Abstract
Background: Primary care is the foundation of health care, resulting in longer lives and improved equity. Primary care was the frontline of the COVID-19 pandemic public response and essential for access to care. Yet primary care faces substantial structural and systemic challenges. As part of a longitudinal analysis to track the capacity and health of primary care, we surveyed every primary care practice in Virginia in 2018 and again in 2022.
Methods: Surveys were emailed or mailed up to 6 times and nonresponders received a phone call. Questions assessed organizational characteristics, scope of care, capacity, and organizational stress in the prior year. From respondents, 39 clinicians, nurses, staff, administrators, and practice managers were interviewed.
Results: 526 out of 2296 primary care practices (23% response rate) completed the survey, with broad representation across geography, ownership, and payer mix. Compared with 2018, in 2022 there were increases in practices owned by health systems (25% vs 43%, P < .0001) and average percent of patients with Medicaid per practice (12% vs 22%, P < .0001). The percent of practices reporting any major stressor increased from 34% to 53% (P < .0001). The main increased stress was losing a clinician, with 13% of practices in 2018 versus 42% in 2022 reporting losing a clinician (P < .0001).
Conclusions: Primary care practices are resilient and continue to serve their communities, including a broad scope of services and care for underserved people. However, the COVID-19 pandemic caused significant stress. With an increase in clinicians leaving clinical practice, we anticipate worsening access to primary care.
- COVID-19
- Health Policy
- Health Services Accessibility
- Family Medicine
- Health Workforce
- Medically Underserved Area
- Medicaid
- Pandemics
- Practice-based Research
- Primary Health Care
- Surveys and Questionnaires
- Virginia
Introduction
With access to primary care, patients and populations experience longer lives and increased health equity.1,2 Primary care, the only specialty to achieve such outcomes, is the foundation of the US healthcare system.1,2 It is essential to promoting health and to preventing and managing chronic conditions. However, primary care in the US has been chronically underfunded and fragmented.3
As part of a longitudinal evaluation of primary care capacity to support Medicaid expansion in Virginia, our team surveyed every primary care practice in Virginia in 2018 with a plan to repeat the survey every 4 years.4 Our 2018 findings indicated that primary care was severely stressed, under-resourced, and overburdened. Since this time, primary care has experienced additional stressors, with an increased demand for services after Medicaid Expansion, which became effective in Virginia in January 2019. In March 2020, the COVID-19 pandemic began, posing an unprecedented challenge for an already strained primary care system.4
Primary care served as the frontlines of the COVID-19 pandemic.5 Primary care has been critical as a site providing testing, treatment, and vaccination, despite facing major barriers and increased stressors.5,6 Recent literature has highlighted increasing stressors in primary care, including the rapid adoption of telemedicine, strained resources, and workforce shortages.6⇓–8 However, there is a lack of repeated measures completed before and during the pandemic to evaluate specific changes encountered during this time period, as well as changes brought about after the expansion of Medicaid. Repeated measurements over time may provide important insights into the adaptation and stressors of primary care.
The Commonwealth of Virginia serves as an important case example. Virginia has a diverse population and varied geography. It ranks 27 out of 50 states for number of primary care clinicians per resident.9 Importantly, prepandemic metrics were evaluated annually beginning in 2018 and can be tracked longitudinally.4
This study offers insight into the changing landscape of primary care with pre and post-COVID-19 practice comparisons of stressors, strengths, and changes in patient demographics. Our analysis also addresses several factors in the National Academies of Sciences, Engineering, and Medicine (NASEM) 2021 report on Implementing High-Quality Primary Care - specifically, addressing the workforce in the Commonwealth.3 Understanding the state of primary care, in particular postpandemic, is critical to implementing and ensuring the sustainability of high-quality primary care.
Methods
We utilized a methodology for primary care practice identification from our 2018 analysis.4 We update this list each year by querying the 2020 National Plan and Provider Enumeration System, the 2020 Virginia Department of Health Professions (VDHP) licensure data, and the 2018 Virginia All-Payer Claims (APC) database to identify primary care physicians and primary care practices in the state; the APCD lags real-time by approximately 18 months.10,11 Using APC data and internet searches of practices and physicians, we nested every physician into practices in health systems. We compared this to our 2018 list and included new practices to survey. The Virginia Commonwealth University Institutional Review Board deemed our study as not human subjects research.
