Author: Albert “Chip” Hutzler and Scott M. Safriet

Download the Original PDF Here

As the hospital/physician landscape continues to evolve, both parties are continually looking for ways to improve their alignment opportunities. The goals of such improved alignment are often focused on better clinical outcomes and improved financial integration. One area that is gaining increased traction is the professional services arrangement (“PSA”), with a particular emphasis on implementing the PSA model in lieu of a traditional employment arrangement.

Sometimes referred to as a “synthetic employment arrangement” or a “foundation model” the PSA allows the target physicians to maintain their autonomy as a free-standing group practice, receiving compensation from the hospital in exchange for providing clinical services, but without the day-to-day obligations of running a business. From the hospital’s perspective, the model allows for a significant degree of flexibility in aligning with a group while avoiding physician concerns over loss of autonomy or inability to easily “unwind” the arrangement should the relationship not work out as planned.

In a typical synthedic employment agreement, the hospital will employ all non-physician staff of the group and will contract with the group practice to provide professional, clinical services to its patients. In exchange, the group receives a set rate of compensation, typically paid on a “per unit” basis, often calculated to include employment taxes, benefits, and certain other retained practice expenses.

This article discusses this emerging compensation arrangement, with a particular emphasis on the rationale of the structure, determination of the appropriate “per unit” compensation rate, and varying issues that can affect the fair market value1 (“FMV”) compensation in such arrangements.


With the possible exception of television stars, it seems that no one in business re-invents themselves more than physicians and hospitals. In response to ever changing regulations,2 whether actually implemented or simply proposed, these two parties are continually exploring new ways to relate with one another, offset changes in reimbursement structures and address the payors’ (and patients’) demands for ever improving clinical outcomes.

Existing against the backdrop of these environmental dynamics is an issue of lifestyle. Physicians are simply placing less and less emphasis on the days of old (e.g., maintaining weekend office hours, taking many days of call coverage each month, etc.), and focusing on spending more time with family, while educating themselves on technological practice advancements.

Enter the PSA: a contractual vehicle that is gaining in popularity in response to this market imperative to integrate for quality and efficiency improvement, while addressing some physicians’ aversion to making the leap to full-employment.


The Hospital: Gets to leverage its infrastructure, payor contracts and provider-based status (for billing of ancillaries).

The Healthcare Professionals: Receive Fair Market Value compensation for clinical services (and can distribute such payment within their practice group as they see fit, subject to certain constraints), a market-based allowance for benefits and malpractice insurance, plus the opportunity to contract with the hospital for staffing and/or management services. All of this is integral to the efficient operation of the service line.


  • The decline of the independent physician model due to decreasing or uncertain reimbursement trends, increasing risk for physicians, and lifestyle considerations;
  • The emergence of integrated health strategies, accountable care organizations (“ACOs”), and shared savings programs, including the Medicare Shared Savings Program (“MSSP”) with the intent of collaborative efforts improving outcomes and reducing costs; and
  • The offer of a viable alternative to the employment model, which cannot be used in many states and has drawbacks such as the relative lack of physician autonomy.

While the synthetic employment arrangement model may not be desirable or even possible in all states, the corporate practice of medicine restrictions in some states may “completely bar” the employment model, making the PSA model the only viable alternative to independent practice. Many states do have exceptions for hospital-physician arrangements or some other legal vehicle (such as the foundation model in California) that allows for a similar type of PSA model relationship to be utilized.


The typical underpinning of the Professional Services Arrangement model is that instead of a traditional employment arrangement, the healthcare professional (or physician group, as applicable) maintains its group practice entity. The doctor enters an agreement under which a hospital purchases the tangible assets and/or leases certain non-clinical staff from the medical doctor. The hospital then provides compensation to the medical provider in exchange for the provision of clinical services.

