ERISA Litigation and Physician Autonomy

The federal Employee Retirement Income Security Act (ERISA) looms like a colossus over the managed care environment. Originally enacted to regulate employer-sponsored pension plans, the statute also covers health care benefits established by self-insured employers (with few exceptions, such as for governmental employees). According to recent Department of Labor estimates, ERISA applies to approximately 125 million US citizens.¹

ERISA has created a regulatory vacuum by preempting state regulation of managed care organizations (MCOs) and drastically limiting state medical liability lawsuits against MCOs, while providing minimal federal regulation in their place.^{2 - 4} For physicians, ERISA preemption has indirectly caused courts to favor MCOs' cost containment initiatives over traditional notions of physician autonomy.⁵ The treatment a physician recommends is vulnerable to a managed care utilization management process largely unconstrained by state regulation or liability law, inevitably resulting in reduced physician autonomy.

The consequences of ERISA preemption lie at the heart of proposed congressional patients' rights legislation that would restore the primacy of the patient-physician relationship and permit state legal challenges to cost containment programs. At issue is control over physicians' clinical decisions and the ability to challenge improperly operated cost containment programs.

Because ERISA plays a vital role in the relationship between physicians and MCOs, it is important for physicians to understand what ERISA is, how it operates, and how it influences clinical decision making and physician autonomy in the managed care era. In this article, we outline ERISA's major provisions, analyze trends in ERISA litigation applicable to physicians, and conclude by discussing the policy implications and significance of these trends for physician autonomy.

Congress enacted ERISA in 1974 primarily to regulate pension plans, but it also included health benefit plans within its scope.⁶ ERISA's goals are to establish uniform national standards, safeguard employee benefits from loss or abuse, and encourage employers to offer those benefits. To achieve these objectives, ERISA imposes strict requirements on pension plan administrators for reporting and disclosure,⁷ participation and vesting,⁸ funding,⁹ and performance of fiduciary obligations.¹⁰ ERISA does not mandate that employers offer benefit plans, but provides a structure for national uniformity of administration once such plans are extended.

Only a few of these requirements apply to health benefit plans, in part because Congress did not pursue the implications of regulating both pension and health benefit plans under 1 statute. Nor could Congress have anticipated the dominance of the managed care model. As a result, ERISA provides almost no federal regulation of health plans. Nevertheless, ERISA has 3 provisions that directly affect physician autonomy: ERISA's preemption clause, its limited remedial scheme, and its fiduciary duty obligation.

Traditionally, states are responsible for regulating health care delivery, and litigation against health care providers is resolved under state law. Medical liability lawsuits are rarely heard in federal courts. ERISA alters the traditional approach by preempting state law, which means that state laws purporting to regulate health plans may not be enforced in any court.¹¹ In this context, state laws include legislation and regulations, such as those mandating particular benefit coverage, and most medical liability actions targeting MCOs.

Courts have interpreted the preemption clause broadly to prevent enforcement of state laws ranging from laws protecting the patient-physician relationship to litigation challenging how cost containment initiatives are implemented. The courts have held that Congress intended such broad preemption of state law to allow a multistate employer to offer a single, nationally consistent plan to all its workers without the cost and inconvenience of complying with contradictory state regulations, legislation, or litigation. National uniformity conforms with congressional intent to keep the costs of administering an employee benefit plan (EBP) low to encourage employers to offer health care coverage.¹²

In assessing whether a particular state law is prohibited, courts look sequentially to each of the 3 parts of the preemption provision. First, courts must decide whether the state law "relates to" an EBP.¹³ In doing so, courts consider whether the challenged law burdens the administration of plan benefits or has only a remote impact on them. Courts generally hold that ERISA preempts state laws that bind employers or plan administrators to particular benefit choices or that preclude the uniform administration of an EBP.¹⁴ For example, a state lawsuit challenging a benefit determination, such as an MCO's denial of additional hospital coverage, "relates to" a health plan because that challenge would require the court to interpret the plan's benefits, hence binding the administrator to certain actions. But laws with only a remote or incidental effect on plan administration, such as a surcharge on hospital services, may not "relate to" the EBP.

A law is not preempted merely because it "relates to" a plan. Courts must also interpret 2 qualifying provisions, the savings clause and the deemer clause. ERISA's savings clause provides that laws regulating the business of insurance, even if they "relate to" a managed care plan, will not be preempted. This allows states to continue to enforce state laws governing the business of insurance by saving state regulation of health insurance, such as solvency requirements, from preemption.

