Addressing the Supply Problem

From The Practice July/August 2018
How medicine made space for physician assistants

The lead article in this issue of The Practice highlights the State of Washington Supreme Court’s notable break from traditional legal profession hierarchy with its introduction of limited licensed legal technicians (LLLTs)—an entirely new class of legal professionals who can perform some, but far from all, of the functions of lawyers with relative autonomy. While this innovation is lauded by many access-to-justice advocates, some lawyers are left wondering: What does this new model portend for the future of the profession? Will these new entrants revolutionize what it means to practice law and, in doing so, redefine the lawyer role?

While this is a new development for the legal profession, it has played out before in other contexts, perhaps most notably in medicine. During the past 50 years, physician assistants (PAs)—medical professionals who undergo two to three years of graduate training, are largely regulated by state medical boards, and have many (though not all) of the medical authorities of a traditional doctor of medicine (MD)—have managed to carve out a role of their own alongside that of their physician colleagues. Indeed, PAs have become deeply enmeshed in the country’s health care system as they are trained to perform many of the functions of a physician, be it in hospital emergency rooms, primary care facilities, or specialty practices like dermatology or oncology. While PAs are not full-fledged doctors, and instead work under their supervision, one should not understate their impact. According to the most recent data available from the Bureau of Labor Statistics, as of May 2017, there were more than 109,000 employed PAs in the United States—roughly one-sixth the number of physicians and surgeons.

The history of physician assistants can be traced as far back as the first half of the twentieth century.

PAs present a particularly fitting case study for lawyers trying to make sense of the introduction of LLLTs into the legal profession. Since its inception, the PA role was designed to free overtapped physicians and expand access to health care by introducing new professionals who could perform tasks that were hitherto the exclusive realm of fully licensed doctors. Thus, beyond the broad regulatory parallels between law and medicine—for instance, their professional regulation takes place at the state level—the case of PAs in medicine offers lawyers and other observers of the legal profession a glimpse into how an LLLT-like model has been received and operationalized in the past. With their introduction and rise has come the standardization of their education and licensure, the development of related law and regulation, and, all the while, a deepening problematization of the existing hierarchy in the medical profession. We explore each of these progressions in the sections that follow. Taken as a whole, the trajectory of PAs over the last half century suggests one possible road map for how law could incorporate new actors into the profession to improve access to justice and the delivery of legal services.

Educating a new profession

The history of physician assistants can be traced as far back as the first half of the twentieth century. In the wake of World War II, the United States faced a shortage of physicians across the country, especially in poor and rural areas. To address this shortage, the medical profession actually drew on a role developed in the military. Indeed, early conceptions of physician assistants were intended to emulate the medical corpsman of the U.S. military. During the subsequent decades, the PA model gained an institutional foothold in the United States and, in the 1960s and early 1970s, distinct educational programs began to emerge. These programs were typically situated in the medical faculties of major universities, starting with Duke University in 1965. Harking back to its military origins, the earliest students of Duke’s pioneering PA program were all former navy corpsman, and in 1967 the school produced its first three graduates. Many schools were quick to form comparable programs of their own, including George Washington University, Johns Hopkins University, and Yale University. Currently, according to the Accreditation Review Commission on Education for the Physician Assistant (ARC-PA), there are more than 230 accredited PA programs—primarily based at colleges’ and universities’ schools of health sciences, medicine, or a similar subdivision—producing about 8,000 graduates per year.

PAs present a particularly fitting case study for lawyers trying to make sense of the introduction of LLLTs into the legal profession.

Today, the education and training of PAs is largely standardized through an independent accreditation body—the ARC-PA—that collaborates with the American Medical Association and the American Academy of Physician Assistants, among other professional medical organizations, to produce and maintain its standards of accreditation, much like ABA accreditation standards for law schools (see “Teaching Ethics and Professionalism”). Accredited PA programs tend to span two to three years, and the vast majority now culminate in master’s degrees. By 2021, all accredited programs will award master’s degrees. State regulation has validated the education standards endorsed by the ARC-PA, with almost every state in the country now requiring the successful completion of an accredited PA program (Kansas is willing to consider an alternative training via military service), followed by passage of the Physician Assistant National Certifying Exam to obtain licensure. From there, PA licensure certification and renewal requirements vary by state.

