Since the completion of this article, CMS has announced that its final Stark law rule will be published in the Federal Register on December 2, 2020.
The Stark Law has been litigated for the past two decades through the False Claims Act, resulting in case law precedent with far-reaching implications for hospitals and healthcare systems. In this article, we explore hospital-physician arrangements, which are often protected by the Stark Law’s employment exception. On its face, the employment exception appears to broadly protect physician arrangements, but judicial interpretation has limited that protection. At the heart of the controversy is the prohibition against compensation taking into account the volume or value of referrals, and the allowability of a productivity bonus based on personally-performed services under the employment exception. Landmark cases have created a problematic “correlation” theory positing that compensating physicians for professional services resulting in hospital services income, such as a facility fee during an outpatient surgery, takes into account the volume or value of the physician’s referrals and therefore violates the Stark Law. The Centers for Medicare & Medicaid Services (“CMS”) has dismissed this correlation theory in multiple instances in commentary to rulemaking. Nonetheless, courts continue to invoke the theory, using the False Claims Act and its treble damages provision to extract major judgments and settlements out of defendants. Practically speaking, this misalignment leaves hospitals and healthcare systems wading into murky water, transforming decisions about commonplace physician arrangements into a whirlpool carrying potential significant risk.
II. The Stark Law & Physician Relationships
The Stark Law prohibits a physician from making a referral for designated health services (“DHS”), including inpatient and outpatient hospital services, that are payable by Medicare to an entity with which the physician, or an immediate family member, has a financial relationship. It also prohibits that entity from submitting a claim for DHS furnished pursuant to the prohibited referral. A financial relationship includes ownership or investment interests, direct compensation relationships, and indirect compensation relationships. A direct compensation relationship exists if remuneration passes between the referring physician and the entity furnishing DHS without any intervening persons or entities. An indirect compensation arrangement exists between a referring physician and an entity furnishing the DHS if: (1) there is an unbroken chain of financial relationships between the referring physician and the entity furnishing the DHS, (2) in that chain, the referring physician directly receives aggregate compensation from an entity that varies with, or takes into account, the volume or value of referrals or other business generated by the referring physician for the entity furnishing the DHS, and (3) the entity furnishing DHS has actual knowledge, or acts in reckless disregard or deliberate ignorance, of the fact listed previously at element number two. All three elements must be present to find the existence of an indirect compensation arrangement.
Physician arrangements that would otherwise violate the Stark Law are protected if they fit squarely within an exception. One exception for direct compensation arrangements, the bona fide employment relationships exception, protects compensation paid by an employer to a physician if, among other requirements, the compensation is not determined in a manner that takes into account (directly or indirectly) the volume or value of any referrals of DHS by the referring physician to the employer. This prohibition on basing compensation on the volume or value of a physician’s DHS referrals does not prohibit payment of compensation in the form of a productivity bonus based on services personally performed by the physician.
Importantly, a physician’s personally-performed services are not considered referrals under the Stark Law. However, neither the statute nor the regulation clarifies whether the technical component (i.e., facility fee) associated with a physician’s personally performed inpatient or outpatient hospital service constitutes a referral.
III. Judicial Interpretations
The Stark Law’s prohibition on physician compensation methodologies that vary with or take into account the volume or value of a physician’s referrals of DHS has presented abundant opportunities for qui tam relators to bring cases under the False Claims Act alleging noncompliance. Relators and the government have advanced theories that hospital-physician arrangements either (1) do not comply with the Stark Law’s employment exception because they take into account the volume or value of referrals, or (2) do not comply with the Stark Law because they present indirect compensation arrangements that vary with or take into account the volume or value of referrals and do not fit within an exception. Under both theories, relators claim that the hospital-physician arrangements failed to meet a Stark Law exception and therefore resulted in false claims. False Claims Act decisions and settlements have created increased scrutiny for all hospital-physician arrangements. Three particular cases across three different jurisdictions highlight this risk.
