Impact of a Primary Care Value-Based Compensation Model

Introduction

Primary care is a foundational element in delivering high-quality, efficient population health.¹ However, the future of primary care is not certain. Primary care is threatened by high rates of burnout, fewer residency graduates choosing primary care, and experienced primary care physicians (PCPs) retiring earlier. Additionally, due to system operational forces and individual financial compensation based on volume, primary care providers are struggling.^2–5 This is compounded by primary care’s undervaluation relative to specialty care in the US.⁶ Integrated delivery systems⁷ and academic medical centers (AMCs) view primary care as a gateway into the healthcare system, rather than as a direct driver of financial value. This contributes to underinvestment in essential primary care infrastructure, such as care coordination, behavioral health integration, and resources to address social determinants of health. Consequently, PCPs face mounting workloads and insufficient support, which fuels burnout and reductions in clinical effort among PCPs.

For decades, primary care has been paid via fee-for-service (FFS) in which physicians are reimbursed for each service they provide, measured in relative value units (RVUs). FFS rewards and encourages quantity of care as opposed to quality of care, and in some cases can incentivize unnecessary procedures and visits.⁸ FFS payment models are particularly misaligned with the goals of primary care to provide comprehensive, continuous, first-contact, and coordinated care.⁹ The resources available to primary care are closely tied to the number of in-person visits delivered by the individual or group. While primary care practices in an integrated delivery system benefit from the resources of their system, the main financial and operational focus for primary care and healthcare system leaders remains the number of visits, which then translate into relative value units, and subsequently into FFS payment. The difference between the cost and the RVU-generated revenue defines the “support” for primary care.

Beginning in 2010, value-based payment models introduced via the Center for Medicaid & Medicare Services started operating in parallel with FFS payments. These models require participants to take on greater accountability for the quality and cost of care and usually require a greater investment in primary care as a key element. By paying for a group of patients’ overall care rather than individual units of care, value-based payment models have the potential to better align payment with primary care’s fundamental goals. Further, value-based models allow more flexibility for remote and inter-visit care, which is appreciated by patients and physicians alike, provided issues of health equity are addressed and not exacerbated.⁸

Added resources, such as staffing that enables PCPs to engage more frequently with their patients, and that support the management of patients with complex needs, cost money. Organizations can recover the costs of such added resources, however, because they are paid under risk-based contracts that reward the organizations for reducing spending and improving quality across their patient population. However, the impact of these models remains uneven.¹⁰ Many AMCs and integrated delivery systems remain in hybrid environments where FFS revenue continues to dominate, limiting the resources available for robust primary care investment. This dynamic is particularly challenging in AMCs, where financial margins from care delivery are often redirected to support teaching and research missions.

Brigham and Women’s Hospital (BWH) is an academic medical center in Boston and a founding member of the Mass General Brigham health system. Between 2016 and 2017, BWH faced 9%-11% attrition of primary care clinicians, with projected declines in covered lives in the following two years. Covered lives are defined as the number of patients enrolled in the BWH’s insurance contracts or attributed to its care network. In response, BWH proposed and implemented a shift from a FFS model to a value-based compensation model throughout their 16 clinics (see Table 1 for key changes). This happened in phases, which included initial changes to primary care infrastructure from 2017 to 2019, discussions with leadership from 2018 to 2020, and two years of frozen salaries due to COVID from 2020-2022. There was a phased transition from 15 to 30-minute visits that started in spring 2020. A planned change in how primary care physician compensation was determined, from RVU-driven compensation to a value-based compensation model incorporating multiple components, including a panel-size target based on the physician’s clinical FTE was communicated in 2020 (see Table 2 for compensation components). Subsequently, variable compensation changed to value-based compensation in 2021, with the full model implemented in 2022 (see Figure 1 for a timeline overlaid with panel growth rate). This mixed-methods study was designed to understand how a new value-based compensation model impacted the daily practice of PCPs.

• Download figure
• Open in new tab

Figure 1. Trends in Absolute Panel Size Across the Center of Excellence and Percentage Changes in Panel Growth Rate by Year

Panel Growth Rate is defined as ((Year End Panel – Year End Panel of Prior Year)/Year End Panel of Prior Year). This measure was used since organizational accountabilities were based on the rate of panel growth for BWH Primary Care.

Abbreviations: RVU = relative value unit

Methods

Qualitative Data Collection and Analysis

We created a semi-structured interview guide to elicit perceptions of how the value-based compensation model affected PCPs’ approaches to clinical work, identify strengths and weaknesses they perceived with the model, and reflect on whether the model had addressed their levels of joy or burnout with their clinical role (see interview guide in digital supplement). The interview protocol was created by the qualitative researcher (ES) in consultation with the clinical leaders (LR, RG), who had extensive knowledge of, and experience with, the design and implementation of the value-based compensation model.

We recruited voluntary physician interviewees after clinical leaders were informed about the study via emails and announcements at physician meetings. Eligibility criteria included attending PCPs at the primary care sites who were employed by BWH for at least 3 years. This time period enabled participants to have a basis for comparing the value-based model to the prior FFS model in a context unaffected by the COVID-19 pandemic. From among eligible physicians, participants were selected via a random-access generator and were sent an email invitation to participate in the study. Participants were also invited to participate via a division-wide email announcement.

