Abstract
Corporate employment of primary care physicians by hospitals and private equity-backed organizations has increased in recent years, raising concerns that the financial incentives of corporate employment will incentivize provision of low-value care. In this study, we use a difference-in-differences design to examine whether hospital and private equity (PE) acquisitions of primary care practices between 2018 and 2021 change the provision of low-value services within the Traditional Medicare population compared with independent primary care practices. Overall, one-third of Medicare beneficiaries with PE-affiliated primary care providers (PCPs) received low-value services relative to one-quarter of beneficiaries with hospital-affiliated or independent PCPs. We find that hospital and PE acquisitions in primary care did not result in significant changes to the likelihood or number of low-value services received by Medicare patients. As consolidation re-shapes the U.S. health care system, ongoing evaluation of hospitals, health systems and PE-backed organizations remains critical.
Introduction
The primary care landscape in the United States has experienced significant structural changes over the past decade. As of 2024, over half of primary care providers (PCPs) were affiliated with hospitals, health systems and private equity (PE)-backed organizations.(GAO-25-107450, 2025; Kane, n.d.; Physicians Advocacy Institute, Avalere Health, 2024; Singh, Radhakrishnan, et al., 2025) This wave of consolidation has been driven by promises of greater efficiency, care coordination, and enhanced bargaining leverage in negotiations with commercial payers. However, empirical evidence has raised concerns about rising costs, altered referral and utilization patterns, and limited improvements in patient outcomes (Borsa et al., 2023). These trends have renewed scrutiny of consolidation and its impact on the value of care delivered, particularly in the context of pervasive wasteful and “low-value” spending in the U.S. health care system. (Berwick & Hackbarth, 2012; Boudreau et al., 2022; Ganguli et al., 2021; Schwartz et al., 2014, 2019; Segal et al., 2022) Particularly in the context of fee-for-service reimbursement, one potential concern is that the financial incentives of corporate employment will incentivize increased use of low-value care.
While many forms of for-profit ownership exist within health care, the rise of PE has defined a newer wave of health care corporatization in the United States (Fuse Brown, 2025; Singh, 2025). PE firms have invested over U.S.$1 trillion in U.S. hospitals (Appelbaum & Batt, 2020; Bruch et al., 2020; Cerullo et al., 2022; Kannan et al., 2023, 2025; Offodile et al., 2021), long-term care settings (Braun et al., 2020; Braun et al., 2021; Bunker et al., 2025; Gandhi et al., 2020; Gupta et al., 2024), behavioral health clinics (Arnold et al., 2026; B. Brown et al., 2020; Singh, Cantor, et al., 2024; Singh et al., 2025), and physician practices across specialties (Abdelhadi et al., 2024; Singh, 2025; Singh et al., 2024; Singh et al., 2022), including dermatology (Braun et al., 2021; Singh et al., 2022), ophthalmology (Braun et al., 2024; Singh et al., 2024; Singh et al., 2025), gastroenterology (Arnold et al., 2026; Singh et al., 2025), radiology (Khunte et al., 2025; Khunte & Singh, 2025), cardiology (Philips et al., 2025; Sardar et al., 2024; Singh et al., 2024), and primary care (Singh et al., 2025) Previous research has found PE investments in health care increase health care prices and utilization (Borsa et al., 2023), with changes to health care staffing (Berquist et al., 2025; Bruch et al., 2023; Kannan et al., 2025; Singh et al., 2025) that has resulted in adverse patient outcome in some settings (Forgia & McDevitt, 2025). These concerns have prompted multiple federal and state investigations (Cai & Song, 2023), and policy debates focused on limiting risky financial strategies (Bruch et al., 2025), strengthening antitrust enforcement (FTC Challenges Private Equity Firm’s Scheme to Suppress Competition in Anesthesiology Practices Across Texas, 2023), and transparency requirements (E. C. F. Brown et al., 2025; Singh & Brown, n.d.).
In general, hospital acquisitions of physician practices reflect the vertical integration of complementary services (Whaley et al., 2021; Whaley & Zhao, 2024), while PE acquisitions of physician practices can increase horizontal consolidation within specialties or geographic areas via “platform and add-on” acquisitions (Asil et al., 2024; Singh et al., 2024; Singh & Whaley, 2025). A number of studies have documented that both types of consolidation are associated with increased prices paid by commercial insurers for physician services (Asil et al., 2024; Baker et al., 2014; Borsa et al., 2023; Braun et al., 2021; Capps et al., 2018; Forgia et al., 2022; Neprash et al., 2015; Scheffler et al., 2023; Singh et al., 2022; Whaley et al., 2021). In the case of PE, studies have found increases in patient encounters, visit frequency, and higher-intensity billing codes, such as longer office visits, raising questions about financial incentives and clinical necessity (Dixit et al., 2026; Singh et al., 2024; Singh et al., 2026; Singh et al., 2022). Similarly, studies on both, PE acquisition and vertical integration of hospitals and physicians have found higher spending on diagnostic imaging and laboratory tests (Singh et al. 2026; Whaley et al., 2021). While these studies suggest changes in utilization patterns, they generally do not distinguish between care that is clinically appropriate and care that may be of low value.
Low-value care—defined as services that provide little to no clinical benefit to patients, are unlikely to improve outcomes, or may even cause harm in certain clinical contexts—has been a long-standing concern in health policy and clinical practice, contributing to about U.S.$200 billion annually to total U.S. health care spending (Berwick & Hackbarth, 2012). Often described as “overuse” or “inappropriate care,” low-value care can include treatments that are not clinically indicated (e.g., prescribing antibiotics for viral infections), inappropriate for specific patient groups (such as cardiac stress testing in low-risk individuals), or delivered more frequently than necessary (such as unwarranted follow-up visits) (Schwartz et al., 2014). National initiatives such as the Choosing Wisely campaign, launched by the American Board of Internal Medicine Foundation in collaboration with specialty societies, and recommendations from the U.S. Preventive Services Task Force, have sought to identify and discourage the use of such services by providing evidence-based guidelines to clinicians and patients. Despite these efforts, low-value care remains pervasive across the U.S. health care system, contributing to unnecessary spending, patient burden, and potential harm (Boudreau et al., 2022; Colla et al., 2015; Ganguli et al., 2019; Segal et al., 2022).
Both primary and specialty care providers play critical roles in the delivery of low-value care (Baum et al., 2021). PCPs, in particular, serve as the first point of contact in the health system and often act as gatekeepers for referrals, diagnostic testing, and ongoing care decisions (Baum et al., 2021). As such, PCPs have considerable influence over downstream utilization, including the initiation or avoidance of low-value services (Baum et al., 2021; Ganguli et al., 2021). Conversely, specialists are often the direct providers of many procedures and diagnostic services flagged as low value, particularly in the domains of imaging, cardiovascular care, and musculoskeletal interventions (Schwartz et al., 2019). The interplay between primary and specialty care can further complicate efforts to reduce low-value care, as referral patterns, care fragmentation, and practice incentives all contribute to variation in utilization.
The objective of this study is to address this research gap by examining whether acquisitions of primary care practices by hospitals and PE firms are associated with changes in the utilization of low-value services within the Medicare population. While existing studies have examined the prevalence of low-value care in different patient populations or predictors of low-value service use, little is known about how ownership changes, particularly through consolidation, influence the delivery of such services (Ganguli et al., 2021; Mafi et al., 2017; Mafi et al., 2017; Reid et al., 2016; Schpero et al., 2017; Schwartz et al., 2019; Segal et al., 2022). Using administrative claims data from Traditional Medicare and a difference-in-differences (DiD) design, we examine whether hospital and PE acquisitions in primary care change the provision of low-value services relative to independent PCPs.