The Halifax case made headlines in 2014 when Halifax Hospital Medical Center paid $86 million to settle allegations of Stark Law violations; this was a record-breaking Stark Law settlement at the time. In that case, Halifax medical oncologists received incentive compensation in the form of a percentage of the medical oncology program’s operating margin, which was used as a bonus pool. The program’s operating margin included revenue from items and services not personally performed by the referring physicians, such as outpatient prescription drugs, which are DHS. The medical oncologists received a percentage of the bonus pool based on their personally-performed services. The court found that the arrangement violated the Stark Law because the size of the bonus pool varied based on the volume or value of referrals by the medical oncologists. The bonus pool was comprised of a percentage of operating margin, not based upon the physicians’ personally-performed services, but the bonus pool was divided among the physicians based upon their personally-performed services. Because the physicians could directly increase the overall size of the bonus pool based upon their DHS referrals, the court determined that the incentive compensation did not comply with the Stark Law’s employment exception.
Tuomey, which has become known as a landmark Stark Law case, resulted in a jury verdict judgment of $237 million, and a settlement thereafter involving a $72.4 million payment to the United States and the sale of the hospital to a multi-hospital healthcare system. In that case, Tuomey Healthcare System employed certain specialists as part-time employees with particularly lucrative terms, allegedly to avoid a loss of surgical procedures. The physicians were required to perform outpatient procedures at Tuomey Hospital or Tuomey-owned facilities. In return, the physicians were paid annual base salaries conditioned on net cash collections for outpatient procedures. The physicians were also eligible for productivity bonuses “pegged at eighty percent of the amount of their collections.” Overall, the aggregate compensation paid to the physicians, including base compensation and productivity bonuses, was based on collections for the physicians’ personally-performed services. Collections for personally-performed services included collections from the professional component of the service (personally performed by the physician) and collections from any facility fee or technical component of the procedure (not personally performed by the physician).
The case went to trial twice, once in 2010 and again in 2012 after the Fourth Circuit determined procedural errors in the first trial necessitated a retrial. At the second trial, the jury found that Tuomey violated the Stark Law, which finding resulted in a finding that Tuomey violated the False Claims Act. Specifically, the jury found that the compensation methodology took into account the physicians’ volume or value of referrals, based on the facility fee that Tuomey billed with each physician-performed service and the net collections methodology that allowed physicians to receive compensation from that facility fee. The jury found that the facility fee constituted a “referral” under the Stark Law and, therefore, that the physician compensation methodology took into account the volume or value of the physicians’ DHS referrals. The Fourth Circuit affirmed the judgment, clarifying that the facility fees associated with the physicians’ outpatient procedures were “referrals” under the Stark Law, despite the fact that the physicians had personally performed those outpatient procedures. “In sum, the more procedures the physicians performed at the hospital, the more facility fees Tuomey collected, and the more compensation the physicians received in the form of increased base salaries and productivity bonuses.”
In 2019, the Bookwalter case made waves when the Third Circuit Court of Appeals perpetuated the Tuomey standard, making clear that the controversy over the Stark Law’s prohibition on referrals is as relevant as ever for hospitals and healthcare systems. In Bookwalter, a subsidiary of UPMC employed neurosurgeons for professional services and paid those physicians based on their personal wRVU production, a typical industry practice. In the litigation, the parties agreed there was an unbroken chain of entities with financial relationships, since the hospital system owned each hospital as well as the physician groups and the physician groups paid the physicians. The relator argued that the physicians had indirect compensation relationships with UPMC based, in part, on the wRVU-based compensation methodology, which allegedly varied with, or took into account, the volume or value of referrals or other business generated by the referring physicians to UPMC. The relator also argued that those indirect compensation relationships failed to fit within an exception to the Stark Law.
The district court granted UPMC’s motion to dismiss for failure to state a claim. However, in September 2019, the Third Circuit reversed the dismissal and remanded the case to the trial court, citing the Tuomey case. “The surgeons’ salaries rose and fell with their referrals. The more procedures they did at the hospitals, the more referrals they made, and the more they would earn…” The court found that their aggregate compensation, based upon wRVUs, conclusively varied with their referrals’ volume and value.
