Abstract
PURPOSE Physicians have joined larger groups and hospital systems in the face of multiple environmental challenges. We examine whether there are differences across practice ownership in self-reported work environment, a practice culture of learning, psychological safety, and burnout.
METHODS Using cross-sectional data from staff surveys of small and medium-size practices that participated in EvidenceNOW in Virginia, we tested for differences in work environment, culture of learning, psychological safety, and burnout by practice type. We conducted weighted multivariate linear regression of outcomes on ownership, controlling for practice size, specialty mix, payer mix, and whether the practice was located in a medically underserved area. We further analyzed clinician and staff responses separately.
RESULTS Participating were 104 hospital-owned and 61 independent practices and 24 federally qualified health centers (FQHCs). We analyzed 2,005 responses from practice clinicians and staff, a response rate of 49%. Working in a hospital-owned practice was associated with favorable ratings of work environment, psychological safety, and burnout compared with independent practices. When we examined separately the responses of clinicians vs staff, however, the association appears to be largely driven by staff.
CONCLUSIONS Hospital ownership was associated with positive perceptions of practice work environment and lower burnout for staff relative to independent ownership, whereas clinicians in FQHCs perceive a more negative, less joyful work environment and burnout. Our findings are suggestive that clinician and nonclinician staff perceive practice adaptive reserve differently, which may have implications for creating the energy for ongoing quality improvement work.
INTRODUCTION
The Heart of Virginia Healthcare collaborative is 1 of 7 regional efforts supported by the Agency for Healthcare Research and Quality (AHRQ).1 In 2015, AHRQ launched EvidenceNOW: Advancing Heart Health in Primary Care, a multiyear pragmatic trial. EvidenceNOW is designed to generate information about the effectiveness of external quality improvement support in helping small and medium-size primary care practices use patient-centered outcomes research findings to improve the heart health of their patients. The project also is designed to determine what works best in developing organizational capacity for ongoing practice improvement in primary care.
Our EvidenceNOW intervention is based on the theory that, faced with a complex environment, practices must attain the adaptive reserve needed for higher order practice transformation.2 Our approach is predicted on the assumption that attention to overall practice vitality and function—across clinicians and nonclinician staff—is essential to quality improvement activities, including attention to cardiovascular risk factors.3 Adaptive reserve contains many elements, including a positive or joyful work environment and a culture of learning. These elements, along with psychological safety, reflecting a shared belief that a team is safe for interpersonal risk taking, are hypothesized to underlie successful change.4
But how large should the optimal team be? Independent physicians are joining hospital systems or larger groups at a rapid pace, partly to achieve better bargaining power in the face of dominant health plans, and partly to exploit natural economies of scale in information technology, quality improvement, and reporting systems, especially as performance incentives have increased among public and private payers.5 In addition, physicians leave independent practice in response to burnout from a lack of career fit 6 or dissatisfaction with staffing models that require performing major nonphysician, non–revenue-generating work.7,8 Finally, the Karasek model postulates that psychological strain results from the joint effects of conflicting demands of a work situation and the decision-making discretion available to the individual facing those demands.9
Across all physician types, the share of physicians who own their own practices was 51% in 2014, down from 76% in 1983.10 Many of these practices are now hospital owned or affiliated. By 2014 the proportion of physicians who worked in hospital-owned or hospital-affiliated practices had reached one-third. As a by-product, primary care physicians are working in larger groups. In 2015, more than two-thirds of primary care physicians were in groups of 11 clinicians or more.11 Given that our participating practices represented independent and hospital system practices, as well as federally qualified health centers (FQHCs), we hypothesized that practices would differ by their ownership structure in the degree of clinician and nonclinician burnout, work environment, and psychological safety at baseline.
As clinicians shift to larger practices, so do their nonclinician staff. Previous studies have found that burnout varies substantially by occupation and has been found to be higher for clinicians than staff.12 At the same time, nonclinician staff may perceive reduced financial risk and therefore greater job security when practices are under hospital ownership. Existing studies provide little insight on ownership differences on adaptive reserve and burnout, whether for clinicians or staff.
Given the current trend for primary care clinicians to become employees rather than independent practice owners,13 we sought to test for cross-sectional, baseline differences across independent and hospital-owned practices on organizational capabilities and burnout. Previous literature on practice ownership has addressed differences in overall cost of care,14–16 rates of preventable hospital admissions,17 hospital choice,18 structures needed for quality improvement work,19,20 and quality measures.21,22 Our study extends this literature, comparing independent and hospital system practices on both personal and environmental psychological factors that can potentially support quality improvement efforts. We also report on these outcomes for FQHCs, which function as vital access points for underserved populations. We hypothesized that clinicians in FQHCs would report a positive organizational environment, reflecting a shared commitment to underserved populations, but also higher burnout given their challenging financial environment and high patient demands.
