INTRODUCTION

In recent decades, the physician’s role has changed owing to the organizational transformation of health care, the growth of health systems, misaligned financial incentives, and the adoption of electronic health records (EHR).1,2,3 These changes have contributed to an escalating crisis of physician burnout.4 In 2014, 54% of physicians in a nationally representative sample reported symptoms of burnout, an increase from 46% in 2011, and double the rate found in the general United States population.4 Physicians in primary care are at increased risk of burnout. 4,5,6,7

Burnout affects personal relationships and increases the risk of alcoholism and suicide.8,9 It also poses problems for health systems and patients via increased professional turnover, decreased productivity, reduced quality and safety of patient care,10,11,12,13,14 and lowered patient satisfaction.15,16 In order to achieve the triple aim of improving patient experience and population health whilst reducing the cost of health care, some have argued the need to include physician well-being, making it the quadruple aim.17,18 Health systems are increasingly taking steps to enhance physician well-being.19,20 For instance, the Mayo Clinic has used a bottom-up approach to tackling burnout using focus groups with physicians.21 Other approaches target individual factors such as promoting resilience and mindfulness.22,23

Most qualitative studies on physician burnout investigate the individual factors influencing burnout.24,25,26 Other studies analyze survey data, which lack nuanced information about the experience of burnout and do not typically explore strategies to improve well-being. A recent survey of 934 physicians in our organization identified a number of risk factors for burnout,27 and this research is a qualitative companion study aimed at understanding perspectives of frontline physicians and health system leaders on contributors to burnout and potential solutions.

METHODS

We conducted in-depth key informant interviews with frontline physicians and health system leaders, e.g., leaders of individual departments and specialties or executives at Palo Alto Medical Foundation (PAMF), a predominately fee-for-service, multispecialty group practice in Northern California with over 1300 physicians and serving about 1 million patients. Interview participants provided informed consent and the protocol was approved by the organization’s Institutional Review Board.

Key informants were identified through organizational stakeholders and snowball sampling. Interviews were conducted by three researchers (ED, AM, and MTS) between June to October 2017, and the average duration was 1 hour. Interviews included a review of factors associated with burnout at the organization based on findings from the survey27 along with drivers of burnout described in the literature.21,27 Participants were asked to identify the most salient factors contributing to burnout using a card sorting exercise,28 and were then asked questions about these identified factors (see interview guide in online Appendix 1).

Recruitment continued in tandem with analysis until we reached thematic saturation.29 Interviews were transcribed and imported into Atlas.ti software for analysis. The coding team (ED, AM, VM) met weekly to develop a codebook and coding structure. Initial deductive codes were identified from prior research.21,27 The coders began with descriptive and open coding using both pre-identified and emergent codes, including proposed strategies for reducing burnout. The second cycle of coding categorized codes into themes using a pattern coding approach,27,30 and matched factors to solutions suggested to improve well-being.

RESULTS

Seventeen of 21 key informants emailed agreed to participate in an interview. Participant characteristics are reported in Table 1. Participants identified the following organizational factors as primary contributors to burnout and identified relevant solutions: (1) workload and work-life balance; (2) organizational structure and culture; (3) support staff, physician autonomy, and leadership; and (4) social support, feeling valued, and meaning in work (see online Appendix 2 for details of all themes and their frequency). Participants also described how physicians currently cope with burnout. While 9 participants agreed that personal practices are associated with burnout, e.g., lack of exercise and sleep, all 17 noted that organizational factors were more important contributors to burnout. One participant said, “It does not matter how resilient or positive you are, the work environment, especially in primary care will eventually be a problem.”

Table 1 Characteristics of Interview Participants
Full size table

Participants believed burnout was most acute in adult primary care departments, i.e., Internal Medicine and Family Medicine. They cited several reasons: (1) less revenue and fewer resources for staffing; (2) a higher burden of EHR in-basket messages; (3) responsibility for care coordination, e.g., “everything runs downhill” to primary care; and (4) patients with chronic problems: “They tend to have a lot of needs.” Table 2 provides an overall summary of identified contributors to burnout and potential solutions. We describe them below.

Table 2 Causes of Burnout and Potential Solutions as Discussed in Interviews
Full size table

Workload and Poor Work-Life Balance: Desktop Medicine

Many physicians (n = 13) reported a significant increase in workload in recent years largely attributable to the EHR and two “pain points”: in-basket messages and progress notes. In-basket messages were described as a “nightmare” and a “fire hose” that never turned off, particularly in primary care. In-basket messages came from patients, colleagues, laboratories, pharmacies, physicians themselves, and the EHR system.27 The pressure to complete progress notes and close charts also contributed to what one participant described as the “pajama time problem”31: “You close your charts. You do your in-basket… you’re in pajamas, and you’re finishing your work for the day.”

The solutions suggested to address desktop medicine and work-life balance included reducing and compensating desktop medicine work by “redefining the FTE,” enhancing care team support to triage or answer in-basket messages and scribe support to assist with other EHR work such as documentation, along with increasing physicians’ flexibility and control over their schedule.