In line with the developing sophistication of PA credentials, many of these programs maintain increasingly specific and stringent prerequisites for admission. Duke serves as a fitting example. Like the majority of PA programs, Duke requires a bachelor’s degree for admissions as well as a whole set of “premed-like” undergraduate course work, including in anatomy, physiology, biology, chemistry, and statistics. Duke also requires GRE scores for admission, as do the majority of programs around the country. Beyond academic prerequisites, and again similar to many PA programs nationally, Duke requires a minimum of 1,000 hours of “patient care experience,” including roles such as an emergency medical technician, nurse’s aide, or clinical research assistant. The requirement of significant health care experience as a condition of entrance into the program traces back to when PA students were largely skilled corpsmen focusing less on the basics of health care and more how to fine-tune and apply their existing knowledge in new practice settings. Today, it effectively narrows the eligible applicant pool while also serving to raise the level of training PA programs can provide—insofar as everyone who enters a PA cohort has some measure of existing training. Like law schools, many PA programs use a Central Application Service for Physician Assistants for its aspiring students, where it requires applicants to report whether they have ever been convicted of a misdemeanor or felony, are guilty of a license or academic infraction, and, if they ever served in the military, were honorably discharged. Despite all these eligibility requirements, PA programs are highly competitive. Duke, for example, had an acceptance rate of approximately 4.7 percent for its class of 2020 (90 students), and while its medical school had an even bleaker acceptance rate of approximately 1.7 percent for its class of 2021 (116 students), these numbers indicate that acceptance into such a PA program is less attainable than many Ivy League graduate programs in other disciplines. Put together, PA programs serve a gatekeeping and screening function for the PA profession—not unlike the role of law schools in the legal profession.

Once a student is admitted, the typical PA program curricula breaks down into two phases: the didactic phase and the clinical phase. The didactic or classroom learning phase includes required course work on subjects like medical science (anatomy, physiology, pharmacology), interpersonal and other patient-centered skills, technical skills according to current practices and technology, medical ethics, research methods, and even instruction on the PA licensure process, regulations pertaining to practice, and background on the profession as a whole (just as law schools offer, and at times require, courses on the legal profession). The clinical phase takes place under the supervision of licensed health care providers and provides practical experience working in different practice settings (outpatient, inpatient, operating room, and emergency room), under different disciplines (family medicine, general surgery, and pediatrics, just to name a few), and with different types of patients (a variety of ages, genders, conditions, and so on). While the specific standards in place have changed over time, any student graduating from a PA program today will have covered all these components.

Accredited PA programs tend to span two to three years, and the vast majority now culminate in master’s degrees.

Putting this into context, the path to becoming a PA is notably shorter—and less expensive—than the one leading to a doctor of medicine degree and a career as a fully licensed physician. A physician’s education requires a minimum of four years of undergraduate education, four years of medical school, and at least three years of residencies and fellowships—a total of 11 years at the low end. Meanwhile a PA’s education takes roughly half the time, often including four years of undergraduate education and two to three years in a PA program, lasting just six to seven years in total.

The cost of a PA program is also far more affordable than medical school. By way of comparison, according to the Physician Assistant Education Program, the average tuition for a full PA program in 2015–2016 was $78,214 for nonresident students at public institutions. With the average PA program running about 27 months, this would correspond to an average annual tuition of $34,762. Controlling for year (2015–2016), student type (nonresident), and institution (public), the average annual tuition at medical school was $52,464—about 34 percent higher. Since medical school typically spans four full years, that would add up to a total tuition cost of $209,856—nearly three times more expensive than the corresponding PA program tuition and roughly twice as long.