Practice Survey
Practice survey questions were drawn from our prior survey4 and novel questions were added concerning Medicaid Expansion in Virginia, payment programs for primary care including alternative payment models, services that the practice provides, and the impact of COVID-19 and other stressors on the practice. These questions were selected based on experiences from the Virginia Medicaid administrators, the Department of Medical Assistance Services (DMAS), and the Virginia practice-based research network - the Ambulatory Care Outcomes Research Network (ACORN).12 Prior repeat questions were developed based on the VDHP licensure questions, the American Board of Family Medicine Graduate Survey, the American Association of Medical Colleges Physician Survey of Primary Care, and assessments of the patient-centered medical home.12⇓⇓⇓–16 New questions were adapted from the Healthy Michigan Plan’s physician survey.17 We used anticipated or experienced practice changes and competitiveness as markers for practice stress. Finally, we inquired what the practice’s greatest strengths were. The survey is included in the Appendix. 20 out of 36 questions were unchanged from 2018 to 2022.
The survey was fielded, using practice addresses, e-mails, and fax numbers, to all practices in Virginia in an iterative manner. The cover letter included text from the Virginia Primary Care Task Force explaining the use of the survey and its value in completion. The survey included instructions that it should be completed by an individual with detailed information about the practice, in consultation with other practice staff as needed. When we were familiar with a primary care practice, we sent the survey directly to an individual knowledgeable on all aspects of the practice, such as an office manager. Practices were contacted by survey staff via phone between 1 to 6 times as reminders to complete the survey. The survey was distributed by several organizations in Virginia, including the Virginia Academy of Family Physicians, the Virginia Chapter of the American College of Physicians, the Medical Society of Virginia, the Virginia Chapter of the American Academy of Pediatrics, and ACORN. When completed online, only 1 form was permitted per practice. If duplicate surveys were received for a single practice, we included only the first response in our analysis. The survey was fielded between September 2021 and April 2022.
Statistical Analysis
Each practice represented 1 unit. The analysis includes descriptive statistics on practice characteristics, and Chi-square tests to compare practice responses and characteristics between 2018 and 2022. Analyses were conducted using SAS version 9.4 (Cary, NC).
Qualitative Analysis
At the conclusion of the survey, practice staff, clinicians (ie, physicians, physician assistants, nurse practitioners, and nurses), and administrators were offered the opportunity to participate in a semistructured interview to clarify and expand on their survey responses. The interviews were conducted over the phone by 2 research team members. Responses were aggregated and analyzed using an immersion-crystallization process to identify common themes and trends.
Results
526 out of 2296 practices completed the survey (23%). In 2018, 484 practices completed the survey of 1622 primary care practices identified at that time (30% response rate). Respondents had varied roles with detailed information about the practice, with 60% of surveys completed by the office manager or administrative personnel, 24% by a clinician (physician, physician assistant, nurse practitioner, or nurse), and 16% respondents listed as “other.” There was broad representation of practices across geography, ownership, and payer mix (see Table 1). The distribution of primary care practice survey responses in Virginia aligns with the distribution of all primary care practices in Virginia (see Figure 1) and is consistent with the distribution from 2018 survey respondents.4 The percent of rural survey respondents is 48%, similar to 44% of total estimated primary care practices in Virginia. In terms of ownership, 43% of practice respondents are health-system owned, which is consistent with an estimated 35% of practices that are health-system owned across the Commonwealth. Thirty-seven interviews - 36 individuals and 1 group of 3 - were completed with 39 different participants, including clinicians, office managers, administrative personnel, a social worker, and a nurse navigator.
Practice Surveys
In 2022, primary care practices provided care for diverse and often underrepresented people; 91% cared for vulnerable populations and 79% accepted new Medicaid patients (see Table 1). Primary care practices provided varied services (see Table 2), including care coordination or patient navigation (64%), care management for chronic conditions (68%), and follow-up after Emergency Department visits/hospitalizations (57%). Approximately 22% of practices had a Licensed Clinical Social Worker or Professional Counselor, 13% of practices had a Psychologist, and 32% had a case manager, care coordinator, or patient navigator. Practices broadly promoted access; 91% of practices offered telehealth, of which 75% was video and 25% was audio. Services and populations cared for did not vary significantly from 2018, with the exception of an increase in patients with Medicaid (12% vs 22%, P < .0001). Notably, there was an increase in practices owned by health systems from 2018 to 2022 (25% vs 43%, P < .01).
