Academic healthcare settings are complex environments in which a delicate balance exists between clinical training and patient care.1,2 The past 2 decades have seen a surge of physician assistants (PAs) in US teaching hospitals, primarily driven by the 2003 Accreditation Council for Graduate Medical Education (ACGME) resident work-hour restrictions; PAs are intended to improve patient safety and resident education.3,4 Studies have shown that the presence of PAs in clinical learning environments has helped to alleviate the workload for residents, freeing up more time for resident learning.5-7 Furthermore, two recent systematic reviews have confirmed that PAs expand the clinical workforce by providing quality of patient care comparable to that provided by physicians.8,9 Therefore, the use of PAs in academic healthcare settings will likely continue.3
The ACGME seeks to ensure that all residents are trained in settings that provide sufficient opportunities for clinical learning and that a balance is maintained between service and education.10 Residents report that PAs are generally beneficial to their training, primarily due to PAs' effect on reducing resident work hours and workload.5-7,11,12 Furthermore, residents participating in recent ACGME site visits reported that increasing the availability of PAs in the clinical learning environment was an opportunity to improve their program.5,6,11,13 However, a recent cross-sectional study of surgical residents indicated that a sizable minority (31%) perceived PAs as having a detrimental effect on their training, especially when nurses contacted PAs (rather than residents) with patient-care issues.6 Although these researchers did not seek to determine the ways in which PAs affect trainee learning, their findings suggest that varying factors in the clinical learning environment, such as the ways that PAs enact their work, may enhance or hinder trainees' learning.6
Given the contradictory findings in the literature on the influence of PAs on trainee learning, the research team sought to understand how PAs can enhance or hinder resident learning and maintain quality of patient care. Earlier studies have primarily explored trainees' perspectives. Given the unique responsibilities of physician faculty and PAs related to patient care and trainee learning, we sought to identify, from their perspectives, how the presence of PAs in the clinical learning environment may enhance or hinder trainee learning.
We used a qualitative study design and semistructured interviews with physician faculty members (faculty) and PAs to gain a deeper understanding about how PAs may affect trainee learning. A qualitative study design lets researchers explore participants' beliefs, experiences, perceptions, or feelings that otherwise would be difficult to investigate, to gain a more in-depth understanding of phenomena, and to develop ideas for potential quantitative research—especially for phenomena that have not been thoroughly researched.14 This research methodology typically includes a purposive sample of participants specifically selected for the rich information that they might share as to generate in-depth and rich data collection.15
Context and setting
Interviews were conducted between August and November 2016 with participants from the University of Texas MD Anderson Cancer Center in Houston, Texas, one of the largest employers of PAs in the United States, with more than 300 PAs at the time of the study. MD Anderson sponsors more than 70 fellowship and residency programs and offers short-term clinical rotations for trainees from other residency and fellowship programs in and near Houston. Residents and fellows typically are embedded into clinical services that have full-time attending physicians supplemented with PAs and/or NPs. The model used by some teaching hospitals of having separate teaching and nonteaching services is not widely adopted at MD Anderson. Therefore, the site was selected due to the extensive use of PAs in clinical settings where residents and fellows train and the commonly used model of interprofessional medical teams consisting of physicians and PAs and/or NPs. In this setting, PAs are members of the staff and not members of the clinical faculty. Although eligible participants included those on oncology and nononcology services, because the study was conducted at a cancer center, all participants were involved in the care of cancer patients.
