more attractive. This finding is supported by recent research suggesting that physicians practicing in well-functioning teams report higher levels of satisfaction.46,47 Assessing the effects of medical homes and other delivery reforms on physicians’ satisfaction will be an important area for study.
In recent years, primary care resident positions have been filled to capacity, possibly as a result of the willingness of international medical graduates to serve in this role, suggesting that the projected shortage of clinicians is driven primarily by increased demand. Expanding such training programs will certainly help to increase the supply of PCPs. However, our findings indicate that a multipronged approach to solving the primary care workforce shortage may be necessary. First, traditional solutions to workforce shortages in other sectors have included raising wages and making improvements to the workplace environment. In addition, initiatives such as the Johnson & Johnson Campaign for Nursing’s Future34 have sought to improve the image of the nursing profession. Our findings suggest that similar efforts may be required for the PCP workforce. Second, PCPs, in this and other studies, have expressed increasing disillusionment with their jobs and careers. Conversely, PCNPs, as reported elsewhere, have lower incomes, work fewer hours, see fewer patients, and report less workplace autonomy and still report a higher magnitude and degree of satisfaction.48 These differences suggest the need to restructure the work of the primary care physician to increase satisfaction with the career. Taken together, our findings indicate that rather than relying solely on programs aimed at increasing the supply of PCPs, there may be a greater benefit in reimagining how the entire primary care workforce should be structured, with the understanding that without a significant shift in how PCPs view their own careers, the solutions to the primary care shortage are likely to fall short.
1. Colwill JM, Cultice JM, Kruse RL. Will the generalist physician supply meet the demands of an increasing and aging population? Health Aff (Millwood). 2008;27:w232–w241
2. Petterson SM, Liaw WR, Phillips RL, Rabin DL, Meyers DS, Bazemore AW. Projecting U.S. primary care physician workforce needs: 2010–2025. Ann Fam Med. 2012;10:503–509
3. Association of American Medical Colleges Center for Workforce Studies. Recent Studies and Reports on Physician Shortages in the U.S. 2012 Washington, DC Association of American Medical Colleges https://www.aamc.org/download/100598/data/
. Accessed October 21, 2014
4. U.S. Department of Health and Human Services, Health Resources and Services Administration, National Center for Health Workforce Analysis. Projecting the Supply and Demand for Primary Care Practitioners Through 2020. 2013 Rockville, Md U.S. Department of Health and Human Services
5. Ku L, Jones K, Shin P, Bruen B, Hayes K. The states’ next challenge—securing primary care for expanded Medicaid populations. N Engl J Med. 2011;364:493–495
7. Hofer AN, Abraham JM, Moscovice I. Expansion of coverage under the Patient Protection and Affordable Care Act and primary care utilization. Milbank Q. 2011;89:69–89
9. National Governors Association. NGA Paper: The Role of Nurse Practitioners in Meeting Increasing Demand for Primary Care. 2010 Washington, DC National Governors Association
10. Institute of Medicine. The Future of Nursing: Leading Change, Advancing Health. 2011 Washington, DC National Academies Press
13. Everett C, Thorpe C, Palta M, Carayon P, Bartels C, Smith MA. Physician assistants and nurse practitioners perform effective roles on teams caring for Medicare patients with diabetes. Health Aff (Millwood). 2013;32:1942–1948
14. Everett CM, Schumacher JR, Wright A, Smith MA. Physician assistants and nurse practitioners as a usual source of care. J Rural Health. 2009;25:407–414
15. Shipman SA, Sinsky CA. Expanding primary care capacity by reducing waste and improving the efficiency of care. Health Aff (Millwood). 2013;32:1990–1997
16. Auerbach DI, Chen PG, Friedberg MW, et al. Nurse-managed health centers and patient-centered medical homes could mitigate expected primary care physician shortage. Health Aff (Millwood). 2013;32:1933–1941
17. Dill MJ, Pankow S, Erikson C, Shipman S. Survey shows consumers open to a greater role for physician assistants and nurse practitioners. Health Aff (Millwood). 2013;32:1135–1142
18. Dower C, Moore J, Langelier M. It is time to restructure health professions scope-of-practice regulations to remove barriers to care. Health Aff (Millwood). 2013;32:1971–1976
