Every American should have access to the appropriate practitioner at the right time and the right place when necessary. That is the goal of an adequate health care workforce. The “appropriate practitioner” can be any member of a health care team, not only a physician. The “right time” is as close to now as possible when necessary. The “right place” does not need to be the office, emergency department, or hospital; it can be at the patient’s home. And “when necessary” is determined by a negotiation between patients and practitioners. All practitioners would agree that a visit to the emergency department with a cold is unnecessary. However, in fact, many patients visit emergency departments with nonemergency conditions. An adequate health care workforce would make this a much less common event.
Access to care is currently inadequate for many Americans. However, improving access is more complex than increasing the projected number of physicians and nurses. Other approaches to increasing supply are to leverage the workforce, redefine roles of each member of the team, and use integrated systems to improve efficiency. To reduce demand, we must consider ways to address unnecessary demand by both physicians and patients.
I have previously examined the workforce from the standpoint of the educational systems needed.1 Below, I discuss the workforce in terms of increasing supply and decreasing demand; I then offer a few recommendations about how to begin making needed changes. My goal is to show what we can do to improve access to care by improving how we deliver care and who delivers it, thereby creating better systems of care.
The supply of physicians and nurses
In the United States, shortages of 90,000 physicians (45,000 generalists and 45,000 specialists) and 1.2 million nurses are projected to occur a decade from now.2,3 Even now, there is a serious geographic maldistribution of primary care physicians, with 56.8 million people—about 20%—living in Health Professions Shortage Areas (areas with fewer than 1 primary care physician per 3,500 people).4 As the shortages worsen, more positions will become available in more populated areas, and geographic maldistribution will intensify.
Medical and nursing associations have called for increases in the number of graduates. In the 10 years between 2002 and 2011, U.S. MD-granting schools graduated a total of 16,888 physicians, and there is a substantial increase in the number of medical school graduates under way. Even so, it would take much too long to close the gap. Other approaches to increase physician supply could be to determine ways to have physicians retire later, to establish a National Health Service Corps, to have doctors near the ends of their careers to do part-time work in shortage areas, or by contracting with retiring physicians to provide telemedicine consultation or supervision of Grand-Aides (see the next section for a discussion of Grand-Aides). According to one estimate, if physicians retired 2 years later, this would have the effect of adding 30,000 physicians to the workforce, double the number added by schools in the last 10 years.5
Yet any attainable increase in the number of physicians and nurses will not put a significant dent in the projected shortages. However, the projections do not assume any change in the systems of care. I maintain that with new systems,6 the projections will change, and we will see that we do not need as many physicians and nurses.
System change: New roles in the team
Team care is not new. It makes sense. Ideally, teams remove professional silos and improve communication across the team and with the patient. The most effective way to improve the efficiency and cost of a team is to create new roles that make sense, and to leverage each of the team members to perform at the top of his or her competency. If achieving the top of each member’s competency is not supported by current state licensing boards, then the boards need to reexamine their regulations.
The ideal is to begin with the patient as a member of the team and connect the patient and family with the professional members of the team (e.g., physicians, nurses, physician assistants, pharmacists) and with people who leverage the effectiveness of the professionals, such as community health workers (who are typically involved in health advising, information, referrals, translation services, and advocacy with an orientation toward social services)6,7 or other more medically oriented Grand-Aides.8
Grand-Aides are experienced, caring people connecting the patient and care team quickly and cost-effectively; they are becoming parts of innovative delivery systems. Grand-Aides, regardless of age, have the temperament and personality of a good grandparent. They have had prior training in medical care (such as training as a certified nurse aide) and then take an added 200-hour Grand-Aides curriculum, which includes ways to encourage medication adherence as well as medical content. They are usually paid. Under close supervision by a nurse or physician (they make no independent medical decisions), Grand-Aides use telephone protocols and home visits with portable telemedicine to provide transition-from-hospital-discharge and chronic disease management, as well as primary and preventive care. They leverage supervisors, many of whom are current case managers (or care managers), with five Grand-Aides to one supervisor. Their work improves access to needed care and reduces unnecessary emergency, clinic and hospital visits, thus lowering costs.
In studies of primary care Grand-Aides,9 it was shown that (1) patients in a pediatric primary care clinic in Houston (62% of 468 visits), and (2) Medicaid patients in an emergency department in rural Virginia (74% of 402 visits) could have been initially cared for by a primary care Grand-Aide and nurse supervisor, thus freeing physicians from those duties.