Organizational Structure and Culture: the Push for Productivity

The reliance on face-to-face visits for revenue was described as problematic due to the growth of uncompensated desktop medicine work. Participants described an organizational culture increasingly focused on growth, productivity, and physician performance metrics, and noted increased discussion of “business” in physician meetings in recent years.

Physicians noted the limitations of metrics, e.g., a physician whose office visits are all 15 min may appear “efficient,” but may not be spending enough time with some patients. Physicians also disliked the variation reduction practice of comparing physicians, labeling someone “in the red,” and distributing department metrics with physician identities revealed.

Participants suggested that the business model should evolve by changing how revenue is generated and resources are allocated. They suggested that the culture should focus on the well-being of existing patients and physicians. In response to “more and more business-like” physician meetings, they suggested limiting time spent on “business” and instead emphasizing positive stories and discussion of clinical challenges.

Support Staff, Physician Autonomy, and Leadership

Availability and retention of support staff strongly influenced physician well-being. Primary care physicians noted that medical assistants were only allocated during scheduled patient visit time, so physicians lacked support while working on desktop medicine tasks.

Organizational centralization and standardization efforts reduced physician autonomy and contributed to feelings of disempowerment and burnout. One physician described how the transition to a centralized call center meant that individual physicians’ preferences for scheduling were no longer honored, “[we] lost a bunch of workflows that were working pretty well.”

Physicians occupying leadership roles noted having minimal or no administrative support, not enough compensated time for leadership work, and insufficient training and mentorship. One former leader described the transition to leadership: “There wasn’t a lot of support… ‘Now you’re a leader. Figure it out…’ And that was very alienating for me.” Physicians remarked that burnout among leadership had a trickle-down effect and led to loss of institutional knowledge and momentum when leadership changed.

Strategies suggested to improve staffing included enhancing care teams so physicians work at the top of their skill level, allocating more support staff to primary care, hiring advanced practice clinicians (APCs), and improving staff retention. Physicians wanted staff to be allocated to them during all the hours they were at the clinic instead of only when face-to-face visits are scheduled. Strategies suggested to enhance physician autonomy included increasing physician control over schedules, and allowing telemedicine or virtual visits. Increasing leadership support, e.g., by increasing administrative support, would save physician leaders time and energy for work which required their skills. They also desired additional training, mentorship, and stipends for leadership work.

Social Support, Feeling Valued, and Meaning in Work

Physicians described their feelings about work being influenced by three factors (1) social support and camaraderie with colleagues, (2) feeling valued and heard, and (3) finding meaning in work.

Participants noted that a sense of community and social support at work facilitated well-being and was decreasing: “People on the EMR [Electronic Medical Records]… they don’t go to the cafeteria. They eat at their desk looking at the computer.” Nevertheless, participants told us that their colleagues were a strong positive force: “You want colleagues that are fun to be with, smart, and we have that here. So that's a lot of what keeps people here.”

Physicians described an absence of recognition of the positive aspects of their work that led them to feel de-valued, and that being valued required more than a pat on the back.

“If you said, ‘We value you, we value your opinion, we know that you're stressed because you get X amount of messages every day, we’re doing this to try to help you.’ That to me means that you're highly valued.”

Participants remarked that helping patients makes work meaningful, and that burnout was increased by tasks that reduced their face-to-face time with patients, such as desktop medicine time.

Strategies suggested to ensure physicians felt valued and appreciated included reducing the emphasis on productivity, and consciously embracing and celebrating clinical and organizational accomplishments. Participants suggested fostering quality social interactions between colleagues to improve well-being. They pointed to increased organizational efforts to have group lunches, and ad hoc efforts to increase department social events. Physicians also suggested a need to systematically identify and respond to concerns through regular well-being surveys, mentoring programs, and ensuring that physicians’ voices are heard, questions are answered, problems are addressed, and meaningful work is prioritized.

Coping with Burnout: Fractional Quitting and Quitting

Physicians described several coping strategies used to reduce burnout, specifically (1) reducing clinical work hours, dubbed “fractional quitting,” (2) leaving leadership roles, (3) switching departments, or (4) leaving the job (for another position elsewhere or to retire). Physicians described coping with long work hours, long commutes, and evening and weekend work by reducing clinical work hours in order to spend more time with family or pursue other interests. Participants reported that in primary care, the majority of physicians worked less than 1.0 FTE and many new hires requested to reduce clinical work hours once they were past the organization’s initial 2-year introductory period: “The first thing people ask for is to cut back their hours.”

One physician leader described a pattern in primary care she called the “vicious cycle”:

“I’m burnt out at 1.0 now I’m going to cut back to .75 time. Now I’m making fewer RVUs, so they’ll cut my staff, but I’m working the same numbers of hours. Once I get to 0.5 and am still working a full time job effectively, 35 hours, then what is my only option? I’ll work for [local biotech company] or be a same day doc where I don’t have to work on in-basket.”