Whereas a traditional employment arrangement can include a variety of payment approaches, typically the PSA is based upon personally performed productivity, primarily to avoid any possible concern over the physicians having an ownership interest in the entity performing the designated health services (“DHS”) under the Stark law.3 The most commonly used (and likely most supportable) metric is the use of work relative value units (“wRVUs”). wRVUs not only directly relate to the work effort of the physician but as appropriate, also allow the hospital to “gross-up” the rate to include taxes and benefits as well as retained practice expenses (i.e., malpractice insurance, CME costs, etc).4

The development of the applicable wRVU rate necessitates careful evaluation of the Fair Market Value and is the central area where hospitals have tended (to their potential detriment) to overestimate the use of a single methodology.

“Just relying on median rates from the benchmark compensation surveys, or only relying on an internally calculated market approach, and ignoring the results of a cost or income approach can lead to regulatory infringement.5”

By way of background, nearly all transactions between hospitals and referring physicians implicate the Stark Law.6 Therefore, they are required to fit into an applicable Stark law exception, otherwise they are prohibited.7 The potential penalties for failing to meet the Stark law exceptions include repayment of all tainted claims and punitive payments, as well as possible loss of Medicare eligibility. Exceptions to Stark Law which apply to synthetic employment model arrangements require compensation under compliant transactions to be consistent with FMV, as do exceptions related to employment arrangements. Moreover, Stark law defines FMV differently from traditional notions of the term in other settings (most importantly, the IRS definition), thereby affecting the valuation approaches that can be utilized to value a particular arrangement.8

Nevertheless, a healthcare appraiser generally consider the same three approaches that are applied to assets when valuing service arrangements under the Stark FMV standard. The major approaches to value include the Cost, Income and Market Approaches, and their application to PSA model arrangements will be described in greater detail below.9

In many respects, the valuation of PSAs is analogous to valuation of employment arrangements, since both models share many common characteristics. Many of the same approaches and techniques are utilized, with the major difference being that the physicians maintain their own practice entity and direct responsibility for certain costs (e.g., taxes, benefits, etc.) requiring an appropriate adjustment in the analysis.

The determination of FMV with regard to the wRVU rate will often require the appraiser to undertake multiple approaches, as there could be anomalies in the underlying data that negates (or mitigates) reliance on one approach over another. For example, sole reliance on a Market Approach may not take into consideration the high percentage of poor payors in the area and/or the atypical cost structure of the target physician’s practice. On the other hand, an Income Approach may yield over-inflated indications of value if the valuation firm overlooks the need to incorporate an “owner’s return” in the calculations. Furthermore, the Income Approach may not be able to be performed at all if the target physician practice is unable to provide usable financial information.

With regard to a Market Approach, while in many ways it is the most straightforward (and relied upon) methodology in the context of making an FMV determination of a PSA, it is not without its drawbacks, the biggest of which is its singular dependence on making a comparison to available survey data. Essentially, a Market Approach involves a process under which a target physician’s performance is compared against available benchmark data.

Unfortunately, data from otherwise reliable sources can be “misused” in a variety of ways, and in the context of a PSA or employment agreement analysis, there is potential for:

  • Over-emphasis on regional compensation differences;
  • “Cherry picking” from among different surveys or survey tables;
  • Failure to consider ownership/ancillary profits that may be inherent in 90th percentile compensation reported by market survey data;10 and
  • Assuming that there is always a correlation between the reported data for compensation and productivity.

As such, it is prudent to review and analyze multiple metrics of productivity, including wRVUs, professional collections, median compensation per wRVU, etc. Through the use of a “percentile matching technique,” the evaluator will match up each productivity variable with the corresponding expected level of compensation. Since each variable has certain characteristics that may yield unintended results, the evaluator will make a “weighting” determination, primarily based on the unique facts of the arrangement as well as the validity of the data.

The general correlation of reported compensation and productivity is well documented in Medical Group Management Association (“MGMA”) and other physician compensation surveys, though the degree and strength of correlation varies depending on physician specialty and circumstances. This is an inherently logical and fair result.

