Integrated Health Care, Capitated Payment, and Quality: The Role of Regulation

Toward Universal Managed Care

Managed care has a long tradition in the United States, but the passage of the Health Maintenance Organization Act of 1973 boosted the “managed care industry” [14]. Health maintenance organization market share slowly expanded during the 1970s and the first half of the 1980s. Physician decision making in health maintenance organizations had to be cost-effective because the organizations were at risk for costs of care [6]. Lower cost per insured life provided an economic advantage for managed care.

The efficacy of techniques developed by traditional health maintenance organizations was not lost on the rest of the health insurance industry. Cost savings generated by traditional utilization review techniques had started to plateau by the mid-1980s, and insurers began to look for alternative ways to reduce costs. In particular, insurers began to organize physicians into networks. So-called preferred provider organizations evolved rapidly, leading to a blurring of distinctions between the various institutions that had previously been referred to as the managed care industry [15, 16]. For the sake of clarity, we refer to health maintenance organizations, preferred provider organizations, and point-of-service plans that formally underwrite health care risk as managed care organizations.

With the start of the 1990s, another period of change has begun [13, 17]. As care management matures, fewer services are sought from specialists, and hospitals need larger primary care bases to feed their existing hospital-based secondary and tertiary care systems. This situation has caused greater horizontal and vertical integration of health care, with hospitals, physicians, and (in some cases) managed care organizations coming together into one cohesive structure in an increasing number of metropolitan areas [18].

The primary movers of the new care systems are frequently hospitals that fear losing their referral bases and being eliminated by existing managed care organizations. They recruit physicians, often by acquiring group practices, and try to develop vertically integrated working structures. Many of the physicians in these arrangements are salaried by or have exclusive contracts with the hospitals, allowing the hospitals to take on capitated risk for patient care [19]. As such, the systems emulate the ability of the managed care organizations to control costs by managing physicians through utilization management techniques, practice guidelines, critical pathways, and other care protocols [20-22]. In some metropolitan areas, managed care is becoming universal [23].

The relation of vertically integrated care structures to existing insurers and managed care organizations is complex. In some cases, the integrated plans compete directly with managed care organizations, especially for primary care patients. In other cases, integrated delivery systems may still bid for patients from managed care organizations, specifically for hospital services.

Financial Arrangements of Managed Care

The question facing policymakers, health care providers, and patients is, With all of the financial incentives imbedded in managed care and integrated systems of health care delivery, how can we ensure that quality of care is maintained and that physicians can be appropriate agents for their patients? Reiteration of professional ethics is one approach, but calls to reject all market values in health care are unrealistic [12]. Rather, the market should be constrained by rational regulation. To evaluate how well different regulatory policies work, we first need to understand the financial incentives that managed care organizations and integrated delivery systems now use to encourage physicians to be cost conscious.

Managed care organizations and the emerging integrated delivery systems select from several basic financial arrangements when contracting with physicians, but two are most important: salary and capitation. Salary was the traditional payment method used by staff-model health maintenance organizations. Salaries have long been supplemented by bonus arrangements intended to encourage greater productivity among physicians.

Capitation-based payment to physicians in managed care organizations is the alternative and increasingly prevalent method for structuring the financial arrangements between physicians and the managed care organizations. Within capitation, various subarrangements can exist. In the simplest arrangement, an individual physician receives a certain amount of money per month for each member of the managed care organization who designates the physician as his or her primary care physician. In a variation on this arrangement, a group of physicians might contract with the managed care organization and receive a certain amount of money per month for each member of the managed care organization who designates the group as his or her site for receiving medical care.

Another strategy is to hold back part of the capitated payment until the end of the year; this is often known as a “withhold” or a “risk withhold.” The money in the withhold account is similar to the reserves required of insurance companies and banks: It is a mechanism for ensuring that the physicians have enough money to pay for the care of the insured lives in their care. Of course, if all of the money in the withhold account is exhausted, the management of the managed care organization is at risk for additional expenditures. If money remains in the withhold account at the end of the year, the physicians receive it as a lump sum. This is often viewed as a bonus arrangement.

Capitation payments for specialists and inpatient hospital care are arranged in various ways. In some contracts, the managed care organization retains percentages of the premium that are used to pay for specialists, prescription drugs, and inpatient hospital use. Alternatively, the managed care organization may pay a higher percentage of the premium as the base capitation payment to the primary care physician, forcing the primary care physician, in turn, to pay for specialists, prescription drugs, and hospital care. Specialists may be paid on reduced fee schedules or may receive flat consultancy retainers (equivalent to a salary for being on call for a certain number of hours per month). They may even be paid capitated amounts for being available for a panel of insured lives.

The few investigations of the financial incentives used by managed care organizations suggest that these incentives do reduce rates of hospitalization and outpatient visits [24]. Analyses of the quality of care in prepaid practices have not shown systematic deficiencies [25, 26]. Miller and Luft [27], however, have cautioned that the study of quality in managed care has been relatively rudimentary and that careful investigation of mature systems of care should be a policy priority.

State Regulation of Managed Care

Management of care has inherent ethical problems, because incentives to provide low-cost care may be created without sufficient concern for patient welfare [28]. Physicians can be responsible for much of the risk that actual medical expenses will exceed expected expenses for a panel of patients. It follows that economic gains, or prevention of major losses, could be made at the expense of appropriate care. A physician at risk for all ambulatory care could become too parsimonious as reserves dwindle because of unexpectedly sick patients. Some have suggested the appropriate ethical response to such pressures, but many states have also promulgated regulations to protect quality of care.

Regulation of managed care organizations by state governments is generally intended to protect patients by providing them with information and by ensuring the fiscal integrity of the organizations. Most states have long required traditional managed care organizations to obtain specific licenses. The licensing authority, often the insurance commissioner, can regulate in all areas not preempted by federal law. States vary considerably in their regulatory approach, however. The Maryland provisions [29] are an example of an approach that encourages managed care organizations to provide high-quality care. Maryland requires managed care organizations to have regular hours; to provide for 24-hour access to a physician; to ensure that each member seen for a medical problem is evaluated under the direction of a physician; and to give each member an opportunity to select a primary care physician. Maryland also insists that managed care organizations provide appropriate preventive services and periodic health education.

Maryland law also mandates that each managed care organization 1) have a written plan about its implementation of these standards of quality of care and 2) review statistics on the health care needs of its patients. Each organization must have an internal, professional peer review system that in many ways is modeled on peer review in hospitals. The systematic collection of data on performance and patient results is also required, and managed care organizations must provide all members with information on available services and on potential responsibility for payment of services by members.

California, which may have the most advanced managed care industry, also has the most advanced regulation. In addition to having rules similar to those in Maryland, the California state health authorities may do on-site surveys of plans to evaluate peer review and referral mechanisms, internal procedures for oversight of quality of care, and overall performance [30]. Recently, California authorities fined a managed care organization $500 000 for failing to follow its own procedures in a patient's referral for expensive oncologic care [31].

Some states are antagