1.1. Lower quality, higher cost: how did we get here?

In the United States (US) healthcare system, multiple realities exist. The US is a hotbed of innovation, in many ways subsidising healthcare advances for the rest of the world and providing best-in-class care to those with the means to access it. On the other hand, it delivers healthcare at a significantly higher cost than any of its peer nations without the reciprocal improvements in quality to be expected on a national level (Blumenthal et al., Reference Blumenthal, Gumas, Shah, Gunja and Williams2024).

As a country and within the healthcare industry, we talk about this dichotomy ad nauseum, but stakeholders remain divided on the primary culprit. Between insurance companies, pharmaceutical manufacturers, hospital systems, and the innumerable entities that operate in the spaces between them, we can point the finger at many potential drivers of high cost. Furthermore, many argue that Congress has done little to address the structural issues that have created opportunities for cost increases in the first place.

Employers, who bear the primary responsibility for health insurance coverage outside of the government, are bracing for another year of double-digit cost growth, and recently, many have identified increases in unit price – the cost of the individual healthcare services that their employees receive – as a key driver of that growth (Mercer, Reference Umland and Patel2025). While there is no single cause of excess spending on US healthcare, here we endeavour to explain an issue of primary interest to the authors over the course of the last three years: how unit price is inextricably related to provider organisational structure.

We highlight recent research findings that we believe offer a more complete picture of the current environment and suggest pathways to advance the public discourse on the topic.

1.2. A case for consolidation: why is it so hard to be an independent physician?

According to the American Medical Association (AMA), the cost of running a physician practice has increased by nearly 50 per cent from 2001 to 2023, while Medicare reimbursement to physicians has increased by less than 10 per cent over the same time period (AMA, 2023). This is in contrast to hospitals and skilled nursing facilities, for which payment has increased by around 70 per cent since 2001 (AMA, 2023). While these payment rates are specific to Medicare, private payers’ common practice of anchoring their payments to Medicare reimbursement further exacerbates the impact of non-inflationary updates. For providers to have any chance at successfully negotiating payment increases from private insurers, they must either position themselves as critical to network adequacy or take on financial risk, both of which require significant scale.

Beyond reimbursement challenges, providers face greater administrative burden caused by the growing complexity of the system, regulatory whiplash, and technological advances that require significant capital investment. Taken together, the existential question about the future viability of independent physician practice has recently gained attention (Bernard et al., Reference Bernard, Lucarelli, Hertz, Eagle and Swanson2026).

1.3. So, what can providers do to combat the challenge that reimbursement cannot keep pace with costs?

Increasingly, the answer has been that going it alone is simply no longer sustainable.

It is no secret that provider consolidation has increased considerably over the course of the last several decades (Bai et al., Reference Bai, Zare, Eisenberg, Polsky and Anderson2021; David, Reference David2009; Pauly et al., Reference Pauly, Burns, Benitez and Sielski2025; Tsai et al., Reference Tsai, Duggan, Jie Zheng and Epstein2025). Driven by an increasingly complex, burdensome, and expensive operating environment, physicians may leave private practice to become employees of hospitals and hospital systems, sell their practices to large corporations like Optum or McKesson, or choose to find another source of capital or financing partner to help fund their existing practices. In fact, a recent report – which was included, among other studies, in a recent comprehensive analysis of healthcare consolidation published by the US Government Accountability Office (GAO), the government watchdog for Congress – indicates that nearly 80 per cent of physicians in the US were employed by a hospital or another corporate entity (e.g. an insurer, a managed service organisation, etc.) as of 2023, a shift away from small-group, independent private practice (GAO, 2025; Physicians Advocacy Institute, 2024).

In a recent AMA survey, over 70 per cent of recently acquired physician respondents indicated that ‘the need to ‘better negotiate higher payment rates with payers’’ was the number one factor that led to the sale of the practice (AMA, 2025). Therefore, the question of scale becomes critical: how do we balance the fact that physicians must consolidate to achieve sufficient scale to continue to serve their patients with the fact that this scale – achieved through consolidation – is a consistent driver of unit price growth?

1.4. Affiliation model and its impact on care delivery

In the US healthcare system, many services are eligible to be rendered and paid for in multiple settings, including the inpatient hospital setting, the hospital outpatient department (HOPD), an ambulatory surgical center (ASC), the physician office or clinic, and the home, among others. The cost of providing a given service may vary by setting of care (SOC), but more importantly, so may the price that payers – especially Medicare and commercial insurers – are willing to pay. Therefore, the most direct path to lowering the unit price of care may be to optimise the choice of SOC, particularly when differences between SOCs are indiscernible to the patient, or when a service can be delivered in multiple SOCs without diminishing quality or patient safety.

Intuitively, the impact of SOC optimisation – that is, directing patients to the lowest-cost, most clinically appropriate SOC for a given service without negatively affecting quality – has the greatest potential financial and operational impact for high-cost, high-volume services (Avalere Health, 2025; Avalere Health, 2025). However, the growth in consolidation – particularly of physicians into hospital systems – creates incentives that run counter to an effort to optimise care to the lowest-cost settings. The authors see four primary mechanisms at play that have led to growing costs associated with consolidation:

1. 1. Scale itself leads to better negotiating leverage, allowing provider organisations to extract higher reimbursement rates from insurance companies when they represent a meaningful access point for care (US states regulate minimum standards for insurance provider network adequacy).

2. 2. Within larger provider organisations, particularly hospital systems, the lines between SOCs can be blurry – often a function of billing practices rather than physical location.

3. 3. As part of broader efforts to increase the integration of care delivery, physicians often have limited ability to refer care outside of their own hospital system, even when a lower-cost setting may exist.

4. 4. Within large organisations, service lines often cross-subsidise one another, so service delivery is rarely operationalised on a single-service basis.

To test our hypothesis that the shift towards more expensive SOC decisions is in part a by-product of consolidation, we investigated the relationship between practice model and SOC dynamics, as well as the difference in total service cost by SOC across a more nuanced set of practice models, described further below (Kapoor et al., Reference Kapoor, Camel, Eagle, Makhoul, Maroney, Yang and Berggreen2025).

We identified a set of 32 Current Procedural Terminology® (CPT) codes that met high thresholds for material contribution to total payment and volume in four specialties: cardiology, gastroenterology, orthopaedics, and urology. These specialties were selected based on strong representation of surgical and imaging services – which are often more costly than, for example, basic evaluation and management services – performable across multiple SOCs, as well as a history of private investor interest in acquiring managed service organisations supporting these specialties. The analysis investigated the relationship between practice model and SOC dynamics, as well as the difference in total service cost by SOC across models. The four models were unaffiliated private practice, private equity (PE)-affiliated private practice, corporate, and hospital. The first analysis, the cost analysis, compared the highest-cost outpatient SOC for each code, always the HOPD, with the most-utilised lower-cost outpatient SOC, the ASC or the office. (In the US, only outpatient settings have comparable payment systems, hence the absence of inpatient data.) The second analysis, the likelihood analysis, calculated the likelihood that each service would be performed in the lower-cost SOCs based on the affiliation model of the physician providing the service.

The results are compelling. The cost analysis showed that, across the 32 codes analysed, the total Medicare reimbursement in the HOPD setting was 1.2 to 8.6 times higher than in the lower-cost setting (the ASC or the office, depending on the service). The cost differential was even greater and more variable in the commercial insurance market, from 1.1 to 13.4 times greater. For all codes across all settings, commercial reimbursement was greater in magnitude than Medicare reimbursement, and the difference between the higher-cost setting and the lower-cost setting was greater in commercial insurance than in Medicare for all but one code.

Interestingly, when we overlaid the impact of physician affiliation model, the relationship between cost and practice model began to emerge. The likelihood analysis showed that PE-affiliated physicians were most likely to provide a given service in the lower-cost settings (ASC and office combined) for 29 of the 32 codes analysed, including four ties. In aggregate across the 32 codes assessed, PE-affiliated physicians were most likely (63 per cent) – and hospital-affiliated physicians, least likely (37 per cent) – to provide care in a lower-cost setting.

Taken together, these results tell us several things:

1. 1. For a given service in a given SOC, commercial prices are generally higher than Medicare reimbursement rates.

2. 2. In most cases, the gap between lower-cost and higher-cost settings is disproportionately greater in commercial insurance compared to Medicare.

3. 3. PE-affiliated physicians are most likely – and hospital-affiliated physicians, least likely – to provide selected services in lower-cost SOCs.

These findings supplement our earlier research, which showed that patients of physicians who left unaffiliated practice for corporate and hospital models saw increased annual Medicare expenditures, while those whose physicians transitioned to a PE model saw a decrease in annual Medicare expenditures, and that patients of physicians who affiliated with a PE-backed model had fewer inpatient days and emergency department visits (Avalere Health, 2024). Together, these analyses suggest that additional research and consideration of the ways in which private capital can be a positive dynamic within the provider market consolidation discourse is warranted.

While our study introduced new evidence, we acknowledge that (a) our findings are limited to a small subset of specialties within the physician landscape and (b) we focused on a small but impactful subset of outcomes metrics to define quality. Our hope is that this new evidence pushes researchers and policymakers to consider a more nuanced set of dynamics within the provider market consolidation debate.

1.5. Professional versus facility fees: understanding SOC pricing dynamics

It is important to note that the analyses use total service cost. That means considering both the professional fee – the reimbursement paid to the physician for their time – and the facility fee, which is paid to the facility (i.e. the HOPD or the ASC) for overhead costs. The facility fee is typically much greater in magnitude than the professional fee for surgical services. In other words, it is usually the facility fees that make higher-cost settings more expensive.

Our methodological rigour in disaggregating unit price into its facility fee and professional fee components is critical. For example, one well-publicised report found that, in eight of the ten specialties it studied, PE acquisition of physician practices led to statistically significant price increases ranging from 4 to 16 per cent (Scheffler et al., Reference Scheffler, Alexander, Fulton, Arnold and Abdelhadi2023). However, a careful reading of the paper’s methodology reveals that it only considered data from professional claims: it does not contemplate facility fees, which represent the majority of the cost of many services, at all.

We acknowledge this directly in our latest publication:

As the body of research progresses in this segment, consideration of the distinction between professional fees and facility fees is an important nuance, particularly if we are to consider policy solutions with some element of SOC optimisation. This is particularly relevant for care that can be safely and effectively provided in a physician office – where there is no facility fee – but is instead provided in (and billed as) an institutional setting (such as a HOPD or ASC), and a facility fee is added.

1.6. The case for more nuance in the discourse around the role of private capital

Over the last two to three years, stakeholders have spent considerable energy casting PE’s involvement in US healthcare in a negative light. Further, state lawmakers continue to introduce bills that would directly limit PE investment in healthcare, though few – if any – have been enacted (PE Stakeholder Project, 2026). While we do not disagree with criticisms of the documented bad actions that some PE firms have taken and believe that additional oversight is necessary to prevent poor patient outcomes, we would posit that all provider types and financial organisational structures exhibit examples of bad actors and should be assessed equitably, for three primary reasons.

1. 1. PE affiliation is still relatively rare, with recent research indicating that the prevalence of PE-backed providers is as low as 4 to 6 per cent (Avalere Health, 2024; GAO, 2025; PitchBook, 2024).

2. 2. The US healthcare policy environment seems inadvertently designed to all but guarantee consolidation of the provider market.

3. 3. Providers need multiple capital management options in order to remain independent, and affiliation models differ from one another in the services they offer, the business models they use, and the legal, financial, and operational relationships they have with physicians.

1.7. Where do we go from here?

Knowing that (1) physician consolidation is a forced by-product of today’s US healthcare system, (2) physicians’ decision-making with respect to SOC is correlated with their practice model (i.e. affiliation decision), and (3) there are considerable cost differentials between SOCs, what do we recommend?

To start, we argue that we should spend far less time analysing PE in a silo and more time discussing how to ensure providers, regardless of the capital structure they select, have the autonomy they need to care for patients.

Through our research, we have observed that the existing literature typically compares PE affiliation to either independent affiliation (Scheffler et al., Reference Scheffler, Alexander, Fulton, Arnold and Abdelhadi2023; Singh et al., Reference Singh, Song, Polsky, Bruch and Zhu2022) or to ‘all other’ affiliation types (Braun et al., Reference Braun, Bond, Qian, Zhang and Casalino2021; Bruch et al., Reference Bruch, Sameer Nair-Desai and Tsai2022). These approaches neither depict the full landscape of provider affiliation types nor capture the reality that physicians commonly consider hospital employment their alternative to independent practice. Small-group, independent practice is increasingly an unrealistic – or even unavailable – option for physicians, due to the combination of financial and administrative strains placed on private practices (as described above). Across the five specialties we studied in our 2024 analysis, just 12 per cent of physicians remained in unaffiliated private practice in 2022 (Avalere Health, 2024). Furthermore, comparing PE affiliation to ‘all other’ groups in aggregate may not reflect the decision that physicians often face: either to (a) affiliate with a management services organisation that has received capital from a PE firm, or (b) sell their practice to a hospital system and become hospital-employed.

While we believe that all research that aims to provide the public with greater transparency and more information about how healthcare is delivered is important, we highlight these examples to emphasise how specific decisions related to research design may have unintentionally shaded our understanding of the on-the-ground dynamics experienced by providers.

Legislative efforts that single out one financial structure or asset class miss the mark and may actually encourage further consolidation of the provider market to the detriment of patients and the healthcare system overall. Our scrutiny should be consistent, agnostic to the nature of the entity enabling the consolidation.

If our longer-term goal is to unlock advancements in technology and provider training that enable the most efficient care-delivery infrastructures, then we need to fundamentally change our reimbursement system to incentivise providers to deliver care in the lowest-cost SOC consistent with their assessment of a given patient’s clinical needs.

1.8. Why “site neutrality” isn’t the answer

Some policymakers have recently focused on ‘site-neutral payment reforms’ which, as the name would suggest, propose to pay for a given healthcare service at a standard rate, regardless of the site (i.e. SOC) in which it is performed. The recent passage of H.R.1, also known as the One Big Beautiful Bill Act, reignited this debate, as Republicans in Congress considered various mechanisms to pay for their additional budget priorities. In its report sizing the potential savings of a variety of proposals being considered, the Congressional Budget Office estimated that adopting site-neutral payments for nearly all services would save the government $180 billion over the next ten years – a sizeable ‘pay-for’ to consider. Narrower versions of site neutrality – for example, applying only to physician-administered drugs or imaging – were estimated to yield more modest, yet still meaningful, savings of $5 billion and $10 billion, respectively.

Ultimately, those provisions were not included in the bill, in large part due to strong pushback from the hospital community and its view of site neutrality as a funding cut. Since summer 2025, the Trump administration has moved forward with site neutrality for physician administered drugs starting in 2026, and some members of Congress continue to pursue broader site-neutral reform (Hassan, Reference Hassan2024).

Despite broad stakeholder interest and a focus on affordability, site-neutral payment faces two critical challenges:

1. 1. The difficulty of deriving the appropriate benchmark from the existing data, and

2. 2. Reconciling the cost of delivering care with physicians’ discretion regarding patient safety.

Unlike other SOCs, ASCs do not currently report costs, so it is difficult to determine the appropriate price against which payments should be benchmarked. Furthermore, unlike services such as physician-administered drug injections and imaging, surgical procedures require greater consideration of patient-specific factors, and broad clinical evidence and consensus guidelines for these procedures are not yet widely available. In fact, while some stakeholders – specifically, the American Hospital Association – have asserted that patient acuity factors make site-neutral payment untenable, the Medicare Payment Advisory Commission (MedPAC) has researched this issue extensively, finding in their 2023 report to Congress that, for most procedures, the difference in risk profiles of patients treated across SOCs has an ‘insignificant effect on hospital charges’. Specifically, their research led them to conclude that ‘adjustments for patient severity are not necessary for the services in the 66 [Ambulatory Payment Classification codes] in our analysis’ (MedPAC, 2023).

With such critical implementation considerations still outstanding, it is clear that more research and testing must occur before widespread site neutrality can become a viable policy option.

1.9. The CMS innovation center and the opportunity to pilot new models

The Centers for Medicare and Medicaid Services Innovation Center, the ‘healthcare sandbox’ of the US federal government, may be best equipped to pilot a transformational payment model, given the flexibility of its demonstration programmes and its mandate to improve care while reducing costs. One ‘carrot’ approach would be to experiment with enhanced physician professional fees when care is delivered in an ASC as opposed to a HOPD. This approach, when paired with effective safety guardrails, could incentivise a different, more cost-effective SOC decision for clinically appropriate cases.

Let us consider an illustrative example. Medicare has allowed total knee replacements (Current Procedural Terminology® (CPT) code 27447) to be performed in the ASC setting since January 2020. Imagine a model for knee replacements in which the professional fee in the ASC setting is 15 per cent greater than the professional fee in the HOPD setting, encouraging surgeons to perform procedures in an ASC. In this scenario, even modest changes in SOC selection behaviour (e.g. moving 2 to 5 per cent of cases to the ASC) would result in billions of dollars in savings to the Medicare programme over five years for just a single procedure.