Characteristics of Postgraduate Physician Assistant Residency/fellowship Programs: A National Survey

Introduction: The development of postgraduate programs for physician assistants (PAs) began in 1973 and by 2020 there were approximately 100 programs spread across a broad range of medical and surgical disciplines. An assessment of these programs was undertaken. Method: A non-experimental, descriptive research study was designed to obtain information on the characteristics of PA postgraduate education programs in the US. The source of information was from surveyed members of the Association of Postgraduate Physician Assistant Programs. Questions were drawn from consensus discussions. Programs that were operational in 2020 were eligible to participate. Results: Seventy-two programs were invited to the survey and 34 replied. They are geographically distributed across the US in 13 states. The respondents represent a wide range of medicine: surgery, emergency medicine, critical care, orthopaedics, hospitalist, psychiatry, oncology, primary care, pediatrics, and cardiology. Most programs are associated with an academic medical center. The curriculum includes bedside teaching, lectures, mentorship, assigned reading, procedures, simulation, and conferences. The PA fellow serves as house ocer alongside physician residents and fellows. An average program length is 12 months and awards a certicate. Stipends for PA fellows are $50,000-80,000 (2020 dollars) and benets include paid time off, health and liability insurance. About half of the programs bill for the services rendered by the PA. Over 90% of graduates are employed within two months of fellowship completion. Conclusion: A trend is underway in American medicine to include PAs in postgraduate education. PA fellowships occur across a broad spectrum of medical and surgical areas, as well as diverse institutions and organizations overseeing the programs. Most are in academic medical centers or teaching hospitals. This study expands information on PA fellowships and their operation.


Introduction
Postgraduate education for American physician assistants (PAs) has been underway for half a century. Center Emergency Medicine PA Residency followed. Other postgraduate programs were created in rural medicine, neonatology, and primary care (3). By the new century postgraduate programs had broadened into many areas of medical and surgical specialties. For 50 years a trend has been underway to adopt a postgraduate model of PA education (4). The Association of Postgraduate PA Programs (APPAP.org) website lists over 100 programs spanning 32 tracks -some within the same institution (5). Additionally, some programs have been paused during the COVID-19 pandemic of 2020.
Postgraduate PA training programs have been insu ciently researched and more clarity is needed about this aspect of medical education (5). The notion that PAs can supplement house o cers (i.e., physician residents or fellows) is growing as shown by the number of new and established programs. However, little information has been advanced in terms of the broad characteristics of PA postgraduate programs.
Informing administrators and medical directors about this trend is warranted. Documenting when, where, why, and how PAs and nurse practitioners (NPs) are used in American medicine is important to health economists and workforce planners if national goals of access to care are to be met (6). The principal aim of this study was to address this change in the organization of medicine. To address this, a survey of PA postgraduate education programs was undertaken.

Method
A non-experimental, descriptive research study was designed to obtain aggregated information on the characteristics of PA postgraduate education programs in the U.S. The list of programs was drawn from the APPAP membership (APPAP.org). The APPAP website contains information on each program, the sponsoring institution, a contact, degree awarded, and class size. Additional information is gained through reaching the program's website.
Questions were developed based on the author's experience with postgraduate programs and relevant literature. Through an open, iterative process, the questions were re ned with APPAP board member feedback. Fifty-ve questions were selected. The survey instrument was uploaded in Survey Monkey©, an on-line social science survey rm. The web-based questionnaire invitation was sent to active members of APPAP during October 2020 and two separate reminders followed. Each question was singular and not algorithm driven. Responses were aggregated and descriptive statistical analyses were utilized through a statistical package built into the survey software. Text space below the questions provided for impromptu comment by the respondent. The estimated time to complete the questionnaire was 17 minutes. Publicly available information supplemented some of the results.

Results
In total 72 programs were sent a set of questions along with an introduction and an invitation for participation. Thirty-four programs responded to all or part of the questions (47% rate of return). Their distribution is spread across the US in 13 states. Most programs are associated with an academic medical center. On a few occasions, the respondent skipped a particular question (with per question responses varying between 29-34).
All 34 postgraduate programs are focused on the education of PAs, and 12 (35%) include nurse practitioners (NPs). The duration of postgraduate PA education is 12 months (79%), with two at 18 months, and one two-year program. Some programs include research and award a postgraduate degree (7).
The techniques used by programs in role development varies. The most common education delivery methods are bedside clinical teaching, Grand Rounds, professional conferences, case studies, in-person didactic instruction, core discipline rotations, simulation, and other (Fig. 2) Beyond a focus on the education in learning about the particular medical or surgical specialty involved, the program incorporates a wide variety of skill development activities such as procedural training, roles in teambased care, interprofessional communication, and other modalities of education. Nearly all programs provide continuing medical education credits to trainees for maintenance of national certi cation and state licensure. All PAs must be licensed in the state where employed.
Evaluations and assessments of PA fellows incorporate various methods. The more common ones are by the one-on-one observation by a mentor or preceptor, through regular reviews with the program director or supervising (e.g., attending) physician, and by achieving clinical competence (Fig. 3).