However, while compensation and productivity are often correlated, parties are often surprised to learn that compensation is inversely correlated with reported productivity ratios. This result suggests that the correlation between compensation and productivity data is simply not linear (i.e., when compensation is plotted against productivity on a graph, the result is a curve shaped relationship – meaning that as productivity increases, compensation still increases, but not as rapidly at higher levels as it increases at lower productivity levels). This counterintuitive result is where the substantial confusion lies and risks the possibility of a significant valuation error.

In contrast to the Market Approach, the consideration and use of the Cost and Income Approaches can serve to offset and mitigate the limitations of the Market Approach. While sufficient data is not always available to allow the use of all three approaches, particularly with the appraisal of PSAs and employment arrangements, the use of multiple approaches allows the appraiser to consider the totality of marketplace considerations in appraising a service arrangement. Furthermore, the Cost and Income Approaches allow a view into the local marketplace, as the valuator can incorporate into the analysis a full array of economic factors that may be affecting compensation of the target physician.

Under the Cost Approach, a valuator endeavors to understand the historical compensation levels of the target physician in order to make a determination about the FMV of a proposed compensation arrangement. However, the relevance of a physician’s historical compensation depends on the degree of “comparability” between the physician’s practice and the proposed employment arrangement. Historical compensation can be considered as an indication of the FMV for the services provided by a physician as long as the service arrangement and corresponding compensation meet certain criteria.


The physician and the entity contracting with the physician for services did not have a referral relationship that resulted in a non-arm’s-length compensation for purposes of establishing FMV under applicable healthcare regulations.11
The services provided historically are substantially similar to those services that will be provided under the proposed service arrangement.
The services are provided in an operational and/or clinical setting that is comparable to the setting under the proposed arrangement.

To undertake the analysis, the valuator will obtain “historical” practice financials for the target physician, and then make a series of normalizing adjustments to allow for an effective comparison. The adjustments to historical compensation include those that make the historical service arrangement comparable to the proposed arrangement in terms of the scope of services provided, the benefits paid in addition to cash compensation, and certain normalized operating costs associated with the provision of the services. The adjustments for normalized benefits and operating costs are made so as to place the key economics of the arrangement on a comparable basis with those expected in the marketplace.12

Similar to the Cost Approach, the Income Approach also requires certain financial data, but instead of normalizing historical information, the valuator will request a practice pro-forma statement of operations to be developed by the hospital.13 The Income Approach can be used to forecast the “distributable earnings” available for physician compensation in a practice. From this amount, the valuator would make a deduction from the distributable earnings for an estimate of market-level benefits for the physicians.

In addition, recognizing that there is a carrying cost to the deployment of required assets required to operate a practice, the valuator would also make a deduction from distributable earnings to account for an appropriate “owner’s return on invested capital” in order to arrive at the guideline level of compensation for physician services. The valuation concept behind this approach is that a forecast of the distributable earnings reflects the future value of the physician services provided to the practice. The forecast, however, should be prepared consistent with the conceptual framework of FMV, which entails the assumption that the practice will be operated by a hypothetical, typical employer entity.

Once all applicable approaches are completed, the valuator must synthesize the results to arrive at the applicable indication of FMV compensation. In performing this synthesis, the valuator should consider whether one or more approaches yielded values that do not appropriately reflect the intent of the agreement, and each approach is assessed in terms of its relative strengths and limitations.

For example, the valuator may believe that the Cost Approach, reflective solely of past expense structure and physician earnings, may not be reflective of future expenses and reimbursement levels (and thus future compensation) under a proposed arrangement. Alternatively, the valuator may ascertain that the pro-forma developed by the hospital may appear to include unreasonable assumptions. Regardless, the valuator will typically identify one or more of the approaches as yielding relevant values, and will determine the appropriate “weighting” to apply to each in order to calculate the applicable level of FMV compensation. This resulting compensation range is then converted into the proposed compensation structure to allow for a meaningful comparison.



In reviewing salary survey data (e.g., MGMA14), it is important to note that while such surveys are heavily relied upon, they are not always the definitive snapshot of physician compensation in the marketplace. The widely used reference compensation surveys from MGMA, AMGA,15 and other associations and organizations are based on voluntary participation by the respondents, without the use of any statistical sampling methods or means of validating the responses. As such, without a careful review and application of the data, the user may not realize that selecting reference values from different tables could lead to problems (e.g., selecting national wRVU levels but regional compensation), or may fail to recognize that ownership and/or ancillary profits may be inherent in the higher percentiles of reported compensation. Following are four of the more commonly observed misconceptions and/or misuses of the compensation surveys:16


As instructed by the survey questionnaires, it is intended that total cash compensation from all sources, including clinical services, ancillaries (and technical revenues), medical directorships, on-call coverage, other service arrangements and owner compensation, be reported. Therefore, it can be rather easy to create a compensation “stacking” issue if one assumes that the survey data relates only to clinical services. A “stacking” issue refers to a situation where several different elements of compensation are separately evaluated and compared to total compensation survey data and deemed to be consistent with FMV on an individual basis. However, when such elements are “stacked” together in a comprehensive employment arrangement, the results far exceed survey data for total compensation. As such, the appropriate method is to also compare aggregate compensation to total compensation survey information to ensure that aggregate compensation is consistent with FMV. Stacking is discussed in greater detail below.


In fact, the MGMA survey has clearly stated since 2009 that, in its reported survey data, there is an “inverse” relationship between productivity and the compensation per wRVU rate.17 As such, the highest wRVU producers have the lowest comp/wRVU rate. This result is somewhat counterintuitive and leads to substantial confusion and valuation risk.


This assumption defies statistical validity, as by the very definition of the term “median,” 50 percent of the respondents make less (and perhaps far less) than this amount. As such, while it may sound conservative, simply defaulting to the median compensation/wRVU value as an indication of FMV may lead to an overcompensation bias in the arrangement. Thus, while 50 percent of respondents do make more than the median compensation per wRVU rate, 50 percent make less, and thus, determination of the appropriate rate requires analysis of specific circumstances and substantial care to ensure the rate properly accounts for the inverse relationship between productivity ratios and productivity itself (see misconception #2 above).


The major compensation surveys are published annually and reflect the marketplace of the prior year. Thus, they are likely not reflective of any known reimbursement or marketplace changes that are impending or which have occurred since the data was gathered. Compensation from year to year is variable and volatility is due to several factors, including geographic location, cost of living, economic conditions, and most importantly, changes in reimbursement patterns and medical advancements. Furthermore, the surveys only reflect the outcomes for those that responded, an inherent bias in and of itself.


As with most aspects of a transaction, the “devil is in the details,” and the proper calculation and treatment of wRVU’s is no exception, especially in the PSA, as the compensation is almost always exclusively based on physician productivity. By way of additional background, the wRVU is one of three components that make up the “total” RVUs used by Medicare (and many other payors) in the adjudication of physician claims.

Unfortunately, if not careful when compiling the historical productivity information for a physician or group practice, a possible mistake is to inadvertently report total RVU’s (i.e., to also include relative values for practice expense and malpractice risk) and not simply work relative values (i.e., those solely related to the work effort of the physician). As such, a physician’s expected productivity may be significantly overstated, and thus, in any applications of the market approach which attempt to appropriately match compensation with productivity, the resulting indications of FMV compensation will be commensurately overstated.

A second area that can result in an overstatement of wRVUs results from the failure to consider CPT coding modifiers.18 As an example, a physician reporting a CPT code for a surgical procedure could have been an assistant at surgery which carries a reduced wRVU rate than for the primary surgeon. Similarly, when a surgeon performs multiple procedures on the same patient, the appropriate modifier must be used, as all follow-on procedures to the primary surgical event are subject to a reduction in value.



Midlevel providers refer to non-physicians who provide medical care, such as nurse practitioners, physician assistants, imaging technologists, and other similar providers. This is always a tricky aspect of a PSA analysis, as the valuator needs to be careful to address where the midlevel provider resides in the equation (e.g., if the physician is leasing the midlevel to the hospital, which is the most common practice, the benefit of any work performed, whether “incident to” (which refers to office-based services provided by the midlevel under the physician’s supervision which are typically reimbursed at full physician rates) or at the “midlevel provider rate” (typically services performed by midlevels in a hospital setting, where “incident to” billing is not allowed under Medicare rules, and which are reimbursed at about 80 percent of the physician rate) would accrue to the hospital). However, if the midlevel remains the financial responsibility of the physician (for example, as reflected on the practice financials in a PSA arrangement), then it would be appropriate to have the legitimate “incident to” wRVUs count toward the physician’s productivity.


It is becoming increasingly common for hospitals to contemplate entering into PSA’s with multi-specialty physician practices (e.g., a cardiology practice that includes general cardiology, interventional cardiology and electrophysiology). Furthermore, for ease of administration, the hospital may want to pay the same compensation rate per wRVU regardless of the subspecialty. In this scenario, it is important not only to (i) calculate an applicable rate per wRVU for each subspecialty, but (ii) also obtain historical information relative to the percentage mix of wRVUs for each subspecialty, such that the applicable “weighted average” rate can be calculated. Lastly, it is also paramount to run a series of sensitivity analyses at varying levels of potential total wRVUs to ensure that the blended number still yields FMV results, regardless of the level of productivity.

One final area that deserves comment relates to the inclusion of applicable taxes and benefits in the wRVU rate (i.e., the aforementioned “gross-up”). Up until several years ago, it was fairly standard practice for valuation firms to simply acknowledge a hospital’s statement to the effect that “benefits provided will be consistent with similarly situated employees.”
More recently, many in the valuation community have opened their eyes to the fact that physician benefit plans are becoming more and more robust. In certain situations, a very robust benefits plan, significantly higher than what comparable physicians receive, may be a source of a material increase in economic benefit to the physicians, and such excess benefits would generally need to be regarded as additional compensation in the FMV analysis.

On a related note, since the reasonable benefits and taxes are typically included in the “grossed-up” wRVU value, it is important to consider whether an increase in the physicians’ wRVU productivity will result in an overpayment for benefits and taxes since these expenses are mostly fixed (i.e., once the FICA limit has been reached). This is commonly handled by placing a dollar cap on overall benefits.


While the underlying payment methodology for the employment or PSA Model may be commercially reasonable and produce results that are consistent with FMV, a hospital’s desire to involve the same physician in a variety of additionally compensated arrangements may result in an overall compensation arrangement that quickly becomes problematic. Since most PSAs are established on a pure productivity basis, most of the typical “stacking” culprits may be mitigated.

For example, a common accompanying arrangement is a medical directorship. Since these are most often compensated on the basis of worked and documented hours, there is usually not a large concern about overcompensation in the overall arrangement, since time devoted to these administrative duties may result in the generation of fewer wRVUs. However, what may appear to be a subtle change to the fact pattern (e.g., the physician is paid a flat annual amount for medical director services in exchange for providing a “minimum” number of hours) can result in a situation where the total compensation may exceed FMV. In this scenario, without an effective time tracking mechanism in place, there is a possibility that the physician does not provide the required minimum number of medical director hours.

So, beware the notion that, by calling elements of compensation by different names, one can continue to “stack” the compensation higher. Essentially, all sources of compensation (including benefits, as discussed earlier) should be considered in the aggregate to establish compliance with FMV.


Sign-on bonus Call pay
Quality bonus Service line management arrangement19
Productivity bonus Tall insurance20

With careful planning and execution, a PSA may be an ideal vehicle to allow a hospital and physicians to better align their clinical goals and objectives. When properly structured, the PSA can mitigate physicians’ aversion to traditional employment and the accompanying loss of autonomy. However, as with employment arrangements, PSAs must also comply with the FMV standard. The PSA must be commercially reasonable and consistent with FMV, both with respect to its individual components and when considered in the aggregate.