In turn, the deemer clause qualifies the savings clause. The deemer clause prevents states from deeming (or characterizing) an ERISA-covered plan as the business of insurance. States may not characterize a self-funded plan as an insurer to circumvent the effect of the "relates to" clause.

As an example of how these terms interact, consider a state law mandating certain health insurance benefits. That law "relates to" an ERISA plan since it would involve the structure of plan benefits. Even though the legislation would be saved from preemption insofar as it regulates EBPs that purchase traditional insurance policies, it would still be preempted if, for example, a state attempted to apply the statute to a self-funded EBP. Under the deemer clause, a self-funded EBP cannot be an insurer.

Although much state litigation is preempted, ERISA provides some relief for injuries to health plan participants through its civil enforcement scheme. A plan participant or beneficiary may bring a civil action against an administrator who fails to comply with a request for information about the plan, to recover claimed benefits, to enforce rights under the terms of the plan, or to clarify rights to future benefits.¹⁵ A plan participant may also bring suit against a plan fiduciary who breaches any fiduciary duties and may seek to enjoin practices that violate ERISA or the terms of the plan.

Even if victorious, a plan participant can usually only recover the amount of the benefits that should have been provided, as well as certain incidentals such as attorneys' fees. This is a decidedly more limited remedy than what is usually available under state law, through which the patient might be able to recover damages for any economic losses, noneconomic damages for pain and suffering, and possibly punitive damages (especially in cases alleging bad faith insurance denial).¹⁶

Take, for example, a challenge to an improperly denied benefit filed in state court. If the MCO successfully invokes preemption, the plaintiff will be forced to sue instead under ERISA's limited civil enforcement scheme. Effectively, this insulates the MCO from exposure to monetary damages, except for what it would have paid (the amount of the denied benefit) in the first place.

ERISA imposes a fiduciary duty on those who make discretionary decisions on behalf of the EBP. A fiduciary must discharge his/her discretionary functions "solely in the interest of the participants and beneficiaries" of the plan.¹⁷ In many, but not all cases,¹⁸ courts have held that MCOs are subject to this fiduciary duty when making certain decisions, such as reviewing the appropriateness of a physician's treatment recommendations.¹⁹ On the other hand, MCOs and employers are not considered fiduciaries for establishing or changing the terms of the plan. Thus, the fiduciary duty extends only to decisions made once the plan is in place.^{20 - 21} Employers must provide whatever health benefits they promise but need not offer plans at all and can change what they offer after giving plan beneficiaries proper notice.

In exercising the fiduciary duty, an obvious problem is that the clinical needs of one patient may conflict with the MCO's economic interests. Increasingly, disappointed plan participants have sued for breach of fiduciary duty, often challenging the denial of physician-prescribed benefits, especially when there is a potential conflict of interest.²²

To determine whether an MCO breached its fiduciary duty when denying plan benefits (ie, that the denial is not solely in the interest of the participant), courts use different levels of scrutiny based on the amount of discretion granted to the MCO under the EBP. Generally, courts are very deferential, upholding the plan administrator as long as the decision was not arbitrary and capricious.^{23 - 24} In most cases, courts have equated compliance with the terms of the EBP as, by definition, acting "in the interests" of the plan participant. In this sense, the court limits its review to ensuring that the MCO reasonably comported with the terms of the EBP.²⁵ As a result, MCOs retain power vis-a-vis physicians by controlling the interpretation of EBP terms (including medical necessity). But in a case from which the plan profits directly from the denial, the potential conflict of interest must be considered a factor in deciding whether there was an abuse of the fiduciary's discretion.²⁶

For many years, courts have monitored quality of care through medical liability lawsuits. Originally, physicians were the targets of such suits, then hospitals were added, and now MCOs have been held liable under state tort law.

