Burnout Among Physician Assistants: A Systematic Review

Physician assistants provide patient care in private practices, hospitals, and government agencies to meet the growing healthcare needs of an aging population.(1) Physician assistants can perform many of the same duties as doctors and nurse practitioners, such as examining and diagnosing patients, taking and reviewing medical histories, and prescribing medication, all while collaborating with a team of physicians and other healthcare workers.(1-4) As more and more physicians become specialized and leave primary care, physician assistants are expected to play a more significant role in providing primary care.(1) In fact, it is anticipated that the demand for physician assistants will increase by 30% from 2014 to 2024.(1,2) Reports show that patients are highly satisfied with their interaction with physician assistants.(5,6) In a survey done by the American Academy of Physician Assistants, 93% of patients agreed that physician assistants add value to the healthcare team.(7) Among patients who participated in the survey, 93% think physician assistants will be a solution to the shortage of healthcare providers.(7)

Along with the great responsibilities for patient care come stress and the risk of burnout, a psychological syndrome commonly associated with professions that have “staff–client interaction,” including health professions. The nature of health professions—that is, learning about other people’s problems (psychological, social, or physical) and striving to address them, can lead to burnout if healthcare professionals are under constant stress and frustration. Consequences of burnout can include poor patient care, dysfunction in the workplace, and drug or alcohol abuse.(3,8,9) With physician assistants playing an increasingly more significant role in the medical community, high levels of stress and burnout could lead to widespread negative outcomes for employers, patients, and physician assistants themselves.(10)

Although empirical studies have examined burnout among physician assistants, the overall burnout level among physician assistants as a group remains unclear. A recent systematic review evaluated job satisfaction, burnout, and turnover among both nurse practitioners and physician assistants.(11) However that review identified only two articles related to burnout among physician assistants and was unable to provide comprehensive information on this topic. Our review focused solely on burnout among physician assistants. In addition, we assessed major factors of burnout and discussed solutions and courses of action to alleviate burnout among physician assistants.

Methods

Search Strategy

An electronic search of multiple literature databases was conducted using the search terms presented in Table 1. These databases included ISI Web of Science, PubMed, PsychINFO, Academic Search Premier, and CINAHL (with the last three searched via EBSCOhost). The search results were restricted to studies published in English from January 1, 1980, through June 20, 2018. Table 2 reports the search results acquired from PubMed to exemplify the search strategy. To be included in this review, studies had to examine burnout among physician assistants using primary data. For studies that examined burnout among an assortment of providers, such as physicians, physician assistants, and nurse practitioners, data regarding physician assistants had to be presented separately. Studies that were not conducted in the United States were excluded, because in other countries, job responsibilities, titles, and training requirements for physician assistants could be different.

Two authors (SG and AR) independently searched the five databases and screened the titles and abstracts in the search results to identify relevant articles. The full texts of selected relevant articles were reviewed based on inclusion and exclusion criteria. A third author (XL) decided final eligibility if there was disagreement between the first two authors regarding the eligibility of an article. Finally, the reference lists of the selected articles were reviewed to identify additional relevant articles on physician assistant burnout. The search results are presented in Figure 1.

Figure 1. PRISMA study development chart.

Quality Assessment

The quality assessment tool utilized is an adaptation of the instrument that was developed by Cummings and Estabrooks and used in other published systematic reviews (Figure 2).(12,13) This assessment contains four themes: design; sample; measurement; and statistical analysis. The assessment tool contained 12 items, with a total maximum score of 12 possible points. Items were scored as 0 if the item was not met or as 1 if the item was met. After scoring, studies were rated as strong (10–12 points), moderate (5–9 points), and weak (0–4 points) based on their point totals. Two authors (SG and AR) independently scored the six studies. Afterward, the authors compared their score results and discussed any differences with a third author (XL).

Figure 2. Quality assessment tool.

Results

Search Results

In our initial search, ISI Web of Science produced 41 articles regarding physician assistant burnout, PubMed produced 23 articles, and PsychINFO, Academic Search Premier, and CINAHL produced a total of 33. After a screening process and the removal of duplicates, the full text of 22 articles was reviewed.