New Contributions
This study makes several contributions to the growing literature on health care consolidation. First, it extends existing research on hospital and PE acquisitions by shifting the focus from prices and utilization volume to the clinical value of care provided. Prior studies have documented that both hospital and PE consolidation increase spending and service intensity, yet few have distinguished between appropriate and low-value utilization. By directly examining changes in the use of low-value services following acquisitions relative to independent practices, this study provides novel evidence on whether consolidation enhances or erodes the value of care. Second, this work contributes to the growing literature on PE in health care by examining its effects, and the effects of hospital acquisitions within the same analytic framework. By focusing on low-value care, a persistent source of wasteful spending, we aim to inform ongoing debates about the implications of hospital and PE ownership in primary care and the delivery of high-value, patient-centered care.
Conceptual Basis for the Study
The conceptual model informing our research hypothesis is adapted from Landon et al.’s model of the impact of health care organizations on the quality of care, which posits that organizational structures and ownership arrangements influence patient care through their effects on physician incentives, resources, and the care delivery environment (Landon et al., 1998). Adapting this framework, our model conceptualizes hospital and PE consolidation in primary care as organizational changes that reshape practice-level financial incentives, managerial oversight, and infrastructure for care coordination. These organizational shifts can, in turn, alter physician decision-making and clinical practice patterns, potentially affecting the provision of low-value care. Specifically, hospital and PE consolidation may reduce low-value care by enabling greater investment in data systems and adherence to evidence-based guidelines, or conversely, increase low-value care if financial pressures or productivity incentives promote higher service volume with unclear patient benefit.
Our conceptual framework is further informed by recent work from Dennis Scanlon and colleagues who highlight the role of care delivery redesign—innovations in the organization and delivery of care that may improve production efficiency and quality—as a key mechanism through which mergers and acquisitions exert downstream effects on care processes and outcomes (Scanlon et al., 2025). This lens is particularly relevant to PE acquisition of primary care practices, where ownership changes may be motivated by operational restructuring and efficiency gains rather than market concentration alone, and where the expected impact on care delivery therefore depends on whether ownership translates into meaningful redesign of how care is organized and delivered.
We hypothesize that physician characteristics (e.g., age and gender), physician specialty (e.g., primary care), attributed patient characteristics (e.g., health status), and practice characteristics (e.g., PE or hospital affiliation) can influence patient care patterns. Specifically, given the relative profitability and potential for over use, following acquisition, PCPs in hospital- and PE-affiliated practices can increase the provision of low-value care provision among Traditional Medicare beneficiaries.
Method
There were multiple steps involved in creating the analytic sample. First, we identified PE acquisitions of primary care physician practices from 2016 to 2022; second, using the Medicare Data on Provider Practice and Specialty (MD-PPAS) and the Compendium of U.S. Health Systems made available by the Agency for Healthcare Research and Quality, we identified physicians associated with hospital and PE acquisition; third, we assigned patients to primary care physicians using 20% Traditional Medicare claims; and fourth, we linked this information with outcome measures constructed using the Medicare claims data.
Identifying Private Equity Acquisitions
We followed previous studies and used multiple data sources to identify physicians in PE-acquired practices. PE acquisitions from 2016 to 2022 were identified using proprietary PitchBook Inc. data, used by other studies to examine PE in health care (Gupta et al., 2024; Kannan et al., 2023; Singh et al., 2026). Acquisitions prior to 2016 reflect less than 0.5% of primary care physicians and were excluded from this analysis. In addition, we excluded acquisitions in 2017 as pre-acquisition outcomes for this acquisition cohort would require triangulation between International Classification of Diseases (ICD)-9 and ICD-10 transitions that occurred through 2016. Because Medicare claims data were only available through 2022, we similarly excluded acquisitions occurring in 2022, as these physicians would have no full post-acquisition year of outcome data available. Thus, the universe of acquisitions included in our final sample spanned 2018 to 2021. Each acquisition was manually verified and the list was expanded using press releases, industry reports, and current and archived physician practice websites.
To categorize physicians affiliated with PE-acquired practices, we identified physician owners of practice using web searches for their national provider identifiers (NPIs), following validated methods (Braun et al., 2021; Braun et al., 2024; Singh et al., 2025; Singh et al., 2025). Acquisitions were linked to taxpayer identification numbers (TINs) by matching the owner’s NPI, business name, and location to MD-PPAS from 2018 to 2021, which includes all providers who billed Medicare and are registered in the Provider, Enrollment, Chain, and Ownership System. NPIs were tracked over time for each owner’s practice.