More than half (53%) of practices reported at least 1 major stress in 2022, up from 34% in 2018 (P < .0001). Practice stressors were similar between 2018 and 2022 (see Table 3), but there was a significant increase in total practices reporting losing clinicians (13% vs 42%, P < .0001). Of practices with at least 1 stressor, 79% reported a loss of clinicians in 2022 (39% in 2018, P < .0001). Of practices losing clinicians, 43% had clinicians retire early, 16% fired clinicians, and 3% reported clinicians died. Of note, 1% of all practices reported 1 or more clinicians died of COVID-19. Clinician-owned practices were less likely to report that clinicians/staff were still struggling from burnout, though proportions were still high (51% clinician-owned vs 81% hospital-owned, P < .0001). Hospital-owned practices were more likely to lose a clinician than clinician-owned practices (84% vs 71%, P < .001). However, clinician-owned had fewer patient support services such as care coordinators, health plan coordinators, and registries (P < .001) and reported more difficulty finding access to mental health services (64% vs 50%, P < .001).
Access to mental health services was a major challenge. More than half (56%) reported difficulty finding and referring patients to mental health counseling. When asked how they most commonly care for patients’ mental health needs, 17% regularly spoke to mental health clinicians about their patients and 18% had mental health clinicians in their office. After Medicaid expansion, 5% of practices hired mental health clinicians to improve care for patients with Medicaid, yet 21% reported they would hire more mental health clinicians if they received a net increase in payment from Medicaid through an alternative payment model. Seventy-nine percent reported that if they had better access to mental health clinicians, they would be more likely to see Medicaid patients.
Qualitative Interviews
Burnout was a key theme identified during interviews. In particular, practices highlighted burnout related to electronic health records (EHRs) and moral injury, staffing shortages, and loss of ownership/decision making control (see Table 4). Practices also highlighted concerns about a lack of investment in primary care.
EHR and Moral Injury
Practices consistently reported high levels of stress, burnout, and moral injury. The EHR tended to be a major contributor to stress and navigating systems was regarded as, “the bureaucracy of the EHR.” Beyond the time and energy spent “putting things in the right places [and] checking the boxes,” use of the EHR was equated to a cash register, and clinicians found that it prevented a therapeutic clinician-patient relationship. As one clinician stated, “We do not get paid for taking care of patients. We get paid for doing [things] to them.” Another clinician, reflecting on the complicated documentation and administrative work required to work with insurance companies, described the situation as “a nightmare…instead of improving the knowledge base, instead of facilitating processes to improve the care of patients, we have this monstrous machine that has been created.”
Workforce Shortages Exacerbated by COVID-19
Practices highlighted major financial and staffing stresses exacerbated by COVID-19. Most practices reported being constantly short on clinicians and support staff. Rather than temporary fluctuations in staffing levels, the shortage appeared chronic, with early retirement and turnover at unprecedented levels. As one respondent described, practices are losing the ability to manage and recuperate from these losses - “I am always struggling because I do not ever have enough people. It seems like as soon as I get just enough people to get by, someone’s leaving.” Others reported turnover at more than triple the rate seen in previous years, with some clinics operating down “3 to 4 to 5 nurses at all times.” As 1 practice manager described, this creates intolerable levels of emotional and financial stress for practices. “That to me is the biggest thing that I am burnt out on. It is constant reonboarding, retraining, redoing the whole thing. It is costing us thousands of dollars a person.”
