Primary care is the cornerstone of high-quality health care systems across the world, but medicine in the United States is struggling to meet its current and future primary care needs.1 Students demonstrate little interest in primary care careers,2 and even those who choose primary care training programs often depart for other specialty areas along the way.3 The reasons for this are multiple and well documented: long hours, relative under- reimbursement, excessive clerical burden, and lack of respect for primary care skills all play a role.4 In addition, burnout and job dissatisfaction are prevalent and threaten primary care and the health care system as a whole.5

Although much has been written about physician burnout and job dissatisfaction, relatively little is known about how best to combat or even begin to address these issues. Training in individual skills such as mindfulness has been shown to be beneficial,6 and institutionally-funded physician small group curricula have also demonstrated benefit.7 To this limited literature, Linzer and colleagues8 add evidence that interventions to improve the structure of health care delivery and the environment in which physicians work are important to physician well-being.

In their Healthy Work Place study, Linzer and colleagues conducted a cluster randomized trial of interventions to address clinician work conditions in 34 diverse primary care practice settings. The specific interventions varied across clinics according to local needs identified through discussion and review of baseline assessment of worklife and work conditions among local clinicians and research study staff. This process was intended to stimulate awareness and prompt conversations among clinic stakeholders to inform selection of high-yield interventions at each study site. However, the interventions generally grouped into broad categories that addressed communication, workflow, and quality improvement for health care delivery. Notably, every intervention targeted structural workplace issues rather than individual clinician-level strategies to cope with stress. Although the proportions of clinicians experiencing improvement in burnout and job satisfaction during the study were modest, these outcomes improved more frequently among clinicians working at intervention clinics upon follow-up at between 12 and 18 months.

This study imparts several important messages. First, the multi-site randomized controlled trial design is unique in the literature to date on physician burnout and job satisfaction. Prior studies of interventions in this area have seldom been randomized, but this level of evidence is necessary to truly determine which approaches are effective and for which clinicians. We hope this study emboldens others to test interventions with similarly robust study designs. Second, this study follows an approach with parallels to quality improvement, in which it is important to ‘build capacity’ in primary care.9 This capacity building does not entail a one-size-fits-all approach, but rather entails creating the understanding, the infrastructure, and the commitment to address the problem. Capacity building requires an ongoing focus on collecting and using data to understand and improve the performance of a practice, with the support of leadership and engagement of the community of practitioners. Within the context of an environment committed to reflection and continuous improvement, specific projects can emerge that meet local needs. Third, this study emphasizes the now widespread view of primary care as a team-based specialty in which the nature of a physician’s interpersonal interactions and workflow with others in the practice are critical determinants of well-being. Finally, this study serves to remind us that protecting and promoting physician well-being is a responsibility in which health care organizations, and indeed the profession itself, must share.10 Self-regulation is a privilege of the most esteemed professions,11 and this responsibility should extend beyond clinical competence to a commitment to the welfare of every physician.

The Linzer study advances us toward this goal, but does have limitations. In particular, statistically significant improvements were found only for the secondary outcomes of proportion of clinicians with improvements in burnout and job satisfaction. The prespecified primary outcomes of overall burnout rates and job satisfaction levels did not statistically significantly differ between the intervention and control clinics, although the raw data suggest effects in the direction of improvement. Also, the individualized and heterogeneous nature of the interventions at each clinic site makes it difficult to know which approaches might best be selected by clinics interested in replicating these findings. In fact, the complexity and dynamic nature of the interventions makes it impossible to know whether exporting specific strategies to new practice sites would offer the same benefit. It is important to note that because the confidence intervals for the effects of each intervention category are broadly overlapping, it is unclear whether any category was more likely than the others to reduce burnout and improve job satisfaction. Rather than prescribing any specific intervention, then, this study should stimulate practices to examine their own data and collaboratively consider how best to adapt the categories of interventions studied by Linzer’s group to their local environments. As the authors note, more focused studies of specific types of interventions will be needed to clarify the individual and collective roles of each intervention category.

Finally, the single-item burnout measure used in this study addresses only one domain of burnout, emotional exhaustion. A recent direct comparison of this item against alternative single items and the full Maslach Burnout Inventory suggests that the item used in this study performs adequately in assessing emotional exhaustion, but may not be the optimal measure of this domain.12 Because burnout among physicians also involves other key domains, especially depersonalization, studies incorporating metrics evaluating the full burnout construct are needed. Although physician well-being is a worthy goal in and of itself, and effects on patient outcomes may be mediated by many other factors and therefore be difficult to demonstrate, studies that address physicians’ work satisfaction should also examine the impact of interventions on patients to further inform whether improved physician experience translates to better patient care.

In summary, Linzer and colleagues have applied a sophisticated study design to advance our knowledge of interventions to improve physician work conditions, burnout, and job satisfaction within the context of clinical practice teams. Protecting and promoting physician well-being should be viewed as a shared responsibility. Individual physicians must be accountable for their role in their own well-being, but healthcare organizations and the medical profession itself must recognize their obligation to promote health care delivery within a viable, healthy physician work environment.