METHODS
The Heart of Virginia Healthcare study recruited small and medium-size primary care practices to participate in an intervention to “restore joy in practice” as a means of improving quality. This article reports on cross-sectional analyses of baseline survey data collected in 2016 and 2017 related to practice characteristics, self-reported burnout, a positive work environment, psychological safety, and a practice culture of learning.
Recruitment
Practices were initially recruited through connections with Virginia’s allopathic medical schools, the Virginia Quality Improvement Organization (Health Quality Innovators), and the Health Information Technology Extension Program in Virginia. Additionally, practices were approached at the Virginia Academy of Family Physicians and the Virginia chapter of the American College of Physicians annual meetings. To incentivize participating practices, we offered payments of $500 per practice to complete survey questionnaires. Our study was approved by the Virginia Commonwealth University and George Mason University institutional review boards.
Data and Study
Our primary outcomes included measures of a positive work environment, a culture of learning, psychological safety, and burnout. These were obtained from baseline surveys of clinicians and staff. All survey questionnaires were administered online with paper follow-up. Our primary outcomes were items from the adaptive reserve instrument, specifically items which represent a culture of learning, psychological safety, and a positive work environment,23,24 as well as a measure of burnout.
A positive work environment was captured with 2 adaptive reserve items, “most people in the practice seem to enjoy their work” and “this practice is a place of joy and hope.” A learning culture was captured with “it is hard to get things to change in our practice,” whereas psychological safety was captured with responses to “members of this practice are able to bring up problems and tough issues.”4 Each of these items was ranked by respondents on a 5-point scale from strongly disagree to strongly agree. Adaptive reserve responses were recoded as appropriate so that higher values reflected more positive outcomes.
Individuals also were asked about their degree of burnout using a nonproprietary single-item burnout indicator that instructs respondents to define burnout for themselves. The 5 possible response categories were as follows25: (1) “I enjoy my work. I have no symptoms of burnout.” (2) “Occasionally I am under stress, and I don’t always have as much energy as I once did, but I don’t feel burned out.” (3) “I am definitely burning out and have one or more symptoms of burnout, such as physical and emotional exhaustion.” (4) “The symptoms of burnout that I am experiencing would not go away. I think about frustration at work a lot.” (5) “I feel completely burned out and often wonder if I can go on. I am at the point where I may need some changes or may need to seek some sort of help.” This item was used in the Physician Worklife Study26 and in the Healthy Workplace Trial27 and has been found to perform well against a propriety single-item version of the Maslach Burnout Inventory emotional exhaustion measure.28 In our statistical models, we tested the respondents’ impact on the full burnout scale and also on an indicator of burnout as defined by a rating of 3 or higher.
A second questionnaire was completed by a lead clinician or practice manager and contained measures of practice characteristics, including ownership, practice size, single or multispecialty practice, whether the practice was recognized as a primary care medical home, whether the practice was located in a medically underserved area, and payer mix. Responses on total practice size were used to calculate response rates on the individual-level surveys. Practices were categorized as independent vs owned by a hospital system or as a FQHC. Independent practices included members of independent practice associations that were not owned by a hospital system. Ownership responses were confirmed by telephone, and status as a FQHC and whether the practice was located in a designated medical underserved area were confirmed against public reports maintained by the Health Resources and Services Administration. Practice size was organized into 3 categories: 1 to 5 clinicians, 6 to 10 clinicians, and 11 or more clinicians.
Statistical Approach
We conducted ordered probit regression models because our outcomes were ordered categories. Individual survey-based outcomes were weighted by the inverse of the response rate within practice, and standard errors were adjusted for clustering at the practice level. We first analyzed all responses, and then we separately analyzed clinician and staff responses. We controlled for the following covariates: practice size, whether single or multispecialty mix, whether the practice was accredited as a primary care medical home, whether the practice was located in a designated medically underserved area, the percentage of patients with Medicare coverage, and the percentage of patients with Medicaid coverage or who were not insured.
RESULTS
Study Population
We secured the participation of 67 independent practices, 116 hospital system practices, and 24 FQHCs, which represents roughly 20% of eligible small and medium-size primary care practices in Virginia.
Our response rate for the individual survey was 48% and 93% for the practice characteristics survey. Of 2,109 returned questionnaires, 2,005 with complete responses on all relevant items were used in the analysis. The 2,005 responses were obtained from 190 practices, including 700 questionnaires from 61 independent practices, 931 from 104 hospital system practices, and 374 from 24 FQHCs. Twenty-four percent of survey respondents were clinicians after adjusting for weighting (n = 471). The baseline characteristics of the individual survey responses overall, and by whether they are independent, hospital owned, or FQHC, are displayed in Table 1.