The six studies included in this review met our inclusion and exclusion criteria, presented earlier. One study was published in 1993, another in 2002, two in 2004, one in 2016, and the last in 2018. All of the studies were conducted in the United States. One study focused specifically on emergency medicine physician assistants (EMPAs).(14) Two studies focused on burnout in military physician assistants.(15,16) Another study selected participants from the American Association of Physician Assistants (AAPA) mailing list.(17) One study focused on physician assistants in rural areas.(18) The last study assessed burnout among physician assistants in oncology.(19) Two of the six studies collected results from other healthcare providers, such as physicians, nurses, nurse practitioners, and medics.(15,16) All studies were cross-sectional and survey based.

Quality Assessment

A modified version of the Cummings and Estabrooks assessment tool (Figure 2) was used to assess the quality of the six articles.(12,13) All six of the studies scored as moderate quality, with the lowest score being 7 and the highest being 9 (Table 3). The internal consistency and validity of the instruments evaluating burnout required an alpha coefficient of at least 0.70, which was met by all of the studies. Only one study used a theoretical framework to guide their study, and none of the studies was prospective.(15) All but one of the studies analyzed correlation between multiple effects. Five of the six studies used probability sampling, but none stated justification for their sample size nor did any have a response rate equal to or greater than 60%. Varner et al.(15) justified their response rate, claiming it was consistent with other studies done among family medicine and military providers. Walters et al.(16) attributed the low response rate to a response bias—that is, those who completed the survey were more likely to have burnt out.

Instruments and Scoring Results

Various burnout instruments and scoring methods were used by the six included studies, making it difficult to compare results across studies. Table 4 presents the survey tools used by each study.

Four of the six studies used the third edition of the Maslach Burnout Inventory (MBI).(14,16,18,19) The MBI contains 22 items and measures burnout using three subscales: emotional exhaustion (EE); depersonalization (DP); and reduced personal accomplishment (PA).(10) EE refers to emotional depletion of the professional, DP refers to negative and cynical attitudes towards other people (clients/patients), and reduced PA refers to the professional seeing themselves negatively in their accomplishments at the work place or in their work with patients/clients.(10,14) The MBI manual specifies specific cut-off points to indicate levels of overall burnout as high, moderate, and low, with the cut-off points specified as high EE (>27), high DP (>10), and low PA (7), and PA (<38).(11) Benson et al.,(18) in their study of rural physician assistants, divided cut-off scores into thirds, with the upper third as high, middle third as moderate, and lower third as low.(14) All four of the articles using the MBI reported moderate to high EE, as well as moderate to high DP, and moderate to low scores in the PA subscales.(14,16,18,19)

The Maslach Burnout Inventory–General Survey (MBI-GS) is a 16-item test that reports burnout using three subscales: EE; cynicism; and personal efficacy. Burnout is defined using these cut-off scores: high (>25); moderate (15–24); and low (<14). Varner et al.(15) used a revised version of the MBI-GS to assess burnout. The revised MBI-GS in the study used depersonalization rather than cynicism and personal satisfaction rather than personal efficacy. Moderate EE was reported, along with low DP, and high personal satisfaction.

A 21-item Burnout Scale devised by Pines and Aronson(20) uses a scale of 1 through 7 (where 1 = never, and 7 = always) to score items. Scores between 2 and 3 are considered “doing well.” Scores between 3 and 4 suggest that priorities are to be evaluated and possible changes to be made. Scores higher than 4 reflect burnout.(17) Holmes et al.(17) used this scale to assess burnout, which was assessed by physical exhaustion, mental exhaustion, and EE in the aspects of stress. Due to the various burnout instruments and scoring methods, results for each individual study were evaluated separately.