Eligible participants were PAs with at least 2 years of experience and physician faculty (faculty) from clinical departments that offer residency or fellowship training to physicians and that also employ PAs. An email explaining the study and the informed-consent document was sent to potential participants. Purposive sampling was used to recruit 12 faculty and 12 PA participants with varying experiences by selecting male and female participants from surgical and nonsurgical departments, and with different years of experience.15
Semistructured interviews were used to provide the opportunity for participants to discuss their experiences unconstrained by the interviewer, while also providing the opportunity to probe for clarification and elaboration.14 The interview guide, developed by the research team, consisted of primarily open-ended questions about the perceived effect of PAs on trainee learning and participants' previous experiences working in settings where both PAs and trainees participate in patient care. Minor revisions to the interview guide were made during the course of the study so that new concepts that emerged from earlier interviews could be explored.16
The primary investigator (MP) conducted face-to-face interviews with study participants over about 1 hour. Interviews were audio-recorded by a research assistant and recordings were transcribed verbatim. Names and other identifying information were removed from the transcripts; the participants were designated as F01 through F12 (faculty) and P01 through P12 (PAs). Sampling continued until the research team reached consensus that saturation had occurred and new themes were unlikely to emerge.14
Demographic data are reported here in aggregate; identifying factors such as specific departments and programs are not reported to ensure participant anonymity. Conventional content data analysis, defined as “the subjective interpretation of the content of text data through the systematic classification process of coding and identifying themes or patterns,” was used to allow themes to emerge from the data.17 The analysis was conducted as an iterative process through discussions among members of the research team who identified common and interesting themes, including the selection of representative quotes. The initial phase of coding consisted of the primary investigator (MP) and one research assistant independently coding the interview transcripts, followed by discussions of the identified codes to reach consensus. Preliminary analysis of the data informed probing during subsequent interviews. Emerging themes were then discussed with the other investigators (MG, RS, AE, DD) and as themes were refined, MP independently reviewed coded transcripts using the identified themes.
The research team consisted of the primary investigator (MP), who previously practiced as a PA and who was serving as a program director for education programs in the research setting at the time. In this capacity, she had worked with many of the participants, although she had not supervised or been supervised by any. Additional members of the research team were three educational scientists (RS, MG, DD), one of whom was also trained as a physician (MG), as well as two clinical faculty members from the research setting (AE, DB), providing researcher triangulation.18
Throughout this research, reflexivity, the process of examining oneself as researcher and the research relationship, was used as a means of making explicit the assumptions and beliefs about the research agenda.19 Reflexivity consisted of the primary investigator (MP) writing field notes and memos exploring her viewpoints and potential biases as they emerged throughout the study. These perspectives and those of the other researchers were discussed during team meetings.
The MD Anderson institutional review board approved this study. Written informed consent was obtained from each participant. Only the primary investigator and research assistants had access to the participant list and were aware of participants' identities. All deidentified data were entered into a qualitative data software system (MAXQDA version 12.1.3; VERBI Software, Berlin, Germany), which was used for storing and managing data. No external funding was provided for this study.
Participant background and demographic information is provided in Table 1. Participants identified various ways in which PAs may affect trainee learning, intrinsically linked to the roles PAs assume in the clinical learning environment: clinician, teammate, and clinical teacher.
PAs as clinicians
When PAs serve as clinicians in the clinical learning environment, the available learning time for trainees can be optimized. PAs can address routine patient care issues, reducing interruptions during learning activities for trainees such as during lectures or when surgical trainees are in the OR. PAs also can see patients with less educational value for trainees, freeing up trainees' time to see fewer but more complex patients. For example, new patients and consultation appointments often were directed to trainees because these patient encounters typically provide rich opportunities to learn about complex evaluation and treatment decisions. As one faculty member (F7) noted, seeing fewer patients lets trainees think more deeply about the patients they see. A final identified enhancement to trainee learning was that PA–patient visits can be used by faculty as additional teaching cases for trainees. These visits were described as providing an opportunity to discuss additional patients without trainees needing to perform the clinical tasks required for the visit, such as medical record documentation.
On the other hand, participants noted that PAs in the clinical learning environment may hinder trainee learning if PAs perform procedures that trainees need to learn or when PAs see patients with educational value for trainees. Table 2 provides representative quotes from interviewed faculty and PAs.
PAs as teammates
Enhancements and hindrances to trainee learning also may occur when PAs serve as teammates to trainees (see Table 3 for representative quotes). Having PAs as teammates can provide opportunities for trainees to enhance their interprofessional competencies, such as learning about the PA profession and how to effectively collaborate with PAs. The relationship with PAs was perceived to be a unique interprofessional relationship for trainees; as one faculty member (F11) described, “it's one of the closest [relationships] the fellows have.” However, trainee learning may be hindered if trainees are intimidated by experienced PAs on the team who know the system, patient population, clinical staff, and medical discipline better than do the trainees. As one PA (P09) said, this intimidation may interfere with trainees developing confidence as physicians and may make it more difficult for trainees to establish their own practice style instead of tending to conform to the PA's way. Trainees also may be less involved in patient care activities when PAs are on their team, including if trainees become too dependent on PAs or are excluded from clinical decision-making for their patients. Participants noted that these issues often arose due to the faculty and other healthcare providers (most notably nurses) perceiving PAs as more experienced and feeling “more comfortable” (F02) with PAs, given their long-term relationship with PAs compared with trainees, who were with their service for only a short period of time. PAs themselves described a high level of responsibility being placed on them to ensure details of patient care were addressed, and their concern that trainees were not always dependable to follow up on all aspects of patient care. PA participants also acknowledged that in some situations, they avoided taking the time to include trainees in learning opportunities because of concerns about efficiency of patient care. As one PA (P06) said, “...I think it's easier just to do it. It's faster and then you're done.”