Survey Respondent and Practice Sample Characteristics
Culture of Learning
Survey respondents rated their practices similarly on culture of learning, ie, how difficult it is to get things done, across independent, hospital system, and FQHC practices (Table 1). On a scale of 0 to 5, the overall average was 3.0, or neutral. Unadjusted differences across practice types were not statistically significant, and in multivariate models controlling for practice size, practice type, accreditation for primary care medical home, and payer mix, the differences across independent vs hospital owned and independent vs FQHC were also not statistically significant (Table 2 and the Supplemental Table, http://www.annfammed.org/content/16/Suppl_1/S44/suppl/DC1).
Association of Practice Ownership With Work Environment, Learning Culture, Psychological Safety, and Burnout, All Survey Respondents
Work Environment
On average, staff reported a relatively positive work environment at a mean rating of 3.6 (95% CI, 3.5–3.6) for “enjoyment with work” and of 3.5 (95% CI, 3.4–3.6) for practice is a “place of joy and hope” as shown in Table 1. Given the 5 response categories, the mean responses fall between neutral and agree. Adjusting for covariates, being a hospital-owned practice was associated with a higher rating on enjoyment with work (β = 0.273; 95% CI, 0.014–0.533) relative to independent practices (Table 2). Table 2 also shows the estimated incremental difference of hospital ownership relative to independent practices at each response category. Thus, hospital ownership is associated with being 6% more likely to report strongly agree with enjoyment with work and 1.7% less likely to report strongly disagree. Differences in ratings of the practice as a place of joy and hope between hospital-owned and independent practices, however, were not statistically significant. Further, there were no statistically significant differences in the work environment ratings between independent practices and FQHCs; therefore, estimated incremental differences by response category between FQHCs and independent practices are not displayed in the table.
Psychological Safety
Self-reported psychological safety was rated at 3.5 on average in the total sample, between neutral and agree. Higher ratings were reported in hospital-owned practices than in independent practices, adjusting for covariates (β = 0.363; 95% CI, 0.118–0.609), and hospital ownership is associated with being 8.5% more likely to report strongly agree with the question on psychological safety and 2.7% less likely to report strongly disagree. In contrast, FQHCs experienced lower psychological safety than independent practices, but differences were not statistically significant, adjusting for covariates.
Burnout
Measured on a 1 to 5 scale, burnout was rated at 2.1 on average, which roughly corresponds to occasionally being under stress. Adjusting for covariates, being in a hospital-owned practice was associated with reporting lower burnout values relative to independent practices (β = −0.287; 95% CI, −0.509 to 0.063). Hospital ownership was associated with being 9.8% more likely to report “I enjoy my work. I have no symptoms of burnout” and 2% less likely to report “I feel completely burned out and often wonder if I can go on.” Differences between FQHCs and independent practices were not statistically different.
Separately, we ran logistic models where burnout was defined at a threshold of 3 or higher. Overall, 24% of respondents reported burnout at this threshold or higher, with rates lowest for hospital-owned practices (20%) and highest for independent practices (28%). Adjusting for other covariates, the odds of being at or above the burnout threshold were not statistically different across practice ownership type (results not shown).
Clinicians vs Staff Respondents
We then examined responses from clinicians and staff separately (Table 3a). Clinicians include physicians, nurse practitioners, and physician assistants. Among clinicians there was no statistically significant difference across independent and hospital-owned practices in ratings across culture of learning, work environment indicators, psychological safety, or burnout. Clinicians in FQHCs, however, on average reported lower ratings of culture of learning (β= 0.470; 95% CI, −0.919 to −0.021), lower ratings of the work environment as measured by enjoyment with work (β = −0.552; 95% CI, −0.996 to −0.109), and the practice as a place of joy (β = −0.567; 95% CI, −1.089 to −0.405). FQHC clinicians also reported lower psychological safety (β = −0.723; 95% CI, −1.194 to −0. 252). Incremental differences by response category are reflected in Table 3a for FQHC vs independent practices. Differences in ratings of burnout between clinicians in FQHCs and independent practices were not statistically significant, however. We ran separate models for physicians only and found qualitatively similar results (results not shown).
Association of Practice Ownership With Work Environment, Learning Culture, Psychological Safety, and Burnout, Clinician Respondents
Among staff (Table 3b), there were no differences between FQHCs and independent practices on any of the outcomes. Staff employed by hospital system practices, however, reported greater psychological safety than staff in independent practices (β = 0.372; 95% CI, 0.131–0.613) and lower burnout (β = −0.328; 95% CI, −0.608 to −0.049), adjusting for covariates. The corresponding incremental differences between hospital-owned and independent practices are shown in Table 3b. Differences in ratings of enjoyment with work and the practices as a place of joy and hope were not statistically significant.