Overall Burnout

Four of the six included studies reported that physician assistants do experience burnout.(14,16,18,19) Overall burnout results are presented in Table 5. The study of rural physician assistants reported 64% of the participants experienced high to moderate EE, 64% experienced high to moderate DP, and 46% experienced moderate to low PA.(18) Bell et al.’s(14) study on EMPAs reported 59% experienced moderate or high EE, 66% experienced moderate or high DP, and 34% experienced moderate or low PA.(14) The active duty Army physician assistants study found high EE (mean: 35.67, cutoff point: 27), high DP (mean: 20.72, cutoff point: 10) and low PA (mean: 33.50, cutoff score: 34).(16) Tetzlaff et al.(19) reported that 34.8% of oncology physician assistants indicated professional burnout based on their MBI scores, with 30.4% reporting high EE, 17.6% reporting high DP, and 19.6% reporting low PA.

Two of the six studies concluded that physician assistants are not experiencing burnout. Physician assistants serving in the Air Force reported moderate EE (mean: 15.10, cutoff point: 15), low DP (mean: 10.90, cutoff point: 14), and high personal satisfaction (mean: 27.62, cutoff point: 25). A negative trend in the MBI subscales showed burnout was not significant among Air Force physician assistants.(15) Another study that used occupational stress to measure burnout reported that physician assistants are considered “doing well” (mean: 2.8, cutoff point: 3).(17) The authors concluded that physician assistants were not burnt out in their study.(17)

Factors of Burnout

The included studies identified multiple factors that affect physician assistant burnout (Table 5). Benson et al.(18) found the following factors affected all three subscales of the MBI:

• “I (the physician assistant) feel professionally isolated” (DP: rho = 0.293, p <.01; EE: rho = 0.273, p <.01; PA: rho = –0.410, p <.01);

• Control over workload (DP: rho = –0.281, p <.01; EE: rho = –0.405, p <.01; PA: rho = 0.232, p <.01); and

• The adequacy of administrative support (DP: rho = –0.186, p <.05; EE: rho = –0.314, p <.01; PA: rho = 0.283, p <.01).

Benson et al. reported that the physician assistant’s work location being isolated significantly affected two of the subscales (DP: rho = 0.169, p <.05; PA: rho = –0.175, p <.05), whereas hours worked per week was significant to only the DP subscale (rho = 0.169, p <.05). Satisfaction with and access to the physician supervisor affected only the EE subscale significantly (with supervisor: rho = –0.262, p <.01; access to supervisor: rho = –0.245, p <.01).

The study that focused on physician assistants in oncology reported that the average number of hours worked per week and percentage of time spent on work-related activities were significant factors associated with burnout.(19) Additionally, when compared with physician assistants who were not burned out, physician assistants who reported burnout spent more time on indirect patient care versus direct patient care (p <.001). Physician assistants in oncology also had higher burnout rates if they felt they were not compensated fairly (p <.001). Physician assistants were more likely to report burnout if they did not feel valued by their collaborating physician, did not feel encouraged to achieve professional goals, felt that their accomplishments were not recognized, or did not favor their collaborating physician’s leadership qualities (p <.001). When comparing subspecialties of physician assistants in oncology, those in medical oncology were more likely to report burnout than those in other subspecialties (e.g., radiation, pediatrics)(p = .010).

Using data from the EMPA Characteristics Survey, Bell et al.(14) reported that the following factors were associated with statistically higher probability of reporting burnout at least part of the time:

• Female gender;

• Planning to leave emergency medicine within one year;

• Low satisfaction with physician supervisors;

• Smoking one to two packs of cigarettes a day;

• Drinking more than six alcoholic beverages a week; and

• More frequent insomnia.

Bell et al. also identified the following factors that were statistically significant with all three MBI subscales:

• Self-assessed burnout (EE: rho = 0.620, p = .000; DP: rho = 0.407, p = .000; PA: rho –0.343, p = .000);

• Increasing satisfaction with physician supervisor (EE: rho = –0.346, p = .000; DP: rho = –0.263, p = .001; PA: rho = 0.346, p = .000); and

• Increasing frequency of insomnia (EE: rho = 0.267, p = .001; DP: rho = 0.227, p = .005; PA: rho = –0.193, p = .016).

The following three factors were found to be statistically significant only to the EE and DP subscales:

• Plans to leave emergency medicine (EE: rho = 0.363, p = .000; DP: rho = .230, p = .004);

• increasing use of alcohol (EE: rho = 0.162, p = .044; DP: rho = 0.163, p = .046); and

• Increasing use of recreational drugs (EE: rho = 0.177, p = .028; DP: rho = 0.182, p = .024).