PAs as clinical teachers
Although participants indicated that clinical teaching was primarily the responsibility of the faculty, they also identified various ways PAs can affect trainee learning by serving as (unofficial) clinical teachers (see Table 4 for representative quotes). For this role, only enhancements to trainee learning were identified by participants. These related to PAs enculturating trainees to the learning environment and providing clinical instruction for trainees. PAs were perceived as being valuable to enculturating trainees to the clinical environment in two ways. First, PAs can help to orient trainees to clinic systems, such as use of the electronic medical record. One faculty member (F05) described the essential role of the inpatient PAs in orienting trainees in the hospital because physician faculty in the department only rotated on the inpatient service for a few weeks each year. Second, PAs can be instrumental in helping trainees navigate working with the faculty, for example, learning what to expect from particular faculty members and understanding their individual practice styles. PAs were perceived as being uniquely able to provide this guidance, given their close working relationship with the faculty; as one PA (P05) explained, “I'm going to know my surgeon better than anyone else does.”
PAs were also described as providing clinical teaching for trainees, which may take several forms. PAs can identify learning opportunities for trainees such as directing trainees to patients of particular educational value. PAs also can demonstrate clinical skills to trainees, particularly related to modeling professional behaviors, such as how to manage the workflow of a busy clinic, how to interact with other healthcare professionals, and how to effectively communicate with patients. PAs were perceived as contributing directly to trainee learning through clinical instruction by answering questions about medical management and giving feedback on patient care plans developed by trainees. Often, PAs had provided clinical instruction when faculty members were unavailable, such as when they were with another patient or when they were involved in nonclinical activities. Most participants reported that PAs can participate in the evaluation process for trainees. Some faculty members perceived PAs' evaluations as important to supplement faculty evaluations because of their impression that some trainees displayed “inconsistent” (F08) behavior if a faculty member was not present to observe them. Overall, most participants indicated that PAs were valuable in augmenting the clinical teaching provided by the faculty; as one faculty member (F05) said, “it takes a village.”
This qualitative study provides an in-depth exploration of the potential effect of PAs on trainee learning from the perspectives of physician faculty and PAs working at one academic clinical setting with extensive use of PAs. This study fills an important gap in the medical education literature as it provides a deeper understanding of the roles PAs may assume in the learning process of trainees. Our findings revealed potential enhancements and hindrances to trainee learning related to how PAs work in the clinical learning environment, which may explain the conflicting trainee perspectives indicated in previous studies.5-7,11,12 We also have demonstrated that the roles of healthcare professionals may be multidimensional, illustrating the complex nature of interprofessional healthcare teams. Understanding the roles enacted by healthcare professionals is essential as we seek to enhance the clinical learning environment and advance future educational research.