Identifying Hospital Acquisitions
To identify hospital-affiliated primary care physicians, we followed an approach from previous studies that relies on the MD-PPAS files, the Medicare Provider Enrollment, Chain, and Ownership System (Whaley et al., 2021). We validated our classification of hospital-affiliated physicians using the Compendium of U.S. Health Systems data made available by the Agency for Healthcare Research and Quality (AHRQ) (Contreary et al., 2023). We validated hospital affiliation status to physicians by linking AHRQ Compendium of U.S Health System Group Practice Linkage file, which linked group practices to health systems, with MD-PPAS data by TIN name. Physicians were classified as hospital-affiliated if their primary TIN matched a non-missing health system identifier. The Group Practice Linkage Files are only available for 2016, 2018, 2020, 2021, and 2022; thus, there may be some measurement error in identifying hospital-affiliated PCPs.
Medicare Claims Data
Using 20% Traditional Medicare Part B claims from 2017 to 2022, we identified all claims billed by PCPs (physician specialty codes: 01, 08, 11, 38) in an office setting (place of service code 11) in each year from 2017 to 2022. To focus on primary care, only practices where at least half of physicians had specialties 01, 08, 11, or 38 were included.
To assign accountability for patient low-value care use, we assigned each patient to one primary care physician in a year using an attribution algorithm used previously in studies of Medicare beneficiaries. Patients were assigned to PCPs within a year by the amount of distinct evaluation and management claims (Whaley & Zhao, 2024). Successive tiebreakers included the greater sum of Medicare allowable charge on claims billed within the year and whichever physician billed for the follow-up Medicare annual wellness visit. This method assigned all but 1.67% of more than 55 million patient-NPI combinations, which we excluded from the final analyses.
Study Outcomes
Our primary outcomes are defined at the patient-year level as the probability of receiving any low-value service and the total number of low-value services received. We operationalized 26 claims-based, low-value care definitions described in prior research based on recommendations from the Choosing Wisely campaign and professional medical societies (Schwartz et al., 2014). For each measure, we identified the beneficiaries eligible (i.e., at risk) for the potentially low-value service. Because our outcomes are constructed at the patient-year level and our DiD design requires at least one full pre-period year and one full post-period year, our analytic sample is restricted to PCPs acquired between 2018 and 2021, observed over the full study window of 2017 to 2022. This yields a minimum of one pre-acquisition year (for those acquired in 2018) and one post-acquisition year (for those acquired in 2021).
Statistical Analysis
We used DiD regressions to measure the association between PE acquisition and hospital acquisition and outcomes of interest. Since DiD analyses compare changes in the treatment group with changes in the comparison group, removing the impacts of market-level factors that affect both groups. We used Sun and Abraham estimator to avoid bias that can occur with staggered treatment timing (Goodman-Bacon, 2021; Sun & Abraham, 2021).
We defined two distinct “treatment” groups: patients whose primary care physician became acquired by a PE-affiliated practice, and patients whose primary care physician became acquired by a hospital or health system. Both treatment groups were compared with the same control group, i.e., patients whose primary care physicians did not experience a PE- or hospital-affiliated change in ownership for the duration of the study period.
We compared pre-acquisition patient characteristics by examining parallel trends in pre-acquisition group differences to detect any imbalance in observable characteristics prior to acquisition. In the event study analyses, event time 0 denoted the year of acquisition. We used data from 4 years before acquisition (event time −4, . . ., −1 year[s]) through 4 years after (event time +1, +2, . . ., +4 years[s]), with the year of acquisition as the reference period. We tested for differences in pre-acquisition trends between acquired practices and the control group by performing joint F tests of the hypothesis and that pre-acquisition interactions between the treatment and time indicators were no different.