A transitional and chronic care Grand-Aide accompanies the patient home on the day of discharge and makes daily visits as needed for those with chronic disease (e.g., congestive heart failure, acute myocardial infarction, pneumonia). Early results reveal a 50.2% reduction in 30-day readmission for patients with congestive heart failure.10 A consortium of institutions coordinated by Vanderbilt University School of Medicine is using Grand-Aides to reduce readmissions. Some other programs use nurses and social workers,11–13 whereas Grand-Aides, working as nurse’s aides, leverage nurses (usually current case managers). Grand-Aides are also func tioning in rural areas with nurse practi tioner (NP) supervision, thus helping to alleviate geographic maldistribution.
The Patient Protection and Affordable Care Act calls for health care worker roles that provide a link between professionals and patients.14 Grand-Aides could fill this role as new members of the team.
NPs and nurses.
An Institute of Medicine report15 and the data supporting it make clear that NPs should be able to practice at the “top of their licensure.” At present, 22 states and the District of Columbia currently allow NPs autonomy in diagnosis and treatment.16 In most other states, it is the organized physician lobby that has stood in the way of progress. With the addition of at least 25 million more insured people under the new health care act, most physician practices are likely to be sufficiently busy that they should welcome NPs.
Additionally, perhaps state boards of nursing should consider examining the scope of practice for certain registered nurses who have given patient care for more than 10 years to allow them to do more (e.g., make some independent decisions concerning issues of management) if they pass tests of competency.
The role of the generalist has changed. Marcus Welby was an outstanding role model for all of us, but, unfortunately, his practice is difficult to reconcile with the current eight-minute office visit. How much of what we now think of as “primary care” can be done by others? What if the team functioned well: patients took better care of themselves, Grand-Aides helped nurses to handle the colds and similar issues, and NPs gave most of the routine care that was really needed? What if primary care physicians practiced at the top of their competencies?
Much is said about coordinating the care of patients. There are two kinds of coordination. The first is to be sure that all practitioners communicate with each other and the patient and do not give conflicting information; this type of coordination can be done by a good nurse. The second type of coordination, however, does require a physician, who must manage the other practitioners involved, especially when they may have conflicting recommendations for ways to treat the patient. If a nurse in the first case and a physician in the second took care of these kinds of coordination, primary care physicians would then mainly be involved with complex patients who required advanced, innovative decision making. Physicians could spend greater amounts of time with these patients and develop important relationships with them. Reimbursement will need to change to reward these longer visits. Progress in this direction is beginning in the patient-centered medical home model, which emphasizes care by a team. In this care delivery model, the need for primary care is the same, but others can take some of the burden. The need for primary care physicians is still vital in this model—but if the model were used more widely, we might not need 45,000 more primary care physicians. Of course, many of the more experienced primary care physicians might not like their new roles.
With the aging of the population and the emergence of new specialties (e.g., cardiovascular genetics), at first blush, we seem to need more specialists. However, the arguments for specialist physicians are not dissimilar from those for generalists. Specialists—subspecialists especially—will need to examine the elements of care that they alone must do and then cede the remainder to specialists, primary care physicians, and NPs. If this were done, we might not need 45,000 additional specialists, but perhaps actually more of them than generalists, as the roles of the current NPs can help alleviate the generalist shortage to a greater extent. This would be good, because the schools cannot deliver them and limitations have been placed on the number of federally funded GME slots.
I define physician-induced demand as inappropriate demand, not directed toward the better care of the patient. The vast majority of what physicians do is appropriately directed. However, physician-induced demand does occur. Three lines of evidence support this concept.
First, in Medicare recipients, research showed that the rate of joint-replacement procedures for chronic hip arthritis varied by as much as a factor of five, and the use of surgery to treat lower back pain varied by nearly a factor of six and was higher in regions with more physicians.17 Other studies have found wide regional variation in the treatment of early-stage breast and prostate cancers and in the use of cardiac procedures.17 Several factors (including patient-induced demand) are responsible for such variation; physician-induced demand is also high among them.18
Second, in one study, when uninvolved persons using decision aids, rather than the recommendations of the treating surgeon, explained the need for surgery, patients chose the procedure less frequently. In another study, there was a 38% reduction in hip and knee replacements using shared decision making.19
Last, a study by the U.S. Government Accountability Office showed that self-referring providers ordered about twice as many magnetic resonance imaging or computerized tomography scans as did providers who did not self-refer.20
The fee-for-service payment methodology is one of the factors responsible for physician-induced demand. In one study, physicians who were paid fee-for-service saw 23% more patients than did those who were salaried.21 Physician-induced demand18 was likely responsible for at least some of this difference. Interestingly, the entire projected shortage of physicians in 10 years is 12%.2 Paying physicians in a different way would not completely eliminate the physician shortage, but it would certainly help, as fewer instances of seeing patients unnecessarily and performing unnecessary procedures would free up physicians’ time to see more new patients and those who really need to be seen.