Ownership
Practice ownership transitions were frequently cited, consistent with quantitative findings, and linked to stress in primary care. Practice ownership influenced the types of burnout experienced and many practices in health systems saw underprioritization of primary care compared with specialties, whereas privately-owned practices struggled with the administrative burden of working with payers. Importantly, several practices stated that health system ownership allowed them to financially weather the impact of COVID-19, but there were consistent concerns regarding unsupported mandates and a lack of support for primary care. For example, promises from health systems about integrating new EHR functions to increase the quality and efficiency of primary care conflict with the gross lack of informaticists to configure the EHR to make full use of potential functionalities. Among independent practices, there were reports of an inability to compete against health systems when recruiting new clinicians. In addition, we heard that health systems approach independent practices for purchase and promise to provide robust support for primary care in the form of additional staffing, resources, or space, but fail to follow through on these agreements and fail to pass along any increased payments procured through a practice’s participation in pay-for-performance programs or cost saving initiatives. “The hospital systems that I am employed by, they get the money, and it does not go to primary care,” as one clinician described. Finally, practices perceived health systems as largely indifferent to the systemic factors affecting clinician burnout. While focused on promoting greater “resilience” among beleaguered practices, one clinician explained that such priorities are incongruous with practices’ real needs. “What I cannot get them to understand is how to better take care of people in the office. And to do that you have got to fix the system too.”
As a result of large-scale health system buy-outs, independent practices also struggle to recruit new clinicians to replace those taking early retirement. “We of course try to sell [new recruits] on the other features; more personal care, more personal involvement, maybe even the possibility of buying into the practices and owner as a partner” 1 clinician reported. However, health systems retain an intractable competitive advantage. “There's a lot more that can be offered in some regards, but not financially.”
Insufficient Primary Care Reimbursement
Practices emphasized that chronic underinvestment in primary care has a deleterious effect on their ability to care for medically underserved populations, like people with Medicaid. One clinician reported “It is like death by a thousand cuts. You are basically just trying to survive.” Many clinicians had concerns about low Medicaid reimbursement, especially after the influx of patients with Medicaid after the expansion in Virginia in 2019. One clinician reported, “I feel like payment for Medicaid has not kept up with the expenses of a practice.” Although practices owned by larger health systems seemed relatively better equipped to care for this population, independent practices, in particular those in rural areas, reported an inability to manage the increased costs in time and resources. One clinician explained, “I could say I will not see you again because I do not accept Medicaid anymore. And if that happens, the majority of my patients will probably not be seen anywhere else except in the emergency department.”
Discussion
Primary care practices struggle with the financial and staffing stresses accrued during the high acuity COVID-19 surges in 2020 and 2021. In 2018, primary care anticipated that the stresses they experienced would “be better the next year.” They projected having sufficient capacity to care for more Virginians after Medicaid Expansion in 2019.4 In 2022, primary care practices had 2 lived experiences: while they expanded services to Medicaid beneficiaries and maintained their quality of services, they struggled to make ends meet with significant financial burdens exacerbated by the pandemic.7,18⇓⇓–21 The COVID-19 pandemic added multiple stressors to primary care – patients were sicker and had more mental health needs, practices struggled to maintain staff, and payment for care remained low compared with payment to specialists, hospitals, and the pharmaceutical industry.22⇓–24 Practices managed to provide a broad spectrum of care and continued to care for the diverse population of Virginia, yet their clinicians left practice and burnout remained rampant.
The vast majority of primary care clinicians and staff suffered from burnout or mental exhaustion, found both in quantitative and qualitative analysis. However, the qualitative themes provided a more complex picture than the quantitative results can describe. Practice surveys indicated similar stressors reported between 2018 and 2022, yet a notable increase in the percentage of practices in 2022 that reported losing clinicians (including unexpected loss due to early retirement, fired, or death). Qualitative interviews highlighted moral injury and burnout related to administrative burden and EHR demands, dire workforce shortages, as well as significant challenges accompanying changes in ownership and low reimbursement for primary care services. Practices underscored how impactful the loss of clinicians’ time and clinicians themselves had on care and the morale of the practice at large. The complexities faced by primary care practices undermine their mission to provide person-centered, comprehensive, and continuous care. The loss of team members and consolidation of primary care by health systems will likely have an impact on workforce shortages for years to come. It is possible, though, that using all team members to the highest level of licensure would alleviate some of the increased burden of sicker patients caused by the COVID-19 pandemic and increased Medicaid enrollment. Ensuring appropriate panel size, based on patient complexity, may play an important role in providing high-quality patient care while also reducing clinician burnout.
