The Affordable Care Act (ACA)1 has brought new challenges and new opportunities to academic health centers in the United States. With the advent of the ACA, one of these challenges has been continuing to maintain access to specialty medical care as the U.S. health care system shifts its focus toward ensuring access to primary care. That is, as more patients gain health insurance coverage and begin accessing primary care, how will current budget levels be able to pay for the higher-tier care and increased specialty consults that these new patients will likely require?
The Royal College of Physicians and Surgeons of Canada has an intriguing idea which may offer a way to answer this question. To provide specialty care in rural Canada, where the population density is unable to support specialists and there is patchy access to specialty care, the Royal College created a new training pathway called the Areas of Focused Competence (Diploma) Program (AFC Diploma Program).2
In addition to the traditional residency experience, Canada offers additional one- to two-year discrete subspecialty training options. Practitioners receive additional training in a subject area to further develop their competence, or skill set, beyond that of a graduating resident but short of a fellowship-trained, board-certified subspecialist. These expert–generalists are deployed to rural regions (and other areas) to help bridge the gap between primary and specialty care providers. Officially, the Royal College notes,
Among the many challenges currently facing postgraduate medical education, one ongoing tension is how to best support the growth and advancement of specialty education with national standards of excellence, while trying to ensure an appropriate blend of generalist and specialist physicians to meet the health human resource needs of Canadian society, and without causing harm to the existing system.2
They created the AFC Diploma Program to fulfill these objectives.
As of April 2014, the program offered diplomas in 16 areas: adolescent and young adult oncology, adult cardiac electrophysiology, adult echocardiography, adult thrombosis medicine, aerospace medicine, brachytherapy, child maltreatment pediatrics, clinician educator, cytopathology, hepatology, hyperbaric medicine, interventional cardiology, solid organ transplantation, sport and exercise medicine, transfusion medicine, and trauma general surgery.2 AFC Diploma Program candidates can hail from any number of prior primary training areas.3
According to a representative of the Royal College, each diploma program is initiated by a communication from an existing group representing that clinical discipline.4 The Royal College’s Committee on Specialties (COS) which
has as its primary function oversight of the recognition of disciplines in Canada … receives applications from interested groups—typically National Specialty Societies or existing committees related to the discipline, called Specialty Committees and overseen by the Royal College. These applicant groups are welcome to come forward to the COS with proposed disciplines and, through an application consultation and review process, the COS considers them.4
It is important to note that “the Royal College does not independently choose the disciplines which should be recognized—this process is instead driven by the community of professionals.”4
Not all of these diploma programs are open to generalists, but many are, and most have multiple pathways to enrollment. For example, a holder of a diploma in adolescent and young adult oncology could have a primary specialty in pediatric hematology/oncology, adult hematology, medical oncology, or radiation oncology. Adult echocardiography has a broader range of feeding disciplines, including cardiology, anesthesiology, internal medicine, diagnostic radiology, cardiac surgery, and critical care medicine. A current roster of eligibility requirements has been posted.3
A Canadian Import?
The Royal College’s model of responding to the geographic and pedagogical challenges of Canadian medicine can help meet the fiscal goals and access needs of U.S. medicine as we seek to broaden coverage in a fiscally parsimonious fashion. Expanding the reach of primary care practitioners into the realm of more advanced subspecialty practice could be a way to enhance both access and quality in a cost-effective fashion. Access would be enhanced for patients who would otherwise not have a ready opportunity to see a subspecialist. Quality would be improved because these expert–generalists would be better able to provide treatment than their colleagues who have not received such targeted training.
As I envision the model, expert–generalists would continue to do primary care and serve alongside other primary care providers so as to provide ready access to their services and to ensure that their practices do not morph into subspecialty work for which they were not fully trained. So embedded in primary care, their special skill set could be called on only when needed. At other times, they could serve as an educational resource for their special area of focus. A mix of these expert–generalists could elevate the level of practice in primary care settings that did not have ready access to continuing medical education and serve as a role model for other physicians to emulate. Although these expert–generalists cannot replace board-certified subspecialists, they can help identify patients who need the next tier of care and make appropriate subspecialty referrals, ensuring that these consultations are used effectively and efficiently.