Institutional A liation
Administratively half (55%) of postgraduate programs reside in medical schools, with (40%) in health science colleges, and (5%) in a nursing school. Three-quarters (76%) of the hospitals that sponsor PA postgraduate programs are a liated with an academic medical center, the others are multi-hospital systems such as The Mayo Clinic (located in three states) (15%), community hospitals (6%), and one in an ambulatory clinic. Almost all PA fellowship programs (97%) award a certi cation of completion or a diploma (3%). The combined Army/Air Force-Baylor program awards a Doctor of Science (DSc) for active-duty members and Arrowhead Orthopedics in collaboration with University of Lynchburg offers a Doctor of Medical Science (DMSc) track (7).

Nomenclature:
PAs in postgraduate training are referred to as "Fellows" (59%) or "Residents" (41%). One respondent commented, "We refer to them as fellows, but our program is HRSA grant-funded which is NP only and requires the name resident." The majority of programs (88%) reported that the PA fellow works alongside the physician residents/fellows in the same specialty.

Applicant pool:
Most programs (88%) draw on a national applicant pool with a few relying on regional or local applicants. All PAs accepted into a program are graduates of an accredited program and all are nationally certi ed by the National Commission on the Certi cation of Physician Assistants (NCCPA). Entry into a fellowship most commonly occurs within two years of graduation.

Page 5/10
Employment of PA fellowship graduates: Upon graduation, the respondents reported that 96% were employed within two months. When asked about the demand for graduates from the fellowship program, (78%) reported the demand was "high." Administration All of the postgraduate programs reported a PA administrator. One PA reported being a full-time administrator of multiple tracks, with the remainder varying from 4 hours to 20 hours a week (63%). Most  Organizational costs associated with the PA Fellowship: The majority of costs associated with the PA fellowship were derived from the medical group (42%) or the hospital system (46%) with the remainder from graduate medical education or a private donor (12%).
However, program administrative costs (excluding resident salary and bene ts) are unknown.
Billing for PA Fellow services PA trainees enrolled in postgraduate education are NCCPA-certi ed and licensed in the state where the program resides and are eligible for billing of their services. Billing is done by the hospital, medical group, or academic medical center. Not all the respondents answered this question but of those that did, 23 of the 29 programs (79%) reported billing for services provided by the PA fellow. Additionally, four programs said they intended to begin billing for services rendered by their PA trainees. It should be noted that not all PA fellowship programs bill rst assist services. Billing for these services typically requires an attestation from the surgeon that no quali ed resident is available to assist with eligible operative cases. Medicare reimbursement for a PA/NP assisting in surgery is 13.6% of the primary surgeon's allowable fee and some surgeries are 'restricted' from reimbursement.

PA Fellowship compensation and bene ts:
All programs provide a stipend for the PA fellow. The amount varied between $50,000-80,000 (2019 dollars) (Fig. 5). Not reported were the wages for federal employees (e.g., military and VHA) which are government scheduled and at a higher rate than civilian stipends. Employment bene ts (not reported) include paid time off (2-4 weeks), health insurance, CME, liability insurance, and clothing/uniforms.

Value of APPAP Membership
When asked about satisfaction with APPAP, the responses were mixed. Networking, website listing of the program, and as a resource for programs were answers that stood out the most. Research opportunities and job listings were neutral or less satis ed (Fig. 6). All respondents intended to renew membership in APPAP (100%). One respondent offered the following reasons for maintaining APPAP membership: "networking with other programs, annual meetings, listing the program on the website." Another comment was: "I appreciate being a part of APPAP and the history behind its beginnings." The top areas of satisfaction for APPAP membership include program listing on the APPAP website (79%), accreditation updates (71%), networking with other programs (61%), and APPAP annual conference (57%).

ARC-PA accreditation
Accreditation of PA postgraduate programs has been evolving in various forms since 2010 but the process was put in "abeyance" in 2014. One reason for suspending was that not all programs participated in accreditation, and none were bound by the accreditation process. The following survey question was about an accreditation process. "If ARC-PA accreditation becomes available for your fellowship program will you pursue it?" Two-thirds (67%) said they would decline. Another two-thirds (68%) said they had discussed ARC-PA accreditation funding with their institution. The main reason for declining accreditation was cost (48%), as well as lack of support (18%), and other (joint PA/NP program or cost of multi-track programs). A plurality (37%) would be interested in applying for joint accreditation (not yet available), if it included NPs.

Including Nurse Practitioners in APPAP (selected comment):
There is an undeniable need for a JOINT accreditation process … as more programs are accepting NP and PA applicants.

