Cochrane in CORR®: Preventing Occupational Stress in Healthcare Workers

Importance of the Topic

Wellness and burnout among healthcare workers, including physicians, has been prominently featured recently within the medical community and media, including in the pages of Clinical Orthopaedics and Related Research® [9, 19]. The term “burnout” was first coined by American psychologist Herbert Freudenberger, who described burnout as “the consequences of prolonged stress and anxiety experienced by people working in the healing professions” [7, 12]. A survey of 6880 physicians among all fields in 2011 found that 45.5% reported symptoms of burnout [16]. The 2014 update of a survey of 3310 physicians showed a rise to 54.4% [14]. Most recently in 2018, a survey of 1643 Chinese arthroplasty surgeons with 202 responses reported burnout in 85.1% of respondents [20].

Burnout is often accompanied and complicated by feelings of depression, substance abuse, interpersonal conflicts, and suicidal ideation [12]. As of 2015, approximately 300 to 400 physicians commit suicide each year in the United States [8]. This is approximately one physician per day, or three to four averaged sized American medical school classes per year, committing suicide [3].

Burnout is on the rise among physicians, and orthopaedic surgeons are no exception. The American Orthopaedic Association (AOA) performed a survey of 195 orthopaedic chairs and program directors to assess the severity of burnout and career dissatisfaction [1, 14]. In 2002, 10% reported being dissatisfied; by 2007, this had already increased to 26% [14], which attracted the attention of AOA and led them to encourage the integration of strategies to identify at risk physicians in a nonpunitive manner with treatment and education programs [1, 14]. Quite apart from its effect on physician satisfaction, burnout places patients at risk, as it is associated with decreased quality of care, increased medical errors, and depersonalization [6].

Upon Closer Inspection

This Cochrane review [13] contained 48 randomized controlled trials, eight cluster-randomized trials, four crossover studies, and four controlled before-and-after studies of a work directed intervention for a total of 58 studies. The total number of participants was 7188 with 3592 in intervention groups and 3596 in control groups. Interventions were grouped to cognitive-behavioral training techniques (CBT), mental relaxation techniques such as mindfulness, physical relaxation like a massage, or organizational interventions such as schedule changes and mentoring.

Relaxation interventions were found to have low-to-moderate quality of evidence for stress reduction. Low-quality evidence was found for the reduction of stress using cognitive-behavioral training and organizational interventions. Overall, the quality of evidence for the studies included was low, and risk of bias was high because of difficulties with blinding, incomplete outcome data, selective reporting, and low compliance with interventions. We also question whether the interventions studied in these trials were adequate to the task. Among surgeons, stress commonly arises from one or more of these factors: Interpersonal relationships and responsibility, sense of subjectivity around many decisions, and pressures of quality patient care [1]. We do not believe that in most cases a single study intervention can address all three of these components, and most trials in this Cochrane review focused on single-intervention trials.

Additionally, most of the included randomized clinical trials contained fewer than 120 participants. Of the 58 studies included, the number of participants included was larger than 60 in only 26 studies [13]. In order to find a meaningful effect, it is critical that an intervention has appropriate power and sample sizes. When designing a randomized clinical trial for assessing burnout and stress, the authors of this Cochrane review suggested that a sufficient sample size would need to be at least 110 participants [13]. We agree with the finding that most of the studies included in this review were of smaller sample sizes, only 17 containing more than 110 participants, and an increase in sample size would produce more-reliable results [13, 17]. A small trial could only detect modest-to-large differences with adequate power and may miss the subtle but still important effects of the intervention being studied. A study with 110 participants in this setting might only have 80% power to detect a reduction in stress or burnout.

Finally, there was a great deal of heterogeneity of populations studied, and less than one-quarter of the studies (14 of 58) included physicians at all (many focused on nurses and other providers); none focused on orthopaedic surgeons. Clearly, we need more studies in our own specialty, as results of some of these studies may not apply to orthopaedic surgeons at all.

Take-home Messages

The proportion of orthopaedic surgeons who show signs of burnout range from 50% to 85%, and the prevalence of burnout is increasing [1, 6, 20]. Stress is a major component of burnout and has been shown to be detrimental to the health of both physicians and patients [6]. Future studies need to focus specifically on the impact of stress-reduction interventions; endpoints of interest might include validated outcomes tools that measure symptoms of stress such as the Maslach Burnout Inventory, an inclusion of self-reported metrics on suicidal ideation or attempts, or performance metrics related to patient outcomes or patient satisfaction on matched cohorts.

This Cochrane Review [13] found low-to-moderate evidence favoring stress management with CBT and organizational changes such as schedule changes and mentoring. Both mental and physical relaxation techniques (mindfulness, massage, and exercise) were associated with a modest reduction in stress compared to no intervention at 1 and 6 months followup. Additionally, CBT (including mental relaxation) moderately reduced stress when compared to no intervention. However, CBT was not found to be more effective at stress reduction than computer training or passive attendance of a psychologist at staff meetings. By contrast, organizational interventions such as changes in working conditions, schedule changes, or mentorship had small reductions of stress.

Given the risk of physician suicide and patient harm, we need more and better quality studies to adequately evaluate the effects of interventions not just on physician wellness but also on patient care [1, 6, 13]. Ideally, future studies should focus on physician stress management, stress levels, and patient outcomes in a randomized clinical trial setting, with at least a sample size of 110 with orthopaedic surgeons as participants. Although, to our knowledge, there have not been any large randomized clinical trials since the review, there have been multiple smaller randomized clinical trials [2, 4, 5, 10, 11, 15, 18], many of which include physicians as the primary focus and these studies merit an update. Multiple orthopaedic professional societies including the AOA and Canadian Orthopaedic Association currently are investigating the prevalence and causes of stress and burnout among their membership, as well as prevention strategies. We look forward to their findings.