Varner et al.(15) reported that age was a protective factor against burnout, with health professionals in their study over the age of 44 years reporting lower EE and higher personal satisfaction. Walters et al.(16) reported that having a large number of leave days accrued (62–93 days) was linked to high EE and DP (mean EE: 35.11, p <.001; mean DP: 34.18, p <.001)

A common factor that affected burnout among physician assistants was satisfaction with their physician supervisor or collaborating physician. Physician assistants who were satisfied with their physician supervisor or collaborating physician reported lower scores on the EE subscale in two studies, lower DP and higher PA scores in one study, and lower burnout overall compared with others who were not satisfied with their physician supervisor or collaborating physician in another study.(14,18,19) Another common factor was hours worked per week, with the hours significantly affecting the DP subscale in one study and overall burnout in another.(18,19)

In addition to assessing burnout via the MBI, MBI-GS, or Pines survey, the studies also allowed participants the opportunity to provide comments on what they believe attribute to burnout.

Walters et al.(16) mentioned that many physician assistants attributed the burnout to lack of sufficient time to complete their tasks beyond seeing patients, attesting to work overload. Other participants in the Walters et al. study mentioned that hospitals require many patient contact hours but do not consider other duties of these military providers, such as field exercises, meetings with their units, and so on. Varner et al.(15) study also gathered comments from their participants to suggest what can be done to enhance their well-being and efficiency: smaller patient loads, more staff to help with patient load, having more time with patients, and better quality of care. Both studies suggest physician assistants have heavy patient loads, interfering with their ability to provide quality care to each patient or to manage their other responsibilities.(15,16)

Discussion

Summary of Findings

To determine whether physician assistants experience burnout, we conducted a review of six studies to evaluate burnout levels among physician assistants. The results were mixed: four of the six included studies report physician assistants in their study experience burnout, whereas the two remaining studies reported physician assistants do not experience burnout. Burnout appears to be a moderate problem among physician assistants in certain specialties observed by the included studies.

Burnout Among Physician Assistants Versus Other Health Professionals

When compared with other health professionals in the included studies, physician assistants show as much, if not more, burnout and/or stress. Some of the studies included in this review evaluated burnout not only among physician assistants, but also other health professionals. According to Holmes et al.,(17) physician assistants reported less overall stress than nurses when comparing the results of the Health Professions Stress Inventory, and physicians reported less stress than physician assistants in subjects that dealt with career advancement, income, or role recognition and respect. Varner et al.(15) found that differences in the mean scores between nurse practitioners, physicians, and physician assistants were small, and none of the effects were statistically significant. Nonetheless, physician assistants scored slightly higher in EE and personal satisfaction subscales of the MBI-GS. Walters et al.(16) found that, when compared with combat medics and physicians, physician assistants scored highest in the EE and DP subscales and lowest in the PA subscale. Bell et al.(14) compared the results of emergency medicine physician assistants’ scores on the MBI with those of emergency physicians in other studies and found that EMPAs reported higher on the PA subscale than did emergency physicians. They attributed this to the fact that physician assistants start in lower-level positions compared with emergency physicians, and advancing in the healthcare field gives them a greater sense of PA.

Implications of the identified factors of burnout. Many different factors are related to burnout in physician assistants, with some of those factors related more strongly than others. The higher the score on the EE and DP subscales, the stronger the relation to burnout. Walters et al.(16) established a relationship between leave days accrued and burnout among army health professionals. Leave days are given to military members to take necessary time off to recover from missions and relax. There are two possible explanations for the large number of leave days accrued among army physician assistants: (1) the rigid command structure in the army may deter physician assistants from taking days off; or (2) it may be that physician assistants, like other healthcare professionals, put more effort and energy into taking care of their patients rather than taking care of themselves.(17,21)

Another factor of burnout was the number of hours worked per week. Both Benson et al.(18) and Tetzlaff et al. found this to be a significant factor. Tetzlaff et al.(19) found that almost half of the physician assistants felt burned out if they worked 51 to 60 hours a week. Appropriate measures should be in place to encourage physician assistants to take adequate time off periodically and to manage the hours worked per week to avoid burnout.