The DiD regressions for outcomes at the patient-year level included physician, patient, and time fixed effects. Thus, the DiD regressions estimate the within-physician and within-beneficiary change in outcomes following acquisition. The inclusion of fixed-effects accounts for time-invariant attributes of physicians (e.g., gender) and beneficiaries (e.g., baseline presence of chronic conditions) that are associated with outcomes of interest. Given this, we did not include further covariates in our regression model to account for additional fixed characteristics of physicians or beneficiaries. Standard errors were clustered according to each physician.
Sensitivity Analysis
In sensitivity analysis, we used non-linear Poisson specifications for count outcomes. In addition, we separately examined low-value service that were typically ordered by primary care physicians (PCP-driven) or specialists (specialist-driven), following methodology outlined in previous research (Boudreau et al., 2022; Schwartz et al., 2015, 2019).
We also estimated DiD event study regressions using a matched sample, where patients with PE-acquired and hospital-acquired PCPs were matched to control patients. Matched controls were identified using 1:5 matching without replacement that required an exact match on patient gender, race, and state of residence, and within one standard deviation match on HCC scores and age. Finally, we estimated stacked DiD with stacked cohort fixed effects (Wing et al., 2024).
Results
The analytical sample included 535,281 patients attributed to hospital-affiliated PCPs, 25,362 patients attributed to PE-affiliated PCPs and 2,023,380 control Medicare beneficiaries attributed to independent PCPs. Across all years, the mean age was 76.8 years for patients with hospital-affiliated PCPs (SD: 6.7 years), 76.6 years for patients with PE-affiliated PCPs (SD: 6.7 years), and 76.2 years for patients with control PCPs (SD: 7.1 years) (Table 1). Across all years, the racial/ethnic composition of the sample was 4.1% Black, 91.9% White, 2.5% Other, 1.5% Unknown for patients with hospital-affiliated PCPs; 2.9% Black, 93.7% White, 1.9% Other, 1.5% Unknown for patients with PE-affiliated PCPs; and 5.3% Black, 89.5% White, 3.6% Other, 1.6% Unknown for patients with control PCPs.
Characteristics of the Study Population, by Hospital and PE Affiliation Status of Primary Care Physicians, 2017 to 2022.
Note. “Hospital-affiliated” and “PE-affiliated” represent patients whose attributed PCPs become acquired by health systems or PE-affiliated entities over the study period. “Independent” refers to patients whose attributed PCPs were never affiliated with hospital- or PE-affiliated primary care practices.
SMD reflects the Standardized Mean Difference comparing hospital-affiliated PCPs to independent PCPs, and PE-affiliated PCPs to independent PCPs. We considered SMDs greater than 0.10 to reflect meaningful differences between groups.
Figures 1 and

Percentage of Beneficiaries Receiving Any Low-Value Service, by PCP Affiliation.

Number of Low-Value Services Per Beneficiary, by PCP Affiliation.
Changes in Likelihood and Number of Low-Value Care Services
After hospital acquisition, the likelihood of receiving any low-value service changed by a modest and statistically insignificant amount, –0.00057 (95% confidence interval [CI]: –0.00163, 0.00048] or −0.21% per patient-year relative to patients in non-acquired practices (Figure 3A and Table 2). The number of low-value services also declined by a modest and statistically non-significant, 0.00116 (95% CI: –0.00285, 0.00052) services per patient-year or 0.45% after hospital acquisition, but this decrease was not significant at p = .05 (Figure 4A

Differential Change in Likelihood of Low-Value Service.(A) Hospital Acquisition. (B) PE Acquisition.
Differential Change in Outcomes of Interest, Hospital and PE Acquisition Relative to Controls, 2017 to 2022.
Note/Source. DiD regression coefficients are estimated using a difference-in-differences regression that compares patients in hospital- and PE-acquired practices to patients in non-acquired practices. All regressions include patient, physician, and time fixed effects. Standard errors are clustered at the level of the physician.

Differential Change in Number of Low-Value Services. (A) Hospital Acquisition. (B) PE Acquisition.