One of the study co-authors (ES) conducted all 17 interviews one-to-one with each PCP. This co-author had extensive experience in conducting qualitative interviews in primary care settings and had no relationship with the BWH implementation of value-based compensation. Semi-structured interviews lasted an average of 20 minutes and were conducted via Zoom from June through October 2023. The timing of the interviews ensured that the value-based compensation model had been fully implemented with sufficient time for PCPs to perceive potential changes to their practice. Participants provided verbal consent before being interviewed. All interviews were recorded, transcribed verbatim, and de-identified. Participants received no compensation for their participation. Interviews were conducted until data saturation occurred.

Two authors independently reviewed a sub-set of (5) transcripts to create an initial codebook. Codes were identified based on the interview guide as well as through inductive coding. The lead analyst (LU) then coded all 17 interviews; another analyst (ES) cross-checked the coding to ensure accuracy.¹² Given some overlapping constructs in the data, the analysts engaged in discussions to reach a consensus about final codes during regular team meetings.¹³ Major themes were identified by consensus within the analysis team and then presented to other co-authors for validation.

This study was approved by Mass General Brigham IRB (study number 2023P000464).

Results

Discussion

Primary care is critically important for improving both patient outcomes and addressing rising healthcare spending,^14–16 yet challenges remain for both patients and physicians. This study explored PCPs’ perceptions of shifting to a value-based compensation model in an AMC. PCPs appreciated moving away from FFS and the recognition of inter-visit work, but they also described the strain of larger panels, persistent access barriers for their patients, the growing burden of patient-portal demands, and worsening system-wide issues such as clinic understaffing and longer wait times for specialty care. These concerns mirror national trends of workforce shortages¹⁷ and appointment delays,¹⁸ as well as prior evidence linking electronic health record demands to frustration,¹⁹ reduced job satisfaction,²⁰ and burnout.^21,22

Our findings illustrate a shift in primary care toward asynchronous, inter-visit work, much of it managed through electronic health records (EHRs) and rewarded by value-based models.²³ This has altered physician-patient engagement, increasing time spent on messaging, remote monitoring, and care coordination outside of visits. While EHRs enable this work, they also contribute to administrative burden and potential burnout.²² The American Medical Association has noted generational differences in burnout drivers, with electronic documentation being a top concern for many baby boomers (born 1946-1964) but less so for millennials (born 1981-1999).²⁴ Physicians who entered practice more recently often prioritize work-life balance and flexibility,²⁵ which align with the structured workflows of value-based care. In contrast, those who trained in systems centered around in-person visits may find the shift more challenging.²⁵ These generational differences are shaped not only by age but by formative experiences with technology, cultural expectations, and workplace norms. Younger clinicians, many of whom are digital natives,²⁶ tend to embrace new communication tools and remote patient monitoring, and they expect supportive, flexible work environments. By contrast, more experienced physicians may be accustomed to hierarchical team structures and in-person interactions, making adaptation to new models of care and documentation practices more difficult.²⁷

These evolving dynamics influence both physician satisfaction and the effectiveness of primary care delivery in a rapidly changing healthcare landscape. While major healthcare systems do participate in value-based contracts, revenue from these contracts represent a small percentage of overall system revenue^28–30 and is unlikely to prompt the development of robust infrastructure that is necessary to support value-based practices. Our findings emphasize that health systems must make appropriate financial investments in the structural supports required for value-based compensation models if these models are to improve patient care and clinician wellness. As our study demonstrates, efforts to provide higher-quality care aligned with clinicians’ values can be hindered by structural gaps that limit effective implementation. Physicians noted decreased patient access to visits, ongoing staff shortages, and long specialty care wait times, indicating that a value-based approach and panel expansion in primary care will have limited impact if broader system incentives remain misaligned with primary care. These challenges reflect national post-pandemic trends,³¹ which are especially pronounced in smaller cities with multiple AMCs, where competitive job markets intensify staffing shortages. Our study complements other reports of reorienting primary care compensation in an academic medical center context away from a pure FFS focus.^32,33

Our study has both strengths and limitations. One clear limitation is the presence of confounders identified by respondents, including the COVID-19 pandemic, some major health-system changes that impacted all of BWH, and structural forces related to the organization and delivery of primary care (eg, staff shortages). This made it difficult for PCPs to attribute their perceptions about their practice exclusively to the shift in the compensation model. Further, we attribute the decline in quality metrics to COVID-19;³⁴ however, given the concurrent changes, it is difficult to isolate the specific impact of the compensation model from these other influences with complete certainty. Additional limitations include a relatively small sample size within one health system, self-selection bias, and recall bias, as the implementation of the model had stretched over the prior 2 years. We did not calculate coder agreement, instead relying on verifying accuracy of coding in a sub-sample and a consensus-based approach to identifying themes. Strengths include a rigorous qualitative analysis of interview data as well as evaluations of this model from the physicians’ perspective. To our knowledge, this is the first report describing PCP perspectives on the implementation of such a shift at an academic medical center.