Bell et al.(14) and Benson et al.(18) found that physician assistants who were satisfied with their physician supervisor had scores indicative of less burnout (Benson et al., lower EE score; Bell et al., lower EE and DP, higher PA). Physician assistants who were unsatisfied with their physician supervisor experienced higher burnout, stress, or both. One source of dissatisfaction was having to work with multiple physician supervisors and having to adapt continually to their different characteristics. In order to limit this factor, there should be an effort for physician assistants to work with a limited number of different physicians.

Actions should be taken to reexamine and improve the collaborative practice model by enhancing teamwork, leadership development, and communication skills.

A lack of administrative support also causes burnout among physician assistants.(14,18) As a consequence, some physician assistants have to take on administrative responsibilities in addition to their clinical responsibilities. Supervising physicians can alleviate these factors by providing extra or adequate training to physician assistants and administrative staff. Supervisors also can allot more time for assigned tasks to be completed.(14,18) Additionally, Tetzlaff et al.(19) found that physician assistants’ opinion on their collaborating physician’s leadership qualities had a significant impact on burnout. Actions should be taken to reexamine and improve the collaborative practice model by enhancing teamwork, leadership development, and communication skills. Our review highlights the importance of the incorporation of physician assistant–physician professional relationships as a part of the curriculum in physician assistant educational programs.

Managerial recommendations: As mentioned earlier, burnout can lead to many negative outcomes if it is not addressed. Consequently, it would behoove healthcare managers to adopt strategies to prevent or alleviate burnout among physician assistants. First, healthcare managers need to monitor work hours so that physician assistants and other clinicians do not consistently exceed the 40-hour work week. Studies show that long working hours are associated with a decreasing rate of productivity and a negative impact on workers’ health.(22-24) In addition, it is critical that these clinicians use their leave or vacation time to avoid burnout and improve productivity. Second, this review indicates that physician assistants were more likely to report burnout if working with multiple physician supervisors and adjusting to their different leadership styles. Efforts should be made to reduce the disruption to teams so that the team structure can remain stable and efficient. Furthermore, leadership training should be part of physicians’ continuing education curriculum, because they often are tasked with overseeing physician assistants and other medical personnel. Finally, in order to reduce the administrative burden physician assistants often are required to shoulder in addition to providing medical care, strong consideration should be given to hiring administrative staff to simplify physician assistants’ scope of work. This not only will alleviate burnout, but also can increase productivity or allow physician assistants to spend more time with patients, which may increase patient satisfaction.

Limitations of the Current Review

This review was limited to the findings of six studies that had different methods of evaluating burnout for physician assistants and had limited sample sizes and settings. These factors made it difficult to compare and assess factors and levels of burnout across different studies and to make the findings generalizable to the larger physician assistant population. Additionally, this review intentionally did not include studies on physician assistant burnout conducted in countries other than the United States, because physician assistants in other countries may have different training, job responsibilities, and titles, among other things.

Future Research

With physician assistants becoming a growing and more integral part of the healthcare system, more studies of burnout conducted among physician assistants are needed. A single validated instrument should be used in future studies, with a larger sample size and in more practice settings, which would allow for better comparison of factors and burnout and to compare these findings across the physician assistant profession. There was a lack of consistency of significant factors associated with burnout across the included studies. However, a few factors were found to be significant to burnout by two or more of the included studies, such as hours worked per week and satisfaction with the physician supervisor or collaborating physician. It may be worth looking at those factors more closely to uncover how they manifest in various physician assistant groups. Furthermore, health-related behaviors such as drugs, sleep habits, and so forth were not measured at all or as in depth in many studies as they were by Bell et al.(14) These factors could be playing a more involved role in burnout than can be summarized in this systematic review based on the included studies. More studies are needed to further evaluate reported factors and explore any additional health-related factors of burnout.