When a patient receives care under a health plan not governed by ERISA (as when a person buys his/her own health insurance), MCOs have been held directly liable for their own actions, such as the failure to maintain safe and adequate facilities, select and retain competent physicians, oversee all patient care within the institution, and ensure quality care.²⁷ Managed care organizations can also be held vicariously (indirectly) liable for malpractice committed by physicians who are independent contractors. The primary factors affecting whether a court will impose vicarious liability include the amount of influence the MCO has over the clinical decision, patients' perceptions of the relationship between the physician and the MCO, and the manner in which the health plan is marketed. Managed care organizations operating in contexts other than ERISA may also be subject to state consumer protection or bad faith insurance laws for improper processing of claims that results in delayed or denied care.¹⁶ In non-ERISA cases, courts are essentially following the pattern of establishing liability that was applied to hospitals beginning in the 1960s.²⁸

Managed care organizations covered by ERISA operate under different rules. ERISA preempts many state law claims alleging that the MCO's denial or delay in care caused an adverse medical outcome. Those types of lawsuits may be brought in federal court as actions under ERISA's civil enforcement scheme, but the limited remedies available effectively insulate MCOs from liability and, therefore, accountability for these medical outcomes. Liability may be borne instead entirely by physicians. The practical effect is that MCOs often control resource allocation, but physicians (and patients) bear the costs when resource allocation decisions produce adverse outcomes. In these cases, the patient's only remedy is to sue the physician, regardless of how much influence over the clinical decision the physician actually exercised.

Courts have not been consistent in deciding whether all or merely some state law claims against MCOs will be preempted. Until recently, led by early Supreme Court doctrine, lower federal courts have interpreted the phrase "relates to" very broadly, preempting most state law tort suits challenging health plan innovations and medical decisions. For example, courts generally have held that challenges to delayed or denied care relate to an EBP and are preempted,²⁹ including litigation alleging that the structure of the EBP was responsible for poor medical outcomes. But in New York State Conference of Blue Cross and Blue Shield Plans v Travelers Insurance Co,³⁰ the Supreme Court permitted New York State to impose a tax on all insurers except Blue Cross and Blue Shield, reasoning that a uniform tax only tangentially relates to ERISA plan administration. This decision signaled a scaling back on the breadth of preemption. After this decision, courts have been less vigorous in finding ERISA preemption.

Narrowing preemption has inspired other related changes. The most important change is that courts have erected a critical distinction between state tort law challenges to the technical quality of care (ie, liability claims for substandard clinical care) and state law challenges to the quantity of care (involving improper plan benefit decisions). The latter must be brought in federal court subject to ERISA's limited remedies; the former would be heard in state court.

In practice, the quantity/quality distinction may signal a nascent trend toward holding MCOs accountable at least in some circumstances, especially, if courts strain to characterize a dispute as involving quality. By way of example, the court in Bauman v US Healthcare, Inc³¹ recently held that the defendant's policy of discharging a newborn within²⁴ hours without adequately considering the medical appropriateness in a given case could be challenged in state court as substandard quality of care. As their liability expands, MCOs may begin to reconsider the ways in which they review clinical decisions, as the United Healthcare decision³² signaled recently by shifting greater clinical authority back to physicians. It is one thing to deny treatment when potential liability rests with the treating physician, but it is another to deny the claim when the organization might also be held responsible.

Consider, for example, state litigation seeking to hold an MCO indirectly liable for the actions of an affiliated physician. Because substandard care is litigation about the quality of care and not the quantity of benefits, the case will probably be heard in state court. And because state courts assess liability based in part on the amount of influence the MCO exerts over clinical decision making, MCOs may seek to avoid liability by loosening their control below the threshold required by state law. The result is increased physician autonomy. While this result is what we anticipate, greater liability exposure could alternatively lead an MCO to protect itself by exerting stricter oversight of clinical decisions. Exactly how MCOs respond will need to be studied.

Preemption is not the only area in which this trend toward judicial reconsideration may be emerging. In recent years, many lawsuits charging MCO misconduct that resulted in adverse outcomes from delayed or denied care have been cast as breaches of fiduciary duty under ERISA.³³ No patient has yet recovered a judgment in such a case³⁴ ; but to the extent such suits are successful, MCOs might be less likely to second-guess clinical decision making. To date, the decisions have been inconsistent and no truly coherent doctrine has yet emerged. Most courts have explicitly refused to be the agents of a major overhaul of ERISA doctrine, preferring to leave such a role in the legislative arena.⁵

A central aspect of the managed care environment is the emergence of new organizational forms, including utilization management processes, which have mixed clinical and financial functions. Managed care organizations rely heavily on utilization management techniques, such as preauthorization for high-cost medical interventions, to reduce costs. The more courts uphold utilization management decisions, the less control the treating physician has over the clinical encounter.