Put together, during the last five decades, PA education has gathered structure and developed into a reliable method of training future PAs, even taking on a gatekeeper role for the new profession. A consistent curriculum has emerged across PA programs with the help of a consensus, independent accreditation entity, and the financial and time commitments to pursuing PA credentials are significantly less than the costs associated with becoming a physician. This has all taken form within a fluctuating regulatory picture, which encompasses not just a PA’s education but their responsibilities and limitations once they enter practice. Indeed, to fully appreciate how PAs have come to fill the space they currently occupy in the medical profession, one must consider how regulators—and which regulators—have framed the profession that awaits today’s graduates.

Regulating a new kind of professional

The story of how PAs are regulated has grown in complexity. Initially, the primary basis for their legality was a 1966 opinion from the North Carolina attorney general (the state in which Duke had begun pioneering the country’s first PA program the year prior), which sanctioned medical care that was not performed by physicians “when such care and activities are performed under the orders or directions of a licensed physician,” as documented by health law scholar Ruth Roemer. As interest in the burgeoning profession grew, the pressure to address the lack of regulation around these new medical service providers increased. The question centered on how best to fit the nascent PA role into the existing regulatory terrain.

The path to becoming a PA is notably shorter—and less expensive—than the one leading to a doctor of medicine degree and a career as a fully licensed physician.

Like law, the medical profession has long been regulated primarily at the state level. In the late 19th and early 20th centuries, “medical practice acts” were drafted at the state level to regulate physician activity in the interest of protecting the public from substandard care. These acts often included a combination of establishing regulatory and licensing bodies, defining the practice of medicine, and narrowing the range of professionals who could practice—namely, physicians. The introduction of PAs upset all these existing definitions and regulatory structures, which left it up to states to sort out how to situate a new medical role that was quickly catching on. In one attempt to find national consensus, the U.S. Department of Health, Education, and Welfare (later to be divided into the U.S. Department of Education and the U.S. Department of Health and Human Services) contracted Duke University to formulate model legislation for state legislatures to regulate the newly created PA role.

Harvey Estes Jr. and Reginald Carter outline in a 2005 article how the drafting process for this model legislation proceeded: Duke performed initial research into medical licensure and produced a white paper with the purpose of informing debate at a subsequent series of conferences. These conferences were composed of members of the medical profession, government, academia, and other groups of stakeholders who had meaningful interaction with PAs to that point, including some PAs themselves, and weighed the regulatory options presented in the white paper. Estes and Carter summarize the five possibilities considered:

The “status quo” option was a continuation of the existing policy. Physicians would delegate functions to their assistant, and custom and useage would validate the process over time. This option was seen as leaving both the employing physician and the assistant vulnerable to legal action for improper delegation. Independent licensure of physician assistants was the second option. Difficulty in precisely defining the duties to be permitted was seen as a major problem with this option. The third option was to license the physician or institution that wished to utilize a physician assistant. This was seen as a new function of the Medical Board. The fourth option was to create a new statute authorizing general delegation by physicians. Four states were found to have some features of legislation authorizing general delegation within their medical practice acts. Lack of protection for the public was an identified problem with this approach. The fifth and last option was to create a Committee on Health Manpower Innovations, which would report to the medical board. The Committee would evaluate and pass judgment on new types of health workers, based on the need and the ability of the applicant individual or institution to support and supervise the innovation. The need for representation by all health professions on the new Committee was recognized, but at the same time, it was predicted that each of the representatives would tend to be protective of their own turf.