Association of Practice Ownership With Work Environment, Learning Culture, Psychological Safety and Burnout, Staff Respondents
DISCUSSION
We found more favorable ratings of work environment, psychological safety, and burnout in hospital-owned practices compared with independent practices, controlling for a variety of practice characteristics. These differences were largely driven by staff rather than clinician respondents. Although our results broaden our understanding of practice ownership with respect to a variety of outcomes, our results do not explain why nonclinical staff at hospital-owned practices appear to be in a better place psychologically compared with their clinician colleagues.
We also found that clinicians in FQHCs were less likely to report a positive work environment than clinicians in independent practices, although clinicians in FQHCs did not report greater burnout. That clinicians in FQHCs generally see their practices as more of a learning environment is not surprising. The relatively strong culture of learning in FQHCs may reflect a history of learning support from the US Healthcare Resources and Services Administration and the state primary care association (the Virginia Community Healthcare Association). Our hypothesis of higher scores among FQHC clinicians on other indicators of a positive work environment was not supported, possibly related to the FQHC’s patient population, staffing, resources, reimbursement, or other factors. We are in the process of conducting a qualitative assessment of practices to better understand these differences.
Our study has several limitations, including that results may not generalize to other practices within or outside Virginia. Because the literature provides little guidance on the clinical importance of the differences we measured for select adaptive reserve items, results should be interpreted with caution. In addition, each dimension of change was measured with a limited number of questions. Given the size of our sample, we were not able examine physicians separately from other clinician types. Our survey also was not designed to identify the mechanisms that drive the differences in outcomes across staff roles. The association between ownership and outcomes may be causally driven by the ownership structure, but it also may be driven by sorting of clinicians into practices by type. Further research should explore the potential reasons for ownership differences in outcomes and the clinical significance of differences in responses to adaptive reserve questions.
To our knowledge, this study is the first that compares independent practices, hospital system practices, and FQHCs on elements from the adaptive reserve survey that address culture of learning, enjoyable work environment, psychological safety, and burnout—elements that we see as essential aspects of joy in practice.29 Our findings suggest that staff may be critical to building overall practice adaptive reserve and thus creating the energy for ongoing quality improvement work. Staff also provide a novel validation of the importance of team-based care models in primary care.
Footnotes
Conflicts of interest: Dr Nichols. received honoraria from the nonprofit Rocky Mountain Health Plans (Grand Junction, Colorado) for organizing content about national trends and facilitating board retreat discussions, and from the American Medical Association for speaking at their annual state advocacy conference on antitrust issues. Dr Nichols is also the Director of George Mason University’s Center for Health Policy Research and Ethics, whose 501(c)(3) Foundation account received a grant from America’s Health Insurance Plans to support graduate student work on health care market issues. Dr Nichols is a member of the unpaid Board of Trustees of the National Committee for Quality Assurance, and an unpaid advisor on payment reform matters to the Patient-Centered Primary Care Collaborative. All other authors declare no conflicts of interest.
Funding support: Publication of this article was supported by the Agency for Healthcare Research and Quality (AHRQ) through contract No. HHSA290201200019I, and grant No. R18 HS023913.
Disclaimer: This work represents the opinions of the authors and should not be interpreted as official positions of the Agency for Healthcare Research and Quality or the US Department of Health and Human Services.
Supplementary materials: Available at http://www.AnnFamMed.org/content/16/Suppl_/S44/suppl/DC1/.
- Received for publication June 15, 2017.
- Revision received December 15, 2017.
- Accepted for publication December 22, 2017.
- © 2018 Annals of Family Medicine, Inc.
References
In this issue
Jump to section
Related Articles
Cited By...
- Emergence of Gun Violence as a Patient Priority
- The EvidenceNOW Practice Support Initiative: The Heart of Virginia Healthcare
- Emergence of Gun Violence as a Patient Priority
- Family physician perspectives on primary care reform priorities: a cross-sectional survey
- Engagement of Small to Medium-Sized Primary Care Practices in Quality Improvement Efforts
- Medicare Access and CHIP Reauthorization Act in Small to Medium-Sized Primary Care Practices
- Implementation of Community-Based Resource Referrals for Cardiovascular Disease Self-Management
- Indicators of Workplace Burnout Among Physicians, Advanced Practice Clinicians, and Staff in Small to Medium-Sized Primary Care Practices
- A Qualitative Analysis of Implementing EvidenceNOW to Improve Cardiovascular Care
- Does Ownership Make a Difference in Primary Care Practice?
- Finding a Parsimonious Path for Primary Care Practice Transformation
- The Capacity of Primary Care for Improving Evidence-Based Care: Early Findings From AHRQs EvidenceNOW