Value of Postgraduate Programs at Sponsoring Institutions
To probe the reasons why programs have developed and remain operational, a series of questions were asked. Most agreed that recruitment and retention of postgraduate trainees is part of meeting workforce demands (83%), and retention of career staff (55%). Most program directors (76%) believed their program improved decision making and autonomy of PAs/NPs in the clinical setting. Additionally, (93%) of respondents believed their postgraduate program fostered interprofessional collaboration. Lastly, (34%) of respondents felt their postgraduate training program improved compliance with physician resident regulated work hours.

Discussion
The development of postgraduate fellowships for PAs began in 1973 in New York City and residencies for NPs at the Community Health Center, Inc., of Connecticut in 2007 (8). Because program records have not been centralized, it is di cult to determine the total number of new programs versus those that have closed over the years. As a result, a summary of the number of graduates from PA fellowship programs remains largely unknown. Our conservative estimate, based on an average number of programs with graduates per year spread over 45 years, is approximately 1200.
Why did PA postgraduate fellowships develop and why do they remain operational? From the program director's viewpoint, having a fellowship program in clinical medicine and surgery provides the new PA an opportunity to focus on a specialty and receive intensive training in a structured and abbreviated period. Value for the employing hospital or medical practice is from providers who can also back-ll the reduction in physician residency and fellowship work hours. The need for more medical personnel is growing as hospital bed expansion is underway without concurrent physician trainees to staff house o cer roles and PAs and NPs provide readily available sources of trained personnel for these roles (9,10).
Administratively, adding a PA into an existing fellowship (AKA 'residency') is not likely to impact the 'span of management' for an attending physician. Adding another trainee appears to require little more in the human resource o ce or in department commitment (10). From the student's viewpoint, additional education with supportive mentors provides more con dence to be gainfully employed with a marketable skill set (11). The economist views additional education as an opportunity cost (12). In essence for additional (optional) clinical education the PA foregoes full employment (at 50% higher wages) for one year. The exception being the military where the trainee is also a full-time government employee. From the educator's point of view the salary of a postgraduate trained PA does not differ from on-the-job trained PAs and therefore, some may be reluctant to recommend a PA fellowship (13) (14).
As for the future, based on this census and survey project, half of the respondent programs have been operational for more than 10 years. Older programs include Monte ore Hospital (Surgery, Critical Care), Norwalk Hospital/Yale Medical University (Burn/Surgery), The Army/Air Force-Baylor doctoral program (Orthopedics, Emergency Medicine) and Arrowhead Orthopedic Surgery PA Fellowship program. Most of the programs (74%) train PA fellows in both inpatient and outpatient settings. A growing number of institutions, such as The Mayo Clinic, have multiple postgraduate programs for PAs and NPs located on three campuses suggesting their demand is growing in large medical centers (15)(4). The VHA, within the Department of Veterans Affairs, has eight postgraduate programs for PAs and NPs ranging from primary care, geriatrics, psychiatry, to orthopedics (16). Given the VHA has 160 medical centers and most are a liated with an academic medical center, the growth of PAs and NPs is likely to continue (17).

Limitations
All surveys have limitations and this one was no exception. The APPAP survey was lengthy (average 17 minutes), which may have contributed to a low rate of return (47%). In addition, many clinically focused educational programs were suspended, or activity restricted, during the COVID-19 crisis. For example, there were 29 emergency medicine postgraduate programs for PAs in 2019 (18). This number was reduced in half in deferment of the 2020 Coronavirus pandemic. Suspending clinical postgraduate education may have affected the ability or interest of program personnel to complete the survey. The survey was also restricted to APPAP members.
One shortcoming is the limited information about NP residencies in this census. Centralized information of postgraduate NP information remains in development and not available at the time of this project. The next iteration of a postgraduate census survey should include NPs as their role appears to be indistinguishable from PAs.
Another limitation is that it was a single-mode survey delivered by email invitation. No incentives were offered even though they can improve response rates in certain situations. However, clari cation of some responses was sought which improved the reliability of the answers. The strength of this survey is its uniqueness in providing a contemporary overview of the characteristics of American clinical postgraduate education.

Conclusion
The movement to enhance the newly graduate PA with additional clinical education began in 1973. A half century later more than 100 PA postgraduate programs are operational across the US and producing medical specialists in 32 medical and surgical disciplines. Those most represented are surgery, emergency medicine, critical care, and orthopaedics. Most postgraduate programs are one year in length and utilize the PA as house o cer alongside categorical residents. The 34 participating PA programs in this APPAP member survey have several elements in common as they align with General Medical Education approved postgraduate programs and the PA is added at the margin of infrastructure. The prediction is that PA and NP postgraduate programs will grow as the shortfall in medical and surgical specialties worsens. As a source of readily available and reliable medical and surgical workforce their addition appears to be a necessity. Areas of needed research include return on investment for the sponsoring institution, external funding (i.e., grants) opportunities for postgraduate PA programs, assessing the career arc of PA/NP graduates, comparing postgraduate to on-the-job trainees, and assessing the opportunity cost of PA and NP postgraduate programs.

Declarations
Ethics approval and consent to participate Not applicable as this study does not involve human subjects research

Consent for publication
No consent required