On December 20, 2019, the same Third Circuit panel of judges issued a rehearing decision, finding that plaintiffs had sufficiently pled that the employed physician’s compensation took into account the volume or value of referrals and that it was UPMC’s burden to prove the arrangements fit within an exception. However, the court removed the controversial conclusion that wRVU compensation by default “varies with” the volume or value of referrals. “We need not resolve the meaning of varies with here. Regardless, the complaint plausibly alleges that the surgeons’ compensation takes into account the volume or value of their referrals. Under the Stark Act and its regulations, compensation takes into account referrals if there is a causal relationship between the two. And here, the surgeons’ suspiciously high compensation suggests causation.” Essentially, instead of concluding that the compensation varied with the volume or value of referrals, the court said the suspiciously high compensation sufficiently suggested that the compensation took into account the volume or value referrals. Additionally, the court withdrew its earlier reliance on Tuomey.
IV. Summary of the Controversy
Three federal appellate courts have now weighed in on the volume or value standard of the Stark Law: (1) the Eleventh Circuit in the Halifax case, (2) the Fourth Circuit in the Tuomey case, and (3) the Third Circuit in the Bookwalter case. Most interestingly, these decisions are at odds with CMS’s interpretation of its own regulations.
In 2004, before any of the three aforementioned decisions, CMS stated in the preamble to the Stark Law Phase II regulation, “the fact that corresponding hospital services are billed would not invalidate an employed physician’s personally performed work, for which the physician may be paid a productivity bonus.” Nonetheless, Tuomey later held that the inclusion of the facility fee created a Stark-prohibited referral. This tension caused CMS to respond and address the Tuomey decision directly in its 2019 proposed rulemaking, clarifying its position on the “correlation” theory vis-à-vis the technical component of hospital services:
“[…] commenters expressed concern that, following the July 2, 2015 opinion of the United States Court of Appeals for the Fourth Circuit in United States ex rel. Drakeford v. Tuomey Healthcare System, Inc., CMS may no longer endorse this policy.
[…] for clarity, we reaffirm the position we took in the Phase II regulation. With respect to employed physicians, a productivity bonus will not take into account the volume or value of the physician’s referrals solely because corresponding hospital services (that is, designated health services) are billed each time the employed physician personally performs a service…”
This language was used in UPMC’s petition for rehearing in Bookwalter. In appealing the Third Circuit’s initial decision from September 2019, counsel for UPMC cited the 2019 proposed rule, issued in October 2019, as grounds to rehear the case. Counsel for UPMC argued:
“Until now, nothing—not thousands of pages of CMS guidance and regulations, not any enforcement action by DOJ, not any opinion of a court—gave hospital systems a hint that wRVU-based compensation triggered a viable FCA suit based on the Stark Act. Under the panel majority’s novel interpretation, all a relator need do to survive a motion to dismiss is allege that a defendant follows the standard industry practice of tying physician compensation to wRVU productivity. If the unprecedented nature of this interpretation were not enough to warrant rehearing, CMS’s newly-issued disagreement with this interpretation should be.”
One may infer that the Third Circuit gave weight to CMS’s language and stance, because, as described in Section III.c. above, the same panel of judges reversed the decision three months later, withdrawing the conclusion that the term “varies with” categorically means “correlation.” As noted above, the court also withdrew its earlier reference to Tuomey. Nonetheless, the court held that the case could proceed because the relators alleged facts that the arrangement violated the Stark Law that were sufficient to survive the motion to dismiss.