Primary care is known to be overburdened and under-resourced.3,25 This study provides specific insights into critical factors contributing to clinician burnout, including the COVID-19 pandemic, loss of clinicians, and changes in ownership. Our findings indicate that the widespread buy-out of independent practices by health systems seems to be quickly reshaping the landscape of primary care in Virginia. There has been a national transition of ownership to hospital-owned primary care practices, which has accelerated in recent years.26 This transition may be a result of increased stressors facing primary care practices. Yet, our survey findings indicate that hospital-owned practices had more clinicians and staff suffering from burnout or mental exhaustion compared with clinician-owned practices. Effective strategies on the part of health systems are critical to addressing primary care burnout, including investment in additional primary care clinicians and staff, retention of staff, adequate resources for the breadth of services provided by primary care, and reduction of administrative and EHR hurdles. Additional research is needed to better understand the broader impact of the transition of ownership (eg, in terms of overall capacity, workforce shortages, and clinician burnout).
Broader national investment is essential to recruit, train, and retain a robust primary care workforce. The number of practices that report losing a clinician doubled from 2018 to 2022, foreshadowing an impending workforce crisis. A 2022 primary care survey suggests that 1 in 4 primary care clinicians plan to leave the field within the next 3 years.25 Although workforce shortages are critical today, primary care access will likely become more dire and could have important implications for patients and health systems (eg, patients left without primary care, increased costs for health systems, greater burden on Emergency Departments, etc.). Recruitment strategies for primary care clinicians, including training opportunities for medical students, loan repayment options, and adequate funding for primary care residency slots, are essential. Furthermore, strategies to improve the desirability and viability of primary care as a career path are critical, including reduction of administrative burden, increased investment in primary care, reduced specialty disrespect, and facilitation of interdisciplinary teams that allow clinicians to better address patients’ mental health needs and social determinants of health. Advanced practice providers who enter into and stay in primary care settings, such as nurse practitioners, can bolster the primary care workforce, however, cannot replace the clinician workforce.27
Importantly, payment reform is critical to ensure the sustainability of primary care. Primary care plays an essential role in our health system, including providing management of acute and chronic conditions as well as preventive care.1 Primary care helps prevent morbidity and mortality and promotes equity,1 yet makes up just 5% of health care spending.28 This is notably lower than other high-income countries and is associated with decreased access to primary care compared with other high-income countries.29,30 Furthermore, Medicaid primary care spend is between 1% to 2%, which has made adapting to the influx of Medicaid beneficiaries postexpansion extremely financially challenging.31 Clinician interviews highlighted financial concerns related to patients with Medicaid, for whom practices are paid 20 to 30% less than Medicare.32,33 Any primary care payment reform must come with a true increased investment in primary care from private and public insurers. For independent practices serving rural populations, the need for payment parity is dire.
This study has limitations. Although our response rate was strong compared with most large-scale surveys, it remained modest at 23%. The geographic distribution and composition of the practices seems to be consistent with most practices across the state. However, a segment of primary care practices with a different experience of care may be excluded. Our analysis only includes self-reported scope, stressors, and changes. There are claims-level analyses that could be explored in the future to provide additional context to practice scope, volume of care, and payer distribution. Finally, the practice representation is skewed toward adult and family medicine - and does not include as large of a pediatric population. Pediatric practices should be further investigated.
The Virginia primary care physician shortage was projected to be 1600 individuals by 2030.34 In a previous analysis by this team, the growth from 2010 to 2019 was only 400 physicians, with the broadest definition of those who provide primary care.35 If left unaddressed, the shortage, stress on primary care, consolidation, and reduced scope of care will leave Virginians without a foundational element of care.
Appendix.
Notes
This article was externally peer reviewed.
This is the Ahead of Print version of the article.
Funding: Funding for this study was provided by the Department of Medical Assistance Services and the National Center for Advancing Translational Sciences (UL1TR002649).
Conflict of interest: Dr. Britz is a JABFM Fellow.
To see this article online, please go to: http://jabfm.org/content/00/00/000.full.
- Received for publication April 15, 2023.
- Revision received June 26, 2023.
- Revision received July 14, 2023.
- Accepted for publication July 24, 2023.