The other frequently mentioned cause of physician-induced demand is the fear of litigation. It is difficult to tease out the desire for monetary gain from fear of litigation, as both result in increased use of services. In one study, it appeared that the desire for monetary gain was more prevalent.22
If it were possible to reduce physician-induced demand by 12%, I believe the other expenses that accompany that demand (e.g. hospitalizations, procedures, tests) could also be decreased. Such savings, related to avoiding waste, have been projected to occur once the Affordable Care Act is implemented.23 These savings would not reduce access to appropriate care.
As with physician-induced demand, I define patient-induced demand as inappropriate demand by patients. In one study of a Medicaid population of children, 62% of walk-in patients did not need to be seen in the clinic, and 74% did not need to be seen in the emergency department.9 What if we could reduce that demand by even half with improved education and access to providers (such as Grand-Aides and nurse supervisors) and disincentives, like those being used in many states, such as increased co-pays for “unnecessary” emergency department visits? We could consider the medical care associated with unhealthy lifestyle choices24 to be patient-induced demand. For example, if morbidly obese patients reduced weight and people stopped smoking—combined with incentives such as higher taxes and higher deductibles for unhealthy behavior, restrictions on advertising that promotes such behavior, and simple labeling such as using red, yellow, and green labels on foods to tell calories at a distance—the resulting decrease in lifestyle-related illnesses would free physicians and nurses to provide more care to others.25
System Change: Integration
Integration of the work of practitioners, clinics, and hospitals will lead to markedly improved quality and efficiency of care.26 A number of authors27,28 have put forward the framework of the accountable care organization (ACO). This outstanding concept has been adopted by increasing numbers of organizations. I believe that integrated systems, whether formally named ACOs or not, are the way of the future, as I explain below.
ACOs can use methods such as capitation or bundling that permit them to pay practitioners (whether physicians, NPs, or others) using criteria of increased quality and elimination of waste. ACOs can be an incentive to achieve appropriate patient volume (i.e., increased payment for seeing more patients—not more visits per patient—as long as patient satisfaction is high for “my practitioner spends enough time with me”). One stimulus to quality could be to tie the defense of a physician accused of malpractice to national practice guidelines and appropriateness criteria that are applicable to the specific patient in the case. Or the reverse: If a patient is harmed by treatment or a test that clearly is contraindicated in national guidelines, this fact could be used by the plaintiff.
Payment should also be independent of procedure and test volume. Payment should provide total compensation that is roughly similar to the physician’s current compensation, and it should increase at least at the rate of general inflation. Any payment methodology that results in, for example, a 30% decrease in payment for the average physician would not be acceptable. In a “pay for performance” system, the amount at risk should be large enough to matter to the physicians involved, but not so large that physicians who perform adequately lose so much money that they drop out of the system. If performance measurement reveals an unacceptable performer (by criteria agreed on by physician organizations), this should not be dealt with in the payment system. Rather, this person should be put into a remedial program and, if performance does not improve, lose privileges or licensure.
Finally, payment should reflect factors taken into account in the current, resource-based relative value scale,29 such as complexity. Payment should also be based around the “market value” of different types of practitioners—how much a well-informed consumer would be willing to pay for a surgeon, cardiologist, or internist. (This is, of course, difficult to calculate, but if we are to compensate practitioners “fairly,” there will need to be these types of discussions and decisions). Those that do the same work should be compensated similarly: If a physician and an NP perform the same service, they should be paid the same.
Salary with a significant amount (e.g., 10%–20%) tied to performance could reflect the previously mentioned principles of stimulating quality, eliminating waste, and being an incentive for appropriate patient volume. A practitioner’s salary could begin at his or her current average annual income, and then over perhaps a three- to five-year period, migrate to a salary and bonus more reflective of the income of those in the practitioner’s specialty. Fee-for-service could continue, but it would need to reflect the principles mentioned above. For example, the volume of services and tests ordered would need to be monitored carefully, severity-adjusted, and tied to nationally applicable guidelines as well as appropriateness criteria. In some “high-end” concierge-type practices, fee-for-service would likely continue, with the market setting the compensation.
However practitioners are compensated, these principles should be applied by the integrated system. It makes no sense for an integrated system, or even a hospital employer of practitioners, to be paid using a capitation approach and then pay fee-for-service based solely on relative value units generated.