V. Policy Discussion
At this point, hospital and healthcare organization executives are likely scratching their heads, and with good reason. Halifax, Tuomey and Bookwalter, in conjunction with CMS commentary, present puzzling, seemingly insolvable, contradictions regarding permissible hospital relationships with physicians. One lesson is that although CMS commentary to regulations may be helpful, and at times persuasive, hospitals cannot rely on such commentary to protect an arrangement from Stark Law or False Claims Act scrutiny or liability. Even when courts are willing to consider agency commentary, there is no guarantee that all commentary will be considered. By way of example, in reversing and remanding for a new trial in Tuomey, the Fourth Circuit relied heavily on CMS commentary from the Stark Law Phase I preamble, published in 2001, when it found that physicians’ personally-performed inpatient surgeries included referrals by way of the technical component of the surgical service. However, the court did not rely on another CMS statement, in that same commentary, that an arrangement may still fit within an exception if a hospital pays a physician for professional services and that physician refers to the hospital. Nor did the court consider more recent CMS commentary from the Stark Law Phase II preamble, published in 2004, asserting that a corresponding hospital service (facility fee) would not invalidate employed physicians’ personally performed work. Why did the court not consider these statements in its decisions in Tuomey? This peculiar outcome is one that any defendant in a False Claims Act case risks: the court either may ignore CMS commentary, or it apply some portion of the commentary as it sees fit.
This begs the question, why has CMS not confirmed its interpretation by adding it into the regulation itself? In the 2019 proposed rule, CMS doubled-down on its Phase II interpretation from 2004 by saying that, even in light of the Tuomey decision in 2015, it “reaffirmed” that physician employment relationships would not inappropriately take into account the volume or value of referrals solely because a corresponding facility fee was billed during a hospital service. However, in observing the decisions from False Claims Act cases such as Halifax and Tuomey, which came after CMS commentary had been published, why did CMS not take that position in regulation? It is clear that commentary will only go so far; the regulation itself must reflect CMS’s interpretation of permissible compensation.
Where does this leave hospitals and healthcare systems trying to enter into Stark Law-compliant physician arrangements? The risk vis-à-vis the False Claims Act is apparent. Even though the Third Circuit reversed its controversial “varies with the volume or value of referrals” conclusion in Bookwalter, perhaps considering the CMS commentary on the matter, the court nonetheless allowed the case to move forward. The parties could end up with the same result, that is, a finding that the wRVU compensation varied with the volume or value of physician referrals due to the inclusion of the technical component of the services. Even without knowing the conclusion to Bookwalter, it seems clear that hospitals and healthcare systems are at risk for suits such as this given the Third Circuit’s acceptance of the correlation theory as a basis for a False Claims Act action.
VI. Considerations and Conclusion
Hospitals and healthcare systems have a couple of options when considering how to structure arrangements for physician services. The first is a direct employment relationship between the hospital and the physician. Although this may seem the simplest approach, the risks are evident in the negative case law regarding the correlation theory we’ve summarized above. If a hospital employs and pays a physician based on wRVU methodology, and that physician orders DHS from the hospital, there is apparent False Claims Act risk. Unless and until CMS clarifies its interpretation about this issue in formal regulation, the risks will remain.
Alternatively, hospitals may form a subsidiary physician group as a group practice under the Stark Law. Under this model, the physician group would employ physicians who would provide clinical services to or at the hospital. To the extent the physician group provided any DHS, the group practice model would allow the physician group to avail itself of (1) the in-office ancillary services exception for DHS provided by the group, and (2) the special rule for productivity bonuses and profit sharing for group practices. Specifically, physicians in a group practice may be paid a bonus that directly relates to the volume or value of DHS referrals by the physician if the referrals are for services “incident to” the physician’s personally performed services. Again, this would apply for DHS provided by the group itself. Additionally, this model creates distance between the hospital and the referring physician, since there would be an intervening entity (the group practice). As such, there would be no direct compensation relationship between the referring physician and the hospital-provided DHS, such as inpatient and outpatient hospital services.
Although the second option is likely preferable to the first option if a physician is being compensated based on wRVU methodology, it is not without risk. If the hospital owns the group, it creates an unbroken chain of financial relationships: an ownership relationship between the hospital and the physician group, and a compensation relationship between the physician group and the referring physician. The arrangement could be considered an indirect compensation arrangement between the hospital and the physician if all elements of the indirect compensation arrangement definition are met. If the arrangement is considered an indirect compensation arrangement, the parties will need to fit within the indirect compensation relationship exception. Indirect compensation analyses are complex and difficult for many reasons, including the complication of court interpretations of the volume or value of referrals standard. Once in that realm, the parties are at risk of correlation theory allegations similar to those in Tuomey and Bookwalter as described above.