In state cases for which ERISA does not apply, courts have held generally that physicians and MCOs may share liability for bad outcomes.^{35 - 36} By contrast, ERISA preemption clearly shields MCOs from liability in state courts for utilization management decisions, even when these are arguably medical and not merely administrative in nature. So far, federal courts have uniformly held that utilization management decisions relate to benefit plans and are preempted, regardless of whether medical care recommended by the treating physician is denied. For instance, in Danca v Private Health Care Systems Inc,³⁷ the court supported the prevailing view that a utilization review dispute was preempted by ERISA because it is part of the process used to assess a benefit dispute.³⁸ And in Corcoran v United Health Care Inc,³⁹ the court concluded that United's utilization management program, whose denial of hospital care resulted in the death of a fetus, made medical decisions in the context of determining benefits. Accordingly, the court preempted Corcoran's lawsuit under ERISA. By holding that the administrative aspect of the utilization management process trumps the medical aspect (ie, that it is more a quantity than a quality decision), the federal courts, through ERISA, provide wide latitude for health care plans to control costs, at the possible expense of both individual access to health care services and the treating physician's clinical autonomy.

As part of the backlash against managed care, many state legislatures have tried to safeguard physician autonomy. This legislation has ranged from prohibiting gag clauses to comprehensive reforms designed to limit the primacy of cost containment strategies. In many instances, courts have ruled that these laws are preempted by ERISA, although the decisions are by no means uniform. Such rulings have essentially negated state legislative attempts to restore physician autonomy and have reinforced health plan control over clinical decisions.

The most extensive attempt to regulate MCOs is the Texas statute⁴⁰ requiring an external appeals process for health care denials and allowing subscribers to sue the MCO for poor quality of health care. A federal district court recently upheld the right to sue, based on the quality/quantity distinction, yet overturned the external grievance process as preempted by ERISA.⁴¹ Although the case is on appeal, that the court preempted the external review process as a law "relating to" an EBP is an indication of the ERISA-created hurdles facing state laws that try to bolster clinical autonomy.

Just as troublesome, the current uncertainty in ERISA litigation makes it difficult, if not impossible, to predict which state laws will be preempted. As an example, courts have split on whether any willing provider laws are preempted by ERISA.^{42 - 43} Any willing provider laws would require MCOs to contract with any provider willing to meet the MCO's established criteria and are intended to preserve patient choice of physician.

Another important cost containment mechanism used by MCOs is to provide financial incentives to plan physicians to restrain costs. For instance, salary withholds and bonuses are used as compensation incentives for limiting referrals to specialists and other high-cost procedures. No court has yet ruled that these financial incentives violate public policy, though some non-ERISA cases have permitted challenges to be tried before a jury.⁴⁴

ERISA does not regulate how MCOs create incentive structures to motivate contracting physicians' compliance with cost containment measures. More importantly, ERISA preemption may prevent states from trying to regulate such compensation and incentive arrangements through tort law or legislation. A typical case is Lancaster v Kaiser Foundation Health Plan of Mid-Atlantic States Inc,⁴⁵ in which the court held that the plaintiff's state law claim alleging negligence in establishing and operating an incentive program that encouraged physicians not to prescribe certain expensive tests and not to refer to specialists, was preempted by ERISA. The plaintiff claimed that this program was a substantial factor in her physicians' failure to diagnose her brain tumor for 5½ years until it had invaded 40% of her brain. The court characterized the establishment and operation of this incentive scheme as an administrative decision affecting the provision of benefits and therefore dismissed the claim as preempted.

Not all courts have agreed.⁴⁶ A more recent case perhaps presages a different direction based on breach of ERISA fiduciary duties (with, of course, the corresponding limit on remedies under ERISA). In Herdrich v Pegram,⁴⁷ the court held that a patient could sue for breach of fiduciary duty based on an allegation that the nature of incentive arrangements between the MCO and the physicians caused her to be deprived of proper medical care and that the MCO reaped economic gain from this deprivation. Even though the Herdrich court specifically noted that the existence of economic incentives would not automatically be tantamount to a breach of fiduciary duty, this case is a potentially significant extension of the rationale advanced in non-ERISA cases. If read broadly and followed by other courts, this case could augur an attack on the underlying financial incentives at the core of managed care, perhaps by seeking to enjoin their use.⁴⁷ However, Herdrich may represent a legal theory that is viable only in an extreme case in which "a fiduciary jettisons his responsibility to the physical well-being of the beneficiaries in favor of loyalty to his own financial interests,"⁴⁶ and the Supreme Court has agreed to review the decision.