First, by simply fulfilling their primary role of clinician in the patient care workforce, PAs may enhance or hinder trainee learning, depending on PAs' clinical responsibilities and how patients are assigned to trainees and PAs. PAs may assume clinical duties for routine patient care tasks that free up trainees to participate in other patient care or nonpatient care activities that enhance their learning. Alternatively, PAs also may assume clinical duties that are essential to trainees' competence development. Therefore, the addition of PAs as clinicians in the clinical learning environment may affect the learning opportunities provided to trainees. Given the need to ensure quality learning opportunities for trainees, facilities must consider PAs' clinical responsibilities to ensure that clinical experiences for trainees support their education.20
The second role that we identified was that of the teammate, when PAs and trainees worked on the same clinical teams caring for the same patients. Our findings suggest that having PAs as teammates in the clinical learning environment provides unique opportunities for trainees to learn about PAs, whom they will likely work with once in practice, and hone their skills in interprofessional collaboration. Given the recent attention placed on the need for trainees to develop and apply competencies in interprofessional collaborative practice, PAs serving as teammates in the clinical learning environment may provide trainees with essential opportunities to enhance these competencies.21,22
In spite of the opportunities provided by having PAs as teammates, our findings also revealed important challenges to trainees learning on teams with PAs. Identified hindrances to trainee learning revealed in our study are supportive of the analysis by Nokes-Malach and colleagues regarding collaborative learning.23 They described two social aspects of collaboration that can hinder group learning. The first, social loafing (“where some group members do not engage optimally in the task because they believe someone else in the group will pick up the slack”), was identified by participants who reported that trainees may not demonstrate sufficient levels of responsibility due to the impression that the PA would complete the necessary tasks.23 The second aspect, fear of evaluation (when individuals may be afraid of negative evaluation from other group members), can be seen if trainees feel intimidated by more-experienced PAs on the team.23 As emphasized by Nokes-Malach, educators should consider how to mitigate these social factors of learning on teams.20,23
Last, we found that PAs can serve in the role of clinical teachers for trainees, even in settings in which PAs are not formally appointed members of the teaching faculty. PA involvement in teaching seemed to be most frequent when physician faculty were not immediately present, suggesting that PAs may function as an adjunct member of the teaching team. Given the various responsibilities of the physician faculty, including those beyond patient care and teaching, such as research and administration, PA involvement as part of a teaching team may help faculty balance these often competing priorities, as suggested by previous investigators.24,25
This study was conducted at a single institution that has used PAs extensively for many years, providing an opportunity to explore the challenges and opportunities afforded by PAs working with residents in the clinical learning environment. These results may not apply in other settings, such as those with less experience with PA clinicians or where teaching and nonteaching services are separate. Furthermore, study participants were from various medical and surgical disciplines. The study did not aim to explore differences that may exist between disciplines. Further research is needed to determine the degree to which these findings may be transferable to other settings or vary between disciplines.
A variety of questions arise in consideration of PAs as clinical teachers for physician trainees, such as what motivates PAs to teach, particularly when they are not part of the faculty. Several intrinsic motivators for physicians involved in clinical teaching have been identified, including the desire to repay former teachers and to train the next generation of physicians.26 However, it is not known if such intrinsic motivators apply to PAs, especially when teaching across professions. How PAs may balance their various professional roles to foster effective collaborative practice and learning also is unclear. Furthermore, barriers may exist to physicians (both faculty and trainees) accepting PAs in the role of clinical teacher.
This study has several limitations. First, given the sparse previous research about the experiences of physician faculty and PAs, the study sought to explore their perspectives and did not interview trainees; studies have shown that trainee perspectives may differ.5 For example, although no specific hindrances to trainee learning resulting from PAs serving as clinical teachers were discussed by participants, such hindrances may exist, and future research, including studies involving trainees, can explore this question. Second, given the interviewer's background as a PA employed in the study setting, participants' comments may have been influenced by this relationship, and the semistructured nature of the interviews may have led the interviewer to impose biases on the probing questions participants were asked. As is recommended in qualitative research methods, researcher triangulation and reflexivity were used to mitigate the potential influence of these relationships and improve credibility of the analysis.18,19 These relationships also may have provided unique opportunities to explore issues more deeply, given the familiarity of the interviewer with the clinical setting.27 Third and finally, given the overlapping responsibilities and similar educational model of PAs and physicians, we elected to focus on PAs only and did not address other healthcare professionals, most notably NPs.28 Given the differences in educational models and competencies of other healthcare professionals, the degree to which our findings may be transferrable to other professional populations may be explored in future studies.
Some practical implications of this work include the need for PAs to consider the effect they may have on trainee learning while practicing in the clinical learning environment. Furthermore, PAs may benefit from participating in professional development activities to develop competencies in clinical teaching. Program faculty should consider the effect of PAs and other healthcare professionals on educational opportunities for trainees, with the goal of balancing efficient patient care and trainee learning. Furthermore, although PAs should be integrated into healthcare teams with trainees, providing opportunities for interprofessional learning as previously described, individual accountability of each team member also should be promoted.29 Finally, residency and fellowship program leadership should explore opportunities for PAs to be formally included as members of the teaching team as a way to enhance learning opportunities for trainees and relieve the burden of teaching by the physician faculty. Ultimately, rather than exclusively focusing only on physicians as educators of other physicians, we concur with previous recommendations to incorporate other healthcare professionals into the clinical teaching team.24,25
The results of this study indicate that PAs may serve in various roles (clinician, teammate, and clinical teacher) in the clinical learning environment. Depending on how these roles are enacted, PAs may enhance or hinder trainee learning. Given the increasing frequency to which PAs are employed in the clinical learning environment, further studies are needed to understand how to optimize clinical learning and patient care in interprofessional learning settings.