After PE acquisition, there were similar changes in the likelihood of receiving any low-value, –0.00056 (95% CI: –0.00172, 0.00058) or −0.22% per patient-year relative to patients in non-acquired practices (Figure 3B and Table 2). The number of low-value services changed by 0.00038 (95% CI: –0.00147, 0.00224) or 0.10% per patient-year but this change was not significant at p = .05 (Figure 4B and Table 2).
Sensitivity Analyses
We used a non-linear Poisson specification as a robustness check for count outcomes. This statistical analysis confirmed the findings of the stacked DiD; there were no statistically significant differences between PE patients and matched controls, or hospital patients and matched controls (Appendix Figure 1). A sensitivity test that separately examined services initiated by primary care versus specialist physicians also found small and not statistically significant changes in outcomes (Appendix Figures 2–10). Our results were sensitive to using matched controls (Appendix Figures 6 and 7) and to using a stacked DiD design with cohort fixed effects (Appendix Figures 8 and 9).
Discussion
In this national study of Medicare beneficiaries, we examined how consolidation in primary care, specifically through acquisition by hospitals and PE firms, affects the provision of low-value care. In descriptive analysis, we found that patients attributed to PE-affiliated PCPs consistently received more low-value services than those seen by hospital-affiliated or independent PCPs, suggesting underlying and important differences in care delivery models, incentive structures, or practice patterns across ownership types. In DiD analysis, we found no statistically significant changes in the likelihood or volume of low-value services received by patients with hospital- or PE-affiliated PCPs relative to independent PCPs.
In 2022, nearly one-third of patients attributed to PE-affiliated PCPs received a low-value service, substantially higher than one-quarter of beneficiaries receiving low-value care among hospital-affiliated and independent practices. Higher baseline rates of low-value care among PE-affiliated practices can reflect selection in the practices that are acquired by PE firms. For example, PE investors may preferentially target practices with higher patient volume, reduced oversight of clinical decision-making, or billing practices that emphasize service intensity. As such, higher rates of low-value care observed in PE-affiliated practices could partly reflect pre-existing practice patterns, rather than direct effects of PE acquisition or changes newly instituted by PE ownership. Higher low-value service utilization rates among patients attributed to PE-affiliated PCPs may also be explained, at least in part, to differences in referral practices and age and case mix of these cohorts.
We build on prior literature that has examined increased utilization intensity and spending following PE and hospital acquisitions. (Braun, Bond, et al., 2021; Bruch et al., 2020; Forgia et al., 2022; Richards & Whaley, 2024; Singh, Aderman, et al., 2024) While earlier studies have raised concerns about potential overuse (Singh, Song, et al., 2022; Whaley et al., 2021), for example, through increased frequency of visits or lab tests (Singh, Dixit, et al., 2026), to the best of our knowledge no prior studies have differentiated between clinically appropriate and low-value care in the context of emerging waves of health care consolidation. Previous research has found hospital-affiliated primary care practices to be associated with greater use of low-value imaging services. (Mafi, Wee, et al., 2017) Extending this research, we found that hospital and PE acquisitions did not result in statistically significant changes to the probability or number of low-value services received by patients, suggesting that ownership changes alone may not lead to significant shifts in the provision of low-value services. Importantly, our results are consistent even after accounting for patient-level fixed effects, minimizing potential confounding from time-invariant patient attributes (e.g., chronic conditions) that can influence care provision.
Previous research has found both primary and specialty care providers play critical roles in the delivery of low-value care. (Baum et al., 2021) While certain low-value services in our composite measure are performed by specialists rather than ordered directly by primary care physicians, we intentionally retained these services in our primary outcome. Evidence suggests that PCPs in vertically integrated settings exert meaningful influence over downstream specialty utilization through referral patterns, and that PE ownership may incentivize more liberal referrals to affiliated specialists. (Singh, 2023; U.S. Attorney’s Office, District of Rhode Island, 2023; Whaley et al., 2021) Restricting our outcome exclusively to PCP-ordered services would therefore omit a clinically plausible and policy-relevant mechanism through which hospital and PE ownership could affect low-value care delivery. Nonetheless, to address concerns about PCP-driven services, we separately examined PCP-driven and specialist-driven measures and found that primary care acquisitions did not change either PCP-driven nor specialist-driven measures of low-value care.