References

1. Bureau of Labor Statistics, U.S. Department of Labor, Occupational Outlook Handbook, Physician Assistants. www.bls.gov/ooh/healthcare/physician-assistants.htm . Accessed December 12, 2020.

2. AAPA. What is a PA? 2017. www.aapa.org/what-is-a-pa/ . Accessed January 2017.

3. Campbell DA Jr, Sonnad SS, Eckhauser FE, Campbell KK, Greenfield LJ. Burnout among American surgeons. Surgery. 2001;130:696-702.

4. Nurse practitioner vs. physician assistant. Nurse Journal. 2016. https://nursejournal.org/nurse-practitioner/np-vs-physician-assistants/ . Accessed January 2017.

5. Brady MI. A survey assessing patient satisfaction with physician assistant care at the Maple Street Clinic. School of Physician Assistant Studies. 2004: Paper 127.

6. Counselman FL, Graffeo CA, Hill JT. Patient satisfaction with physician assistants (PAs) in an ED fast track. Am J Emerg Med. 2000;18:661-665.

7. AAPA. Attitudes toward physician assistants: a 2014 survey by the American Academy of Physician Assistants. www.aapa.org/wp-content/uploads/2017/01/AAPA-HarrisSurvey-Methodology-and-Tables.pdf . Accessed January 2017.

8. Shanafelt TD, Sloan JA, Habermann TM. The well-being of physicians. Am J Med. 2003;114:513-519.

9. Barden CB, Specht MC, McCarter MD, Daly JM, Fahey TJ 3rd. Effects of limited work hours on surgical training. J Am Coll Surg. 2002;195:531-538.

10. Maslach CJ, Jackson SE, Leiter MP. Maslach Burnout Inventory Manual, 3rd ed. Palo Alto, CA: Consulting Psychologists Press; 1996.

11. Hoff T, Carabetta S, Collinson GE. Satisfaction, burnout, and turnover among nurse practitioners and physician assistants: a review of the empirical literature. Medical Care Research and Review. 2017. doi: 10.1177/1077558717730157.

12. Keyko K, Cummings GG, Yonge O, Wong CA. Work engagement in professional nursing practice: a systematic review. Int J Nurs Studies. 2016;61:142-164.

13. Lin PS, Viscardi MK, McHugh MD. Factors influencing job satisfaction of new graduate nurses participating in nurse residency programs: a systematic review. Journal of Continuing Education in Nursing. 2014;45:439-450.

14. Bell RB, Davison M, Sefcik D. A first survey. Measuring burnout in emergency medicine physician assistants. JAAPA. 2002;15(3):40-42, 45-48, 51-52.

15. Varner DF, Foutch BK. Depression and burnout symptoms among Air Force family medicine providers. JAAPA. 2014;27(5):42-46.

16. Walters TA, Matthews EP, Dailey JI. Burnout in Army health care providers. Mil Med. 2014;179:1006-1012.

17. Holmes SE, Fasser CE. Occupational stress among physician assistants. JAAPA. 1993;6:172-178.

18. Benson MA, Peterson T, Salazar L, et al. Burnout in rural physician assistants: an initial study. J Physician Assist Educ. 2016;27(2):81-83.

19. Tetzlaff ED, Hylton HM, DeMora L, Ruth K, Wong YN. National study of burnout and career satisfaction among physician assistants in oncology: implications for team-based care. J Oncol Pract. 2018;14(1):e11-e22

20. Pines AM, Aronson E, Kafry D. Burnout: From Tedium to Personal Growth. New York: Free Press; 1981.

21. Bond C. The training of the “helpless” physician. MedGenMed. 2007;9(3):47.

22. Pencavel J. The productivity of working hours. The Economic Journal. 2014;125:2052-2076.

23. Kivimäki M, Jokela M, Nyberg ST, et al. Long working hours and risk of coronary heart disease and stroke: a systematic review and meta-analysis of published and unpublished data for 603 838 individuals. Lancet. 2015;386(10005):1739-1746.

24. Virtanen M, Heikkilä K, Jokela M, et al. Long working hours and coronary heart disease: a systematic review and meta-analysis. Am J Epidemiol. 2012;176:586-596.