In a national sample of >14,900 U.S. physicians, those with a physician well-being index score ³3 were at greater risk for a number of adverse outcomes including a 2-fold higher risk of reporting a recent medical error, a 5-fold higher risk of burnout, 4-fold higher risk of severe fatigue, and 2-fold higher risk of suicidal ideation, and 3-fold higher risk of poor overall quality of life [22]. Our Institution had a representative population compared to the United States. Our Institution included 60% physicians and 40% APPs which was reflective of the whole nation. While we analyzed the data of the physicians across the nation in the study that included 14,900 physicians, our own data was also compared to the Well-Being Index scoring across the nation with both physicians and APPs, with a similar make-up to our own group. A total of 41% of providers at our Institution had a well-being index score of ³3 at baseline which was similar to that of U.S. physicians nationally (39%). This striking number of physicians at our Institution and in the United States nationally who possess a high level of distress and numerous features of burnout corresponds closely to the approximately 50% of providers as reported in the literature [1, 4-6].
Following the implementation of the Well-Being Task Force at our Institution, the average overall well-being index scores of our Institution’s providers decreased from 1.76 at baseline to 1.32 2 years later compared to an increase in well-being index scores of U.S. physicians nationally. The myriad tactics ranging from Epic optimization, leadership engagement, onboarding, the NGaged program to heightened communications between providers and well-being champions most likely all contributed to this decline.
Provider Gender
Several studies have reported that female providers have a 20-60% increased odds of burnout [4, 5, 7, 23]. In Houkes and colleagues’ self-reported questionnaires of 340 general practitioners, burnout in men was primarily associated by depersonalization, while emotional exhaustion was most likely to cause burnout in women [23]. These authors speculate that men choose avoidance and withdrawal coping strategies whereas women become exhausted but do not depersonalize. In Shanafelt and colleagues’ survey of 7288 physicians, female physicians were more likely to be dissatisfied with work-life balance compared to their male colleagues (p=0.002) [4]. Both male and female providers’ average well-being index scores at our Institution decreased over the 3 years of this study while providers’ scores nationally increased for both genders. The male and female average well-being index scores were similar at baseline at our Institution whereas the female average well-being index scores were lower than those for males at both 1 year later and 2 years later. These findings may particularly reflect the impact of focused efforts made at our Institution to target a specific provider group such as formal lectures given by female leaders and social events designed for female providers.
Years since medical school graduation
It has been reported that younger physicians are at an increased risk of burnout symptoms [4], with those <55 years old at more than double the risk of those >55 years old [7]. In Del Carmen and colleagues’ survey study of 1774 physicians in 2014 and 1882 physicians in 2017, early career physicians (£10 years since training) were more susceptible to burnout (odd ratio, 1.36), while physicians in their late career (>30 years since training) were less vulnerable (odds ratio, 0.59) [10]. Our current study corroborates these findings as the highest average well-being index scores were highest for providers both at our Institution and nationally who graduated from medical school 15-24 years earlier. We also noted a peak in providers’ average scores who graduated from medical school <5 years earlier and the lowest average scores in those who graduated ³25 years earlier. The two provider groups that were at the highest risk for burnout warrant particular attention and intervention. Onboarding engagement may benefit providers who graduated from medical school <5 years earlier, while Epic optimization, enhanced communication between administrators and providers, and well-being champions who serve as support personnel for providers in need may be valuable for providers who graduated from medical school <5 years earlier and 15-24 years earlier.
Provider medical specialty
In Shanafelt and colleagues’ survey of 7288 physicians, physicians practicing emergency medicine, general internal medicine, family medicine, neurology, or radiology had the highest risk of burnout, whereas dermatologists had a lower risk [4, 5]. Furthermore, physicians practicing dermatology, general pediatrics, and preventive medicine had the highest rated satisfaction with work-life balance, while physicians practicing general surgery, general surgery subspecialties, and obstetrics/gynecology had the lowest rates. These authors attribute the highest burnout rates to working primarily in the front line of access to care, except for pediatrics. Similar findings were encountered in the present study as obstetricians/gynecologists and internal medicine physicians had the highest average well-being index scores compared to other medical specialties, while pediatricians had the lowest average well-being index scores.
Physician turnover rate
A high physician turnover rate not only may lead to diminished productivity, low morale, and diminished quality of patient care but also poses a financial burden. The costs of replacing a physician due to recruitment, onboarding, and lost patient care revenue equates to 2-3 times the physician’s annual salary [10, 12]. The turnover rate of our Institution’s providers was extremely low at 5.6% in 2017 and decreased by 30% to 3.9% in 2019.