1. Philibert I, Friedmann P, Williams WT. New requirements for resident duty hours. JAMA
2. Cleland J, Roberts R, Kitto S, et al. Using paradox theory to understand responses to tensions between service and training in general surgery. Med Educ
3. Pezzi C, Leibrandt T, Suryadevara S, et al. The present and future use of physician extenders in general surgery training programs: one response to the 80-hour work week. J Am Coll Surg
4. Jones PE, Cawley JF. Workweek restrictions and specialty-trained physician assistants
: potential opportunities. J Surg Educ
5. Buch KE, Genovese MY, Conigliaro JL, et al. Non-physician practitioners' overall enhancement to a surgical resident's experience. J Surg Educ
6. Kahn SA, Davis SA, Banes CT, et al. Impact of advanced practice providers (nurse practitioners and physician assistants
) on surgical residents
' critical care experience. J Surg Res
7. Dies N, Rashid S, Shandling M, et al. Physician assistants
reduce resident workload and improve care in an academic surgical setting. JAAPA
8. Foster CB, Simone S, Bagdure D, et al. Optimizing team dynamics: an assessment of physician trainees
and advanced practice providers collaborative practice. Pediatr Crit Care Med
9. Johal J, Dodd A. Physician extenders on surgical services: a systematic review. Can J Surg
10. Nasca TJ, Philibert I, Brigham T, Flynn TC. The next GME accreditation system—rationale and benefits. N Engl J Med
11. Resnick AS, Todd BA, Mullen JL, Morris JB. How do surgical residents
and non-physician practitioners play together in the sandbox. Curr Surg
12. Freiburg C, James T, Ashikaga T, et al. Strategies to accommodate resident work-hour restrictions: impact on surgical education. J Surg Educ
13. Caniano DA, Yamazaki K, Yaghmour N, et al. Resident and faculty perceptions of program strengths and opportunities for improvement: comparison of site visit reports and ACGME resident survey data in 5 surgical specialties. J Grad Med Educ
14. Creswell JW. Educational Research: Planning, Conducting and Evaluating Quantitative and Qualitative Research
. 4th ed. London, United Kingdom: Pearson; 2014.
15. Starks H, Trinidad SB. Choose your method: a comparison of phenomenology, discourse analysis, and grounded theory. Qual Health Res
16. Watling CJ, Lingard L. Grounded theory in medical education research: AMEE Guide No. 70. Med Teach
17. Hsieh HF, Shannon SE. Three approaches to qualitative content analysis. Qual Health Res
18. Reeves S, Peller J, Goldman J, Kitto S. Ethnography in qualitative educational research: AMEE Guide No. 80. Med Teach
19. Hsiung PC. Teaching
reflexivity in qualitative interviewing. Teach Sociol
20. Billett S. Constituting the workplace curriculum. J Curriculum Stud
21. Charles G, Bainbridge L, Gilbert J. The University of British Columbia model of interprofessional education. J Interprof Care
22. Interprofessional Education Collaborative Expert Panel. Core Competencies for Interprofessional Collaborative Practice: Report of an Expert Panel
. Washington, DC: Interprofessional Education Collaborative; 2011.
23. Nokes-Malach TJ, Richey JE, Gadgil S. When is it better to learn together? Insights from research on collaborative learning. Educ Psychol Rev
24. Tsai PI. Commentary on: impact of advanced practice providers (nurse practitioners and physician assistants
) on surgical residents
' critical care experience. J Surg Res
25. Stalmeijer RE. Teaching
in the clinical workplace: looking beyond the power of ‘the one’. Perspect Med Educ
26. Steinert Y, Macdonald ME. Why physicians teach: giving back by paying it forward. Med Educ
27. Gorden RL. Interviewing: Strategy, Techniques and Tactics
. 3rd ed. Homewood, IL: Dorsey Press; 1980.
28. Kurtzman ET, Barnow BS. A comparison of nurse practitioners, physician assistants
, and primary care physicians' patterns of practice and quality of care in health centers. Med Care
29. Johnson DWJ, Johnson RT. An educational psychology success story: social interdependence theory and cooperative learning. Educ Res