Two conferences in Durham, North Carolina, reviewed these options in 1969 and 1970 and ultimately agreed that physicians should be authorized to delegate some of their responsibilities to PAs and that, to protect the public, PAs must meet certain requirements with respect to their credentials and operate only within a prescribed scope of work. The resulting model legislation was passed into law in North Carolina in 1971 with minimal changes. By 1974, 37 states had enacted legislation that carved out a space for PAs alongside supervising physicians. Some states used models similar to the legislation drafted in North Carolina, while others adopted more defined versions, parsing out the role of the PA in more granular detail. Some simply added an exception to the state’s medical practice act whereby PAs were exempt from certain restrictions provided their supervising physician was licensed. Now PAs practice in all 50 states, and in many cases the regulation has evolved to take a more hands-off approach, leaving much of the detail surrounding the role to be set at PAs’ individual practice settings with their supervising physicians. While differences in regulatory regimes remain at the state level to this day, it is also worth stressing that the regulation of PAs is delegated to the same boards that regulate physicians in all but a small number of states. In other words, PAs and MDs are most often governed by the same medical boards.

The introduction of PAs upset existing definitions and regulatory structures, which left it up to states to sort out how to situate a new medical role that was quickly catching on.

With the PA role established, the larger vision for addressing access-to-care issues could begin to fall into place. Again, the genesis of the budding PA role was a nationwide shortage of physicians, especially in rural areas. With this newly sanctioned class of medical professionals (including not just PAs but also nurse practitioners [NPs] and eventually certified nurse-midwives) came the potential to inject trained practitioners into the health care workforce to relieve the strain on physicians and provide services that had been lacking or nonexistent—not unlike the vision for the LLLT model in the legal profession. To that end, Congress passed the Rural Health Clinic Services Act in 1977, establishing rural health clinics (RHCs)—facilities offering specific medical services covered by Medicaid and Medicare in areas that the census defined as “nonurbanized” and that the Department of Health and Human Services identified as suffering from a shortage of certain health care services. The Act (along with its various revisions in the years since) also provided space for PAs and NPs to practice with or without a physician physically present, aligning the work of these new professionals with the goal of servicing underserved populations. Moreover, just as there must be at least one physician staffed at each RHC, so too must there be at least one PA or NP. Today, some RHCs are even owned and operated by PAs.

The role of PAs has, of course, evolved across states in the ensuing decades—and numerous iterations of model legislation have evolved along with it. There are countless instances where PAs’ scope of practice and general autonomy has shifted with corresponding legislation and policy. Below we highlight three prominent areas where PAs have broadened their ability to practice medicine: prescription authority, reimbursement authority, and the physician relationship.

Prescription authority. Over time, the ability of PAs to prescribe medication has expanded across the country. As recently as 1992, less than half (47 percent) of states authorized PAs to write patients prescriptions for controlled substances. By 2000, more than three-quarters (76 percent) of states permitted PAs to directly prescribe medication. Today, 43 states and the District of Columbia allow PAs to prescribe controlled substances in schedules II–V—schedule V being those controlled substances deemed to have the lowest potential for abuse (a strong cough medicine, for example), schedule II being those with the highest potential for abuse while still serving an acceptable medical application (such as Adderall or Ritalin), and schedule I being those deemed to have not only a high risk for abuse but no acceptable medical application (such as LSD or heroin) and thus may not be prescribed by anyone. Six states allow PAs to prescribe controlled substances schedules III–V, and only one state, Kentucky, does not currently permit PAs to prescribe controlled substances of any kind (though there is local pressure to change this). PAs, through organizations such as the American Academy of Physician Assistants, are currently advocating for their ability to appropriately prescribe all available medication throughout the United States, arguing that full authority would facilitate more extensive medical care for patients that PAs are well trained to provide.

Reimbursement authority. Another major regulatory fault line in the advancement of the PA profession has been the recognition of—and compensation for—their services by insurers and other third-party health care payers. In this respect, states have been slower to enact legislation and policy to require payment for services billed directly by PAs rather than routed through their supervising physicians. Symbolically, reimbursement authority signals the PA’s relative independence, declaring unambiguously that the PA provided the service rather than the supervising physician. On a practical level, it more accurately reflects who is performing each service, a distinction that cascades onto numerous other considerations like performance evaluations and, of course, how professionals are compensated. Today, most services provided by PAs throughout the country are covered by Medicare and Medicaid, although the compensation does not always match that of physicians performing the same service (Medicare, for example, typically covers only 85 percent of a physician’s fee for claims submitted using a PA’s identification number).