Primary care practitioners could elect to use these intermediate providers to provide a discrete clinical service like an echocardiogram or to provide direction in a case that did not seem to warrant the care of a subspecialist. In both situations, the expert–generalist would save the patients the trouble of seeing a subspecialist who might have limited availability or whose services necessitated travel which would impose a burden on the patient. The use of an expert–generalist would not be a mandated gate-keeper to more specialized care, but a collegial resource available in a practice community. Ethically, referrals to either an expert–generalist or subspecialist should be at the discretion of the responsible physician.
The expert–generalist model could also be expanded beyond direct patient care to other areas that could help enact the structural reforms required by the implementation of the ACA.1 For example, a similar model could be employed for additional training in preventive medicine, cost-effectiveness of care, population health, or clinical ethics. Physicians who work in these policy domains will bring pragmatic (practice-based) insights to health policy and potentially add immeasurably to health care policy activities.
Training expert–generalists is also a cost-effective solution because the infrastructure for additional training already exists. The Accreditation Council for Graduate Medical Education could certify residency programs in primary care to offer the option of an additional year of training in a subspecialty area. During a fourth year of residency, an internist might learn how to do echocardiograms or hyperbaric medicine. Such cross-training during, and after, formal residency training is a good way to maximize society’s considerable investment in graduate medical education (GME) and sustain the Hippocratic ethos of lifelong learning.
Equipped with these additional competencies, the internist from the above scenario could serve a broader swath of the population and provide care without sending patients to fully trained specialists whose workups would likely be more costly and often not necessary to address the clinical problem, reserving the considerable expertise of these specialists for more complicated cases. In this way, true subspecialists—who are in short supply in underserved areas—could better meet the needs of patients who need their ministrations. This would lead to increased physician productivity, which Kirch and colleagues5 suggest is one way to address the looming shortage of physicians in the wake of health care reform and expanded access.
Thus, a properly trained expert–generalist would benefit both the quality of the patient experience and the bottom line. When the utility of referring a patient directly to a subspecialist is in question or when there is an obvious need for a specialized study, such as an echocardiogram, an initial detour to an expert–generalist could help triage those patients who do need more specialized care. Patients who went on to see specialists for in-depth workups because the expert–generalist referred them would have a higher pretest probability of a diagnosis requiring specialty care than those directly referred by a primary care physician without additional expertise.
The risk stratification of this model would make for a more prudential use of scarce resources, reduce false-positive diagnoses, limit unnecessary treatment, and keep the locus of care closer to a primary care ethos—all factors in the patient’s interest. And if done properly, the evaluative process offered by expert–generalists could yield more interesting specialty consultations for specialist colleagues, decreasing the specialist’s engagement with cases that do not require his or her greater expertise.
Evidence of Success
In British Columbia, more than 75% of patients seen by primary care physicians for an “orthopaedic” problem, and identified by the primary care physicians as potentially needing surgery, did not need to be seen by an orthopaedic surgeon after evaluation by a physiotherapist.6 When patients were referred directly to a surgeon by a primary care doctor, operations were necessary in 22% of cases; when patients were triaged by a physiotherapist before being referred to a surgeon, 91% needed surgery, making better use of the surgeon’s time. All patients in the study were either “satisfied” or “very satisfied” with the process.
The study’s authors suggest that this model could potentially have a “considerable impact on orthopaedic wait times in Canada by minimizing unnecessary referrals; the model could also promote timely and conservative management of non-surgical conditions by physiotherapists.”5 Although this example does not directly speak to the expert–generalist model, as the triage clinician was a physiotherapist, it does demonstrate the value of bringing additional expertise into the primary care setting to decrease unnecessary use of subspecialty referrals. The success of this program suggests that a primary care doctor with additional exposure to physiatry could similarly affect costs and patient satisfaction.
There are other, less concrete benefits too. It is hard to be a generalist physician and keep on top of all that an internist is expected to know. It is a field plagued by higher patient volumes, lower pay, and burnout,7 especially in critical midcareer years.8 Diversifying and focusing aspects of the generalist’s practice space can round out one’s activities and make work more rewarding, improving all areas of the doctor’s practice experience. As Okie9 has demonstrated, innovation in primary care can mitigate burnout.