Electronic health records
Integrated systems will, ultimately, be best able to make use of the real promise of electronic health records (EHRs), although it may take another 10 years. Hopefully, by then, each patient will have personalized-decision support suggestions, based on guidelines and appropriateness criteria, that propose the highest-quality and lowest-cost alternatives for management. I envision that by 2024, there should be no billing clerks; any data required for recording the encounter and billing will be generated by the system and based on records of what treatment was given. The interface between patient, practitioner, and EHRs can be as simple as speaking; perhaps practitioners can use electronic eyewear so that they can look at their patients instead of a computer screen. Finally, team members can have instant access to the record and can communicate seamlessly.
We are currently far from this goal for quality via information technology and the expected cost reduction of $81 billion per year.30 We have numerous barriers, chief among them the needs for greater ease of use and interoperability. But the “automobile” (EHRs) has arrived; we will eventually leave our horses at the hitching post. Although it seems a long way off, EHRs should improve physician efficiency by at least 5% to 10%,31 again putting a dent in the need for more physicians.
How Do We Begin?
Public and private payers should develop mechanisms to pay for quality, shared savings, and the team. Mechanisms need to be developed in the current fee-for-service environment to pay team members (e.g., Grand-Aides, nurses, physicians) appropriately and to reward the team for improved outcomes of quality and efficiency; standardized metrics for team success in improving quality and lowering cost must be developed. Ultimately, salary plus bonus should become the prevalent payment mechanism. All states should permit “independent practice” of NPs; if physicians in certain states require more convincing (which they should not, but some probably will), perhaps the regulations could be phased in over two to three years while data demonstrating effectiveness are collected in the state. All states should examine and consider revision of scope of practice for certain experienced registered nurses so they can perform at their maximum competency, as well as develop state certification programs (with ongoing demonstration of competency) for certain nonprofessionals such as Grand-Aides. Physician organizations should begin redefining the roles of generalists and specialists and then testing those roles in teams. State and federal legislation and regulation should facilitate the development of integrated systems, whether they be ACOs or other forms that appear promising. These formal systems will provide environments for change that increase the likelihood of success of the recommendations made above.
What Will We Do?
There are shortages of physicians and nurses, and these will worsen, with increasing geographic maldistribution. However, the shortages may not be as serious as projected if the systems in which practitioners work can be changed to promote integration, with different payment models and leveraging every member of the workforce to perform at his or her maximum competency.
1. Garson A Jr. Perspective: Leveraging the health care workforce: What do we need and what educational system will get us there? Acad Med. 2011;86:1448–1453
7. Felix HC, Mays GP, Stewart MK, Olson M. Medicaid savings resulted when community health workers matched those with needs to home and community care. Health Aff (Millwood). 2011;30:1366–1374
9. Garson A Jr, Green DM, Rodriguez L, Beech R, Nye C. A new corps of trained Grand-Aides has the potential to extend reach of primary care workforce and save money. Health Aff (Millwood). 2012;31:1016–1021
10. S. Craig Thomas. MSN, RN, Grand-Aides Supervisor, University of Virginia. Personal communication, April 16, 2013.
11. Counsell SR, Callahan CM, Buttar AB, Clark DO, Frank KI. Geriatric Resources for Assessment and Care of Elders (GRACE): A new model of primary care for low-income seniors. J Am Geriatr Soc. 2006;54:1136–1141
12. Coleman EA, Parry C, Chalmers S, Min SJ. The care transitions intervention: Results of a randomized controlled trial. Arch Intern Med. 2006;166:1822–1828
14. . Patient Protection and Affordable Care Act, Pub L No. 111-148, Title V, Subtitle D, Sec 5313 2010
15. Committee on the RWJF Initiative on the Future of Nursing, Institute of Medicine. The Future of Nursing: Leading Change, Advancing Health. 2010 Washington, DC National Academies Press
18. Einstadter D, Kent DL, Fihn SD, Deyo RA. Variation in the rate of cervical spine surgery in Washington State. Med Care. 1993;31:711–718
19. Arterburn D, Wellman R, Westbrook E, et al. Introducing decision aids at Group Health was linked to sharply lower hip and knee surgery rates and costs. Health Aff (Millwood). 2012;31:2094–2104
20. United States Government Accountability Office. . Medicare: Higher Use of Advanced Imaging Services by Providers Who Self-Refer Costing Medicare Millions. http://www.gao.gov/assets/650/648989.pdf
. Accessed September 30, 2013
21. Hickson GB, Altemeier WA, Perrin JM. Physician reimbursement by salary or fee-for-service: Effect on physician practice behavior in a randomized prospective study. Pediatrics. 1987;80:344–350
22. Klingman D, Localio AR, Sugarman J, et al. Measuring defensive medicine using clinical scenario surveys. J Health Polit Policy Law. 1996;21:185–217
24. McGinnis JM, Williams-Russo P, Knickman JR. The case for more active policy attention to health promotion. Health Aff (Millwood). 2002;21:78–93