Physician relationship. Both prescription and reimbursement authorities relate to a fundamental feature of the PA career: their relationship with their supervising physician. PAs must have a supervising physician wherever they practice in the United States—the same as the first PAs in the 1960s—but what that supervisory role entails varies. According to the American Medical Association, the relationship is worked out at the practice level in most states, where it might be more flexibly negotiated by the supervising physician and PA. For example, the relationship could be as minimal as checking in with a mentor to work through problems and ask for advice or as rigid as obtaining physician approval for various aspects of daily patient care. That being said, it is worth noting that other regulatory factors will inevitably impact those relationships, such as cosignature requirements (20 states require that physicians cosign PA medical charts) and ratio requirements (39 states limit the number of PAs that any one physician can supervise). Nevertheless, the way this relationship operates is trending toward flexibility, with very few states still regulating the specific nature of “physician supervision.”

The Parallel Rise of Nurse Practitioners

Any conversation about evolving roles in medicine, especially those whose practices are increasingly autonomous, would not be complete without discussion of nurse practitioners (NPs). Like PA programs, the earliest NP graduate programs trace back to the 1960s with a similar focus on providing care to rural and other underserved populations. But, instead of descending from the physician model, NPs were cut from a different cloth. Whereas PAs are trained in a course of study that in many ways resembles that of physicians (for example, with a particular focus on the treatment of disease), NPs come from the pedagogy of nursing and midwifery (which, many have noted, places greater emphasis on preventive measures and the patient’s overall well-being). NPs are even regulated differently at the state level. While PAs are usually regulated by the same boards as physicians, NPs are regulated by boards of nursing in all states and the District of Columbia.

Philosophical and regulatory nuances notwithstanding, there is plenty of overlap in the roles performed by NPs and PAs. NPs diagnose and treat conditions, counsel patients, and manage all stages of diagnostic tests. Also like PAs, NPs have faced an uphill battle gaining autonomy and recognition for certain features of medical practice. According to the American Medical Association, some degree of physician involvement is required for NPs to prescribe medicine in most states, and in some states, they cannot prescribe schedule II controlled substances. Reimbursement is also an issue for NPs, whereby insurers and third-party payers may reimburse them for their services at a lower rate than if those services were performed by a physician. Thus, as many observers have suggested, NPs and PAs appear to have a common cause as they struggle for greater independence within the medical profession.

The new normal for professionalism in medicine—and cause for concern?

These developments raise important questions about the quality of new entrants to the medical profession that in many ways mirror those posed by lawyers observing the introduction of LLLTs. How has the introduction of new roles impacted patient satisfaction? Perhaps even more important, how has it affected quality of care? Has it actually improved access to medical services? Thankfully, with the benefit of five decades of hindsight, the medical profession does not lack for empirical studies to explore these questions.

Just as regulation carves out space for PAs and NPs to operate with increasing autonomy in the medical profession, it also sets limitations to their respective scopes of practice.

While some in the physician community object to the increasing autonomy of PAs and NPs and the creation of a two-tiered system—the argument often being that these roles require less training and therefore PAs and NPs are less equipped to deal with the full spectrum of medical issues—by and large patients are not complaining. A 2005 survey of Medicare beneficiaries found overall patient satisfaction with physicians, PAs, and NPs to be “statistically indistinguishable from each other, regardless of the patients’ age or gender.” Perhaps more to the point of physicians’ concerns, data on the quality of care suggests that patients may be slightly better off when the appropriate services can be performed by PAs and NPs rather than physicians. A 2017 study measuring quality of care in health centers found that of the nine outcomes measured—which included physical examinations, depression treatment, smoking cessation counseling, statins for hyperlipidemia, health education and counseling services, imaging services, medications, return visits, and physician referrals—seven revealed no statistically significant difference between care administered by PAs, NPs, or physicians. The remaining two outcomes (smoking cessation and health education services) did show statistically significant differences in quality of care, with NPs performing better than PAs and physicians in the former and both NPs and PAs outperforming physicians in the latter. Other research projects on quality of care have similarly found little significant difference between the care provided by physicians and the care provided by PAs and/or NPs, be it a cross-sectional analysis of HIV care or a randomized control trial measuring primary care outcomes, among many others.