The null findings across all specifications warrant consideration of several plausible mechanisms. First, the effect of hospital- and PE ownership on low-value care may be too distal to detect in the absence of more tightly controlled referral networks. For example, corporate ownership in primary care may not reliably translate into changes in referral behavior unless ownership structures more directly coordinate or constrain referral patterns. Second, competing incentives may attenuate any ownership-driven pressure toward higher utilization. PCPs operating under value-based payment arrangements, for instance, or those seeing a larger share of Medicare Advantage patients might face countervailing incentives to reduce low-value care (Boudreau et al., 2022; Duggan et al., 2024; Modi et al., 2019; Schwartz et al., 2015). It is also possible that acquisition-driven changes take longer to materialize than our post-acquisition observation window allowed. Moreover, while acquisitions by hospitals and PE firms may shape organizational incentives, individual clinicians might retain clinical autonomy, resisting or adapting to operational incentives in ways that buffer patients from changes in care value.
Together, these considerations suggest that ownership structure within primary care alone may be insufficient to shift low-value care provision. More broadly, our findings are consistent with previous research that suggests that observable physician and practice characteristics account for only a modest share of variation in low-value service use (Barreto et al., 2019; S. N. Landon et al., 2022; Schwartz et al., 2019). Rather, drivers of low-value care may be more deeply embedded in local practice norms, patient preferences, and clinical culture than in the structural features of how practices are organized or owned. Future research should examine more proximal determinants of low-value care delivery, including the role of referral network design, payment model incentives, and organizational changes in staffing and workflows that may follow acquisitions but are not captured by ownership status alone.
Limitations
This study has several limitations. First, we rely on Medicare claims-based measures of low-value care based on previously published and validated claims-based algorithms (Boudreau et al., 2022; Schwartz et al., 2014, 2019). Nonetheless, our estimates of low-value care may differ from the true rates. Second, given the lack of systematic reporting or disclosure requirements for PE acquisitions of physician practices, it is possible we underestimate PE’s true penetration into primary care and undercount PE-affiliated primary care physicians. Third, we were not able to separately examine care provided by advanced practitioners, who make up a growing part of the primary care workforce and bill under physician NPIs. Fourth, our matched controls may include patients who see physicians from independent practices, payer-affiliated practices, and other corporate entities that may not be a homogeneous comparison group. Relatedly, to the extent physician groups have informal contractual affiliations with hospitals that are not documented in the MD-PPAS or AHRQ data, we would not be able to capture hospital-affiliation for these groups accurately. Fifth, some beneficiaries may have been misattributed to a PCP, and results may not be generalizable to unattributed beneficiaries or other populations. Sixth, our study analyzes data from the Traditional Medicare program, which has limited utilization management tools unlike those deployed by Medicare Advantage or commercial insurers to limit low-value care. As such, our findings may have limited generalizability to other payers or patient populations. Finally, we are unable to track primary care participation in Accountable Care Organizations and similar arrangements that might influence the provision of low-value care.
Conclusion
Hospital and PE acquisitions in primary care did not result in a significant change in low-value care delivery in the short term. At the same time, the higher unadjusted rates of low-value care among patients seen by PE-affiliated primary care practices highlight a need for closer monitoring of care quality and value in these settings. As PE and hospital systems continue to expand their footprint in primary care, ongoing evaluation of their impact on care value, patient outcomes, and health care spending remain critical.
Supplemental Material
sj-docx-1-mcr-10.1177_10775587261453724 – Supplemental material for Low-Value Care Following Hospital and Private Equity Acquisition in Primary Care
Supplemental material, sj-docx-1-mcr-10.1177_10775587261453724 for Low-Value Care Following Hospital and Private Equity Acquisition in Primary Care by Yashaswini Singh, Qi Lei, Meehir Dixit and Christopher M. Whaley in Medical Care Research and Review
Footnotes
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by grants from Arnold Ventures and the Commonwealth Fund.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
The data that support the findings of this study are proprietary and not publicly available.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