Regardless of the physician arrangement chosen, hospitals, healthcare systems, and physicians must exercise caution when wading into these murky waters.
 42 U.S.C. § 1395nn.
 Some courts have also found the Stark Law applicable to Medicaid. That discussion is outside the scope of this article.
 42 C.F.R. § 411.354.
 42 C.F.R. § 411.354(c)(1)(i).
 42 C.F.R. § 411.354(c)(2)(ii).
 All exceptions may be found in the regulations at 42 C.F.R. Part 411.
 42 C.F.R. § 411.357(c)(2)(ii).
 42 C.F.R. § 411.357(c)(4).
 42 C.F.R. § 411.351 (See “Referral” definition – (1)(i)).
 Press Release, DOJ (Mar. 11, 2014) (https://www.justice.gov/opa/pr/florida-hospital-system-agrees-pay-government-85-million-settle-allegations-improper).
 The physicians were employed by a hospital staffing agency; however, the parties agreed that whether there was a direct or indirect compensation relationship, the relevant question for any applicable exception was whether the compensation took into account the volume or value of referrals. For purposes of the medical oncologists’ compensation, the court reviewed the arrangement as a direct employment relationship and analyzed the arrangement for compliance with the employment exception. United States ex rel. Baklid-Kunz v. Halifax Hospital Medical Center, et al., No. 6:09-cv-01002-Orl-31TBS, at *14 n. 6 (M.D. Fla. Nov. 13, 2013).
 There are other allegations and findings included in the Halifax case; however, those are outside the scope of this article.
 United States ex rel. Baklid-Kunz v. Halifax Hospital Medical Center, et al., No. 6:09-cv-01002-ORL-31TBS (M.D. Fla. Nov. 13, 2013).
 United States ex rel. Drakeford v. Tuomey Healthcare System, Inc., No. 3:05-2858-MBS (D.S.C. 2013).
 Press Release, DOJ (Oct. 16, 2015) (https://www.justice.gov/opa/pr/united-states-resolves-237-million-false-claims-act-judgment-against-south-carolina-hospital).
 United States ex rel. Drakeford v. Tuomey Healthcare System, Inc., 792 F.3d 364, 379 (4th Cir. 2015).
 Id. at 379.
 Id. at 379.
 Relative Value Units (RVUs) are determined for each service based upon physician work, practice expense and malpractice. Measuring and tracking work RVUS (wRVUs) is a common way to quantify and compare productivity of physicians.
 United States ex rel. Bookwalter v. UPMC, 938 F. 3d 397 (3d Cir. 2019) rev’d en banc, 946 F.3d 162 (3d. Cir. 2019).
 United States ex rel. Bookwalter v. UPMC, 946 F.3d 162, 171 (3d Cir. 2019).
 United States ex rel. Bookwalter v. UPMC, 938 F. 3d 397 (3d Cir. 2019) rev’d en banc, 946 F.3d 162 (3d. Cir. 2019).
 United States ex rel. Bookwalter v. UPMC, 946 F.3d 162 (3d Cir. 2019).
 The case was remanded to the district court and continues to move forward as of the date of this article.
 69 Fed. Reg. 16,054, 16,088 (March 26, 2004).
 84 Fed. Reg. 55,766 (October 17, 2019). This proposed rule has not been finalized as of the date of this publication.
 84 Fed. Reg. 55,766, 55,795 (October 17, 2019) (emphasis added).
 Petition for Panel Rehearing or Rehearing En Banc of Defendants-Appellees at 16, United States ex rel. Bookwalter v. UPMC, 946 F.3d 162 (3d Cir. 2019) (No. 18-1693).
 United States ex rel. Drakeford v. Tuomey, 675 F.3d 394, 407 (4th Cir. 2012).
 66 Fed. Reg. 856, 941 (January 4, 2001).
 69 Fed. Reg. 16,054, 16,088-16,089 (March 26, 2004).
 42 C.F.R. § 411.352.
 42 C.F.R. § 411.355(b).
 42 C.F.R. § 411.352(i).