Finally, a 2003 study published in the Annals of Family Medicine on California and Washington State medical professionals found that a higher percentage of PAs in California worked in rural areas (21.7 percent), vulnerable population areas (47.7 percent), and health professions shortage areas (35.2 percent) than any type of physician; in Washington, whereas the proportion of PAs serving in these areas did not always surpass that of certain types of physicians, there was a higher percentage of PAs in rural areas than any type of physician (28.8 percent) and the proportion was comparable to that of the best-represented types of physicians in vulnerable population areas (PAs, 50.3 percent; general internists, 54.5 percent) and in health professions shortage areas (PAs, 42.1 percent, surpassed only by family physicians at 43.5 percent). In short, PAs and NPs are not just providing competent care—they are honoring their unofficial founding principle of serving those in otherwise underserved, rural areas.

Warnings that PAs and NPs are not as well equipped as MDs to deal with medical issues often overlook the other side of the coin when it comes to regulating these new models of medical professionals. That is, just as regulation carves out space for PAs and NPs to operate with increasing autonomy in the medical profession, it also sets limitations to their respective scopes of practice. Case in point, notwithstanding all the advancements PAs have made during the past 50 years, it remains true that they must be formally supervised by physicians. Moreover, unlike PAs, physicians still enjoy greater autonomy of practice and, by contrast, can set up their own practices with or without PAs. Similarly, norm-based constraints are built into medicine more broadly, for instance, when cases are elevated—whether from an PA to an MD or even from a nonspecialist MD to a specialist. Viewed in this way, criticisms that the entrance of PAs and NPs into medicine risks devaluing physicians or creating the risk of a slippery slope may be overblown.

Prognosis for the legal profession

What should lawyers take away from all this? The introduction of PAs and NPs has problematized the existing medical profession hierarchy by revealing that certain services previously thought to require an MD may be competently performed by lower-credentialed professionals. It did not, however, upend the medical profession—one in which physicians have generally seen steady and even increasing wages. Neither has the addition of PAs and NPs created a two-tiered system of care as evidenced by numerous studies tracking patient satisfaction and quality of care.

With the benefit of five decades of hindsight, the medical profession does not lack for empirical studies measuring the effectiveness of PAs and NPs.

This, of course, was no accident. Members of the profession had to build these new models, and early examples of PA regulation and the proliferation of PA training programs succeeded in large part because they were deliberate, collaborative endeavors. Moreover, PA education and state regulation have required constant adjustment as the roles of PAs and their potential in the medical profession crystalized over time. The result is almost 110,000 more professionals in medicine today who are more likely to work in rural areas and areas hit hardest by shortages than many of their physician colleagues.

In considering new legal professionals and the potential value they could add to the profession, lawyers, educators, and regulators will have to be clear in establishing a goal and in setting those professionals up to succeed at it. As the lead article notes, while the LLLT model is not perfect, it shows that, at a minimum, there is potential in unbundling certain legal services and diversifying who can provide them. Lawyers, like doctors, have historically done a lot to benefit those around them through their professional efforts and expertise—the question is over whether they still need to be the only ones and whether a professional truly needs no less than those credentials to competently serve clients. Perhaps new legal professionals like LLLTs will not be as transformative as a civil Gideon, but new models should be considered worthwhile if they provide sustainable relief for those who would otherwise be discouraged or shut out from the justice system.