Another physician, a self-described “recovering leader,” relinquished a leadership role to re-allocate time to patient care and improve personal well-being. Participants noted that female physicians were more likely to reduce FTE and prioritize family over work.

“We know what’s actually happening is people fractionally quit, so that they can manage the in-basket. And it’s not just women anymore, but so often it lands on some of the female providers because they are managing children’s schedules. And I honestly think women are like, you know what? I’m working part-time because it’s how I can cope.”

DISCUSSION

Many of the factors causing, and strategies to improve, burnout, mentioned by participants, are greatly intertwined. Participants identified organizational factors as having the biggest impact on burnout and well-being, including the fee-for-service business model, the growth of EHR-related tasks, and growing focus on productivity metrics. These factors contributed to increasing workloads, loss of autonomy, and feeling de-valued. The exacerbation of stressors in primary care, e.g., lack of support staff, increased in-basket messages, and a shortage of physicians, led to a feeling that primary care is facing a crisis. Leaders also expressed a need for enhanced support, and research suggests that helping leaders succeed could have a powerful positive effect on the whole care team.32,33

Strategies suggested to improve well-being included: decreasing and/or compensating desktop medicine work;27 ensuring adequate support staff;34 enhancing physician autonomy; and ensuring physicians feel valued and heard.21 Participants also had recommendations that had not previously been discussed in the literature, including providing support staff for all physician clinic work hours (not just patient visit hours), allocating or scheduling time for responding to increasing in-basket messages,27 and ensuring leadership affirms the value of frontline physicians by telling them, and more importantly, demonstrating through action that “your opinion counts.” Individual factors such as sleep, exercise, and mindfulness were cited as helpful, but insufficient to counter the potent organizational forces contributing to burnout.

The feasibility of implementing the solutions suggested by physicians merits further attention. Some have the potential to be implemented relatively quickly, e.g., reducing the emphasis on metrics, enforcing panel size limits, celebrating accomplishments, and spending more time on medicine at physician meetings, while others would require substantial lead time and potentially challenging negotiations with organizational stakeholders, e.g., increased number or reallocation of support staff, re-allocating how revenue is distributed throughout the organization, ensuring physicians feel valued and heard, and negotiations with internal and external stakeholders to change reimbursement and compensation models.

These interviews highlight how physician well-being is linked to staffing and the EHR. Physicians who report their clinics are under-staffed or have high staff turnover exhibit higher rates of burnout.34 Reducing desktop medicine and increasing staffing were the most commonly mentioned strategies for improving well-being. Expanded care teams have demonstrated value in reducing physician EHR time in primary care.20 Assisting physicians with certain clinical tasks may also increase physician well-being.6 This organization has taken recent steps to improve well-being, including creating an office of physician wellness, piloting multidisciplinary in-basket support, providing enhanced leadership development support and training, and implementing a primary care transformation program including in-basket support from APCs. Physicians also stressed that relationships with colleagues and patients enhanced their own well-being. In addition, increasing physician well-being will likely require addressing the well-being of all clinic staff.

Changes in the profession of medicine and the demographics of the workforce may also be escalating physician burnout. In recent decades, health systems have grown and turned their attention to standardization and productivity metrics. Ariely and Lanier35 suggest that physician work environments are becoming a “fixing-people production line.” Some evidence suggests that organization size is related to burnout; physicians in small solo practices have much lower burnout rates than physicians in large groups, perhaps because of greater autonomy.36 These trends are accompanied by an increasingly female workforce who typically bear greater responsibility for caring for young children, often choose to reduce work hours rather than reduce family time,37,38,39 and may face additional discrimination.40

These interviews contribute personal insights about how physicians experience burnout and solutions they believe would enhance well-being, which are often not available in structured surveys.41 It is essential for organizations to learn from their own physicians when developing strategies to improve well-being. Well-intentioned efforts promoting individual-level interventions (e.g., mindfulness training) to improve workplace wellness may be alienating if they are at odds with physician perceptions of organizational pressures as the root cause of burnout.

Limitations

Interview participants worked at one predominately fee-for-service organization with an oversampling of physicians in leadership roles and primary care. Future research should seek the perspectives of physicians in other specialties with high rates of burnout.7,42 This organization’s location in northern California with its strong economy, growing population, and high cost of living may pose additional challenges for physicians and staff. This organization was an early adopter of the EHR and patient portals, and its physicians may receive an above average number of in-basket messages. However, other systems may soon face similar increases in in-basket work.

Conclusion

Participants in these interviews perceived that the causes of burnout were partly due to the health system's increased focus on productivity in recent years. Ultimately, the biggest gains in physician well-being may require broader changes to how financial incentives are structured in healthcare, changing culture, redistributing EHR work among physicians and support staff, and redesigning current EHR systems and staffing models, so that physicians spend more time doing work they find meaningful.