Causality Between Our Implemented Strategies and Their Perceived Impact
Our internal evaluation prior to the initiation of the 4 strategies (Table 2) indicated that many of our providers were overwhelmed by their work, in particular the large of amount of time spent using Epic outside of work hours. Providers were frustrated by this extraordinary time commitment and lack of efficiency. The lack of work-life balance resulted in sleep loss, diminished energy to enjoy non-work activities, and psychological/emotional burdens. Additionally, there was deficient communication between providers and administration. Combining the feedback from our internal survey as well as Stanford’s and the Mayo Clinic’s established techniques of burnout reduction, we developed our 4 strategies. To address the Epic concern and to improve workflow/office efficiencies, Epic optimization was implemented which allowed providers to spend less time using the EMR during and after office hours and permitted more face-to-face contact with patients. Providers were less frustrated and more resilient and fulfilled, leading to enhanced provider-patient relationships, less administrative burden, and decreased burnout. The APP Onboarding led to happier providers, as reflected by the increased retention over the 3 years of our study. To promote a work-life balance, relationship building through socialization allowed providers and administration to get to know each other in a relaxed, non-work atmosphere. They were able to address any problems or concerns, which resulted in more engagement and less burnout. To enhance the communication between providers and administration, well-being champions forged valuable connections with all providers across different medical specialties which decreased the likelihood of burnout.
The 4 strategies implemented at our Institution were a staged-approach and did not need to be implemented as a package. Communication was the most important strategy as it bolstered trust among providers and administration/managers. The other three strategies built upon the foundation of communication. Workflow/office efficiencies were improved, and provider engagement/growth and relationship building were enhanced.
Strengths and Limitations
The 4 strategies in our study were fashioned from Stanford’s and the Mayo Clinic’s proven success that focused on professional fulfillment and avoidance of burnout. We applied their well-being techniques to the culture in our metropolitan community. Our 3-year study serves as a unique and effective model for incorporating numerous strategies aimed at decreasing provider burnout and boosting well-being in a metropolitan community. These approaches were directed at strengthening our Institution as a whole, replete with building a constructive administration-provider relationship, optimizing providers’ time by decreasing clerical burdens, and developing provider self-care resources to maintain a healthy work-life balance. Following the implementation of changes at our Institution, the well-being of providers at our Institution improved while that of the rest of the nation was either static or declined. The strong association between our implemented strategies and reduction in provider burnout suggests that our strategies substantially contributed to the decrease in provider burnout and improvement in provider well-being. As our approach to reduce provider burnout was based on Stanford’s well-recognized model, our beneficial tactics employed at our Institution in Kentucky may be generalized and applied to healthcare systems in other states.
Our study included providers in all fields of medicine with varied years since medical school graduation which permitted a comprehensive examination of factors that may lead to burnout. Providers were retested annually with the well-being index survey to determine yearly goals and areas for improvement based on provider feedback. The strategies implemented at our Institution led to a more efficient and standardized program that allowed providers to have more dedicated time to fulfilling their clinical and clerical obligations in a more relaxed atmosphere.
While the physician well-being index survey represents a valuable screening tool to improve physician self-awareness and identify physicians who may benefit from further evaluation or support [1, 24], our study only reported the average well-being index scores among providers at our Institution who completed the anonymous survey. In this respect, we were unable to specify the particular provider who may benefit from individualized attention. While we were incapable of identifying the specific providers who scored the highest on the well-being index survey, we hoped that the abundant modifications designed to bolster well-being at our Institution were advantageous for providers who were most at-risk for burnout. Another limiting factor in our work is the relatively low percentage of providers at our Institution who completed the 3 well-being index surveys. The response rates differed by specialties and by whether the respondents were physicians or APPs. There were different engagement levels depending the particular year. With continued dissemination of the strides made by the Well-Being Task Force to all providers at our Institution to combat provider burnout, we are encouraged that more providers will take the survey in the future. We also assume that there was self-selection bias of the providers who took the survey. An additional limitation was that the Engagement Survey was only performed before the initial well-being index survey. As there was an overlap in questions between these 2 surveys, the well-being index survey was the only one given in the follow-up period during our study. We subsequently re-implemented the Engagement Survey which is given concurrently with the well-being index survey.
Our current study is a timely contribution as the National Academy of Medicine proposed systemic changes in healthcare organizations, academic institutions, and all levels of government to create a positive work environment on October 23, 2019 [25]. The goal was to promote professional well-being, enhance patient care, reduce the risk of burnout, and balance job demands and resources [25]. Similar to our study, the proposal intends to regularly assess provider burnout.