19. Elwood TW. Patchwork of scope-of-practice regulations prevent allied health professionals from fully participating in patient care. Health Aff (Millwood). 2013;32:1985–1989
20. Spetz J, Parente ST, Town RJ, Bazarko D. Scope-of-practice laws for nurse practitioners limit cost savings that can be achieved in retail clinics. Health Aff (Millwood). 2013;32:1977–1984
21. West CP, Dupras DM. General medicine vs subspecialty career plans among internal medicine residents. JAMA. 2012;308:2241–2247
22. Ratelle JT, Dupras DM, Alguire P, Masters P, Weissman A, West CP. Hospitalist career decisions among internal medicine residents. J Gen Intern Med. 2014;29:1026–1030
23. Frinter MP, Cull WL. Pediatric training and career intentions, 2003–2009. Pediatrics. 2012;308:2241–2247
24. Sklar DP. A special issue on training the future health care workforce. Acad Med. 2013;88(12, theme issue):1785–1949
25. Metz DE. Redesigning the health care workforce. Health Aff (Millwood). 2013;32(11, theme issue):1865–2060
26. Buerhaus PI, Staiger DO The Future of the Nursing Workforce in the United States. 2009 Sudbury, Mass Jones and Bartlett
27. Buerhaus PI, DesRoches C, Donelan K, Hess R. Still making progress to improve the hospital workplace environment? Results from the 2008 National Survey of Registered Nurses. Nurs Econ. 2009;27:289–301
28. Ulrich BT, Buerhaus PI, Donelan K, Norman L, Dittus R. Magnet status and registered nurse views of the work environment and nursing as a career. J Nurs Adm. 2009;39(7–8 suppl):S54–S62
29. Buerhaus PI, Donelan K, Ulrich BT, DesRoches C, Dittus R. Trends in the experiences of hospital-employed registered nurses: Results from three national surveys. Nurs Econ. 2007;25:69–79
30. Buerhaus PI, Donelan K, Ulrich BT, Norman L, Dittus R. State of the registered nurse workforce in the United States. Nurs Econ. 2006;24:6–12
31. Buerhaus PI, Donelan K, Ulrich BT, Kirby L, Norman L, Dittus R. Registered nuses’ perceptions of nursing. Nurs Econ. 2005;23:110–118
32. Buerhaus PI, Donelan K, Ulrich BT, Norman L, Dittus R. Is the shortage of hospital registered nurses getting better or worse? Findings from two recent national surveys of RNs. Nurs Econ. 2005;23:61–71, 96
33. Buerhaus PI, Donelan K, Ulrich BT, Norman L, Williams M, Dittus R. Hospital RNs’ and CNOs’ perceptions of the impact of the nursing shortage on the quality of care. Nurs Econ. 2005;23:214–221
34. Donelan K, Buerhaus PI, Ulrich BT, Norman L, Dittus R. Awareness and perceptions of the Johnson & Johnson Campaign for Nursing’s Future: Views from nursing students, RNs, and CNOs. Nurs Econ. 2005;23:150–156, 180
35. Norman LD, Donelan K, Buerhaus PI, et al. The older nurse in the workplace: Does age matter? Nurs Econ. 2005;23:282–289
36. Ulrich BT, Buerhaus PI, Donelan K, Norman L, Dittus R. How RNs view the work environment: Results of a national survey of registered nurses. J Nurs Adm. 2005;35:389–396
37. American Association for Public Opinion Research. 2011 Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for Surveys. 20117th ed. Deerfield, Ill American Association for Public Opinion Research
38. Iliopoulou KK, While AE. Professional autonomy and job satisfaction: Survey of critical care nurses in mainland Greece. J Adv Nurs. 2010;66:2520–2531
39. Ellenbecker CH, Cushman M. Home healthcare nurse retention and patient outcome model: Discussion and model development. J Adv Nurs. 2012;68:1881–1893
40. Pasarón R. Nurse practitioner job satisfaction: Looking for successful outcomes. J Clin Nurs. 2013;22:2593–2604
41. Pron AL. Job satisfaction and perceived autonomy for nurse practitioners working in nurse-managed health centers. J Am Assoc Nurse Pract. 2013;25:213–221
42. Van Mierlo H, Rutte C, Vermunt J, Kompier M, Doorewaard J. Individual autonomy in work teams: The role of team autonomy, self-efficacy, and social support. Eur J Work Organ Psychol. 2006;15:281–299
45. U.S. Department of Health and Human Services, Health Resources and Service Administration, National Center for Health Workforce Analysis. Highlights From the 2012 National Sample Survey of Nurse Practitioners. 2012 Rockville, Md U.S. Department of Health and Human Services
46. Sinsky CA, Willard-Grace R, Schutzbank AM, Sinsky TA, Margolius D, Bodenheimer T. In search of joy in practice: A report of 23 high-functioning primary care practices. Ann Fam Med. 2013;11:272–278
47. Friedberg MW, Chen PG, Van Busum KR, et al. Factors Affecting Physician Professional Satisfaction and Their Implications for Patient Care, Health Systems, and Health Policy. 2013 Washington, DC RAND Corporation
48. Donelan K, DesRoches CM, Dittus RS, Buerhaus P. Perspectives of physicians and nurse practitioners on primary care practice. N Engl J Med. 2013;368:1898–1906