ERISA has played an important role in facilitating, and perhaps stimulating, the development of managed care. But this undeniable policy benefit has come at a high cost to some individual plan subscribers and to physicians. From a policy perspective, ERISA has created a regulatory vacuum in which states cannot act and there is no comparable federal regulatory mechanism. From a legal perspective, ERISA has essentially insulated MCOs from liability by blocking state courts from resolving litigation challenging managed care practices. From a clinical perspective, ERISA has facilitated reductions in physician autonomy relative to health plan influence over clinical decisions.

This analysis suggests several consequences of importance to physicians. First, the effect of judicial interpretations of ERISA is to subordinate physician autonomy and the patient-physician relationship to managed care cost containment goals. Implicitly, ERISA reinforces the status quo of the health care delivery market and hence managed care's current market domination.

Second, perceiving themselves bound by ERISA, courts do not champion either physician autonomy or the patient-physician relationship. Courts that once protected physician autonomy, are no longer doing so.⁴⁹ Judges repeatedly suggest that complaints against managed care should be taken to the legislative branches of government rather than to the courts. Since state legislative initiatives are often barred by ERISA preemption, Congress appears to be physicians' best hope for change or relief. As of this writing, congressional action to amend the preemption provision (so that state legislatures and courts may act with fewer constraints) or to create new federal regulations similar to recent state initiatives appears to be unlikely. During the 1999 session, the House of Representatives enacted a bill that would permit patients to sue MCOs for damages in state courts, but the Senate bill does not contain a right-to-sue provision. Although House and Senate conferees have been meeting to reconcile the 2 bills, the prospects for enactment are slight, in part because the House conferees are largely opposed to the right-to-sue provision.

Regardless, physicians should continue to support attempts to remove ERISA preemption. Even though treating physicians would remain accountable, MCOs should also be held accountable for both financial decisions that affect clinical treatment and for their implicit role in making medical decisions. Physicians should not be left in the untenable position of being entirely responsible for cost containment provisions over which they have almost no control. By exposing MCOs to similar liability considerations, MCOs will not be able to influence medical decisions with impunity, which may enhance physician autonomy.

Short of eliminating ERISA preemption, Congress could also amend ERISA by expanding on the available remedies. Consistent with the goal of maintaining national uniformity, Congress could retain preemption but allow individuals to sue for monetary damages in federal court for an ERISA violation. Congress could also direct the US Department of Labor (as the appropriate regulatory agency) to develop regulations that would more effectively protect the patient-physician relationship. For instance, regulations might address patients' rights to notice of a denial of care and to an external grievance panel.

An assumption animating this article is that deference to physician autonomy is a desirable goal for better patient care. To some, that proposition may not be self-evident. After all, public concern with rising health care costs and perceived harms from overtreatment in the fee-for-service era led directly to managed care's cost containment innovations and concomitant restrictions on physician autonomy as public policy objectives. Indeed, constraints on physician autonomy predate the effects of ERISA litigation.⁵⁰

Thus, one scenario suggests that if Congress removes ERISA preemption and helps restore physician autonomy, managed care's cost containment goals may be difficult to achieve. This outcome seems unlikely given the current policy environment. Another possible scenario is that MCOs might respond to eliminating ERISA preemption by imposing more aggressive utilization management controls, ironically reducing physician autonomy below what it is under current financial incentives. Yet there is no indication that MCOs are eager to accept the additional liability consequences that may result from greater control over clinical decisions.

No matter whether Congress changes ERISA preemption or expands ERISA's limited remedies, the tensions among physicians, MCOs, and patients will not be resolved easily. Even if there are some downsides to physician autonomy, patient care ultimately depends on the treating physician's ability to maintain patient trust while balancing patient demands for high-quality care with the MCO's legitimate cost containment efforts. In truth, these cost containment programs are needed corrections to an unsustainable fee-for-service system, and many physicians have worked effectively to mediate managed care's constraints. Yet, inevitably, some patients will not be well served by this system. Either there will be undue delay in arranging health care, or benefits will be denied that should have been provided. In those situations, holding MCOs legally accountable provides incentives for better health care plan administration. ERISA preemption simply goes too far in removing the liability threat for improper undertreatment.