The option of training as an expert–generalist may also help to develop and retain more high-quality primary care doctors by providing additional opportunities for income with these additional qualifications. This would help close the large gap in income levels between specialist and primary care providers in the United States and bring value to any additional expenditures.
From an educational point of view, training as an expert–generalist could help promote self-directed learning strategies as adult learners seek new knowledge.10 This would help organize one’s scholarship and make keeping up with the required knowledge a bit less of a challenge.
Logistics and Implementation
So, how could we establish this model? The simplest way would be to add an additional year to residency training and let trainees obtain the expert–generalist designation during a period of their lives that is already devoted to GME. Some might worry that this would extend training by a year. New ideas about the state—and indeed the fate—of the fourth year of medical school actually complement this proposal. The fourth year of medical school is considered by some to be largely a wasted year.11,12 Perhaps if undergraduate medical education were shortened, GME could be extended one year for expert–generalist education without affecting the overall length of postgraduate medical training.
A second option, drawn from the progressive educational wisdom embedded in Plato’s Republic, would be to offer this additional training to mature doctors who were ready and willing to further develop their skills in order to pursue interests that may have been latent since their student days.13 This could be accomplished part-time, in a modular fashion, and could be supported by state or federal educational subsidies or grants, because such an expansion of the workforce would be in the public interest. These GME subsidies would be directed to academic health centers so as to allow them to expand their training mission.
One model for how to operationalize education for busy practitioners comes from Spain, where the government provided resources for midcareer physicians and other clinicians to obtain a master’s degree in bioethics at Madrid’s Universidad Complutense.14,15 Local hospitals chose students for the program and relieved them of some responsibilities so that they would be able to attend weekly sessions in Madrid. When the students completed their studies, they assumed leadership positions in their hospital ethics committee or institutional review board.
Whatever the pathway to these additional competencies, through the extension of residency or through rotations or course work later in life, candidates for the expert–generalist designation would need to be evaluated. This would require the cooperation of current subspecialists who are the keepers of their field’s knowledge, skills, and attitudes. Ideally, the U.S. specialist community would follow their counterparts in Canada, who did not view the expert–generalist as competitive but, rather, as a means to make better use of scarce resources.
Beyond the Canadian example, palliative medicine and hospital medicine (hospitalism) provide a historic precedent for what is suggested here. Both are derivative areas of practice that require prior board certification with additional skills added onto that strong base.16,17 In both cases, the initial practitioners in these fields were physicians who cultivated additional expertise before there was a formal board (as in palliative medicine) or version of recertification (hospitalism).
Potential Objections and Implementation Challenges
Some might argue that the Canadian model works because these expert–generalists are deployed to rural areas. Others might assert that the presence of this additional training will increase costs over what primary care physicians currently make, thereby driving up expenditures and adding complexity to physician fee structures. A response to this critique might be to restrict expert–generalists to rural (and otherwise-underserved areas) and/or to limit payments to these practitioners to be less than payments to board-certified subspecialists.
Implementation of this proposal should be subject to rigorous analysis and prudential implementation. Demonstration projects, like the one cited above,4 could be done to assess the impact of altering physician mix on the basis of quality, cost, and access. These studies should be undertaken in diverse markets with varying percentages of generalist versus specialist physicians.
Current physician workforce data indicate that the need for medical services is not uniform and that it varies from state to state.18 Across the nation there were 225.6 active physicians per 100,000 population, ranging from a high of 321.4/100,000 in Massachusetts to a low of 164.4/100,000 in Mississippi. These data correlate with 131.9/100,000 primary care doctors in Massachusetts and just 63.4/100,000 in Mississippi. Similarly, Massachusetts had 231.3/100,000 non-primary-care physicians as compared with Mississippi’s 101.0/100,000.19
It may be the case that this proposal would be more beneficial in areas that have poorer access to specialist care and counterproductive in regions where there is relative scarcity of primary care providers. In the latter case, diverting the primary care workforce from their primary mission could decrease the availability of generalists. For example, in urban settings with a high density of specialists, there may not be as pressing a need for expert–generalists as there is in areas where access to specialty care is only available at a distance from the patient’s home and with a delay.
Finding the correct balance will be important because Association of American Medical Colleges projections indicate that the shortage of primary care versus non-primary-care specialties will be on par for the next 10 years: 29,000 versus 33,100 in 2015; 45,400 versus 46,100 in 2020; and 65,800 versus 64,800 in 2025.20
This trade-off between access to primary care and access to specialty care is something of a zero-sum gain. The task of demonstration projects in differing markets would be to find a pareto-optimal balance where the local deficiency is rectified with minimal harm to the opposing need. Ultimately, data obtained from such studies and demonstration projects could help inform regional and national goals for the generalist–specialist mix that maximizes access to both primary and specialty care. It seems that such analyses would be constructive projects for both the Agency for Health Care Research and Quality and the Patient-Centered Outcomes Institute (PCORI), which could look jointly at health outcomes and patient satisfaction when expert–generalists are included in the health care delivery system. These data could also highlight differential physician workforce shortages across the United States. To date, the Royal College of Physicians and Surgeons of Canada has not yet collected workforce data on the AFC Diploma Program because it is too early in the implementation cycle to assess such implications.4
In addition to its impact on patients’ access to both specialist and generalist care, this proposal would also need to be assessed for its impact on quality. I have argued that a properly trained expert–generalist could enhance triage and thus improve access to specialty care. This would enhance quality. But a more nefarious outcome in which expert–generalists are either practicing beyond their competence or insufficiently active in their subdiscipline to provide quality care would also need to be considered.
There is considerable evidence that surgical centers with more experience have better outcomes for complex surgical care than do centers with less experience. This is an important caveat to consider, but not a completely analogous situation. The types of additional services that would be provided by expert–generalists are neither at the same level of complexity nor meant to replace the provision of complex care. Indeed, it could be argued that an enhanced triage ability by intermediate providers like expert–generalists could enhance referrals to high-volume providers of complex care. A recent article describing the differential access of rural versus urban populations to sophisticated endovascular procedures demonstrated that, compared with urban patients, rural patients had increased referrals to high-volume hospitals and had superior outcomes following abdominal aortic aneurysm repair.21 This suggests that informed and timely referrals determine access to high-volume providers of complex surgical care. The proposal offered here should help facilitate this process by inserting additional expertise in intermediate-level providers, thus improving access to quality care.
To be sure, expert–generalists will need to have both sufficient patient volume and a clearly circumscribed area of practice to provide quality care. The former is necessary to maintain their skills. The latter point is perhaps even more critical: Expert–generalists need to know when they must refer patients and seek the skills of a fully trained specialist. The demarcation of practice boundaries and the need for the expert–generalist to appreciate what he or she does not know must be an essential part of the additional training and any recertification process. This reflective ability should be consistent with their skill sets as primary care doctors accustomed to seeking the advice and assistance of physicians with more specialty skills. Nonetheless, with the addition of some specialty skills comes the possibility of overconfidence and even hubris, which must be held in check.
Expert–generalists must learn to be reflective about their new abilities and use their new knowledge wisely, most importantly in appreciating their limits. As Oliver Wendell Holmes, Sr.—the physician father of the famous jurist—wisely wrote:
The best part of our knowledge is that which teaches us where knowledge leaves off and ignorance begins. Nothing more clearly separates a vulgar from a superior mind, than the confusion in the first between the little that it truly knows, on the one hand, and what it half knows and what it knows, on the other. That which is true of every subject is especially true of the branch of knowledge which deals with living beings.…22
Dr. Holmes, a poet, anatomist, and Harvard Medical School professor and dean,23 wrote these words in 1862 in a volume entitled Border Lines of Knowledge in Some Provinces of Medical Knowledge. It is an apt title as we seek to redraw some of the traditional boundaries between generalists and specialists and to add an intermediate layer of practice to better meet the often-competing imperatives of quality and access.
With these caveats in mind, U.S. policy makers should look to emulate our Canadian neighbors to the north and expand training options for physicians as they prepare to care for growing numbers of patients who will need a favorable blend of primary and advanced care as we carry out the implementation of the ACA.1 Subject to careful analysis, expert–generalists might be just what the doctor ordered.
Acknowledgments: The author acknowledges the assistance of Lisa Gorman of the Royal College of Physicians and Surgeons of Canada and acknowledges the helpful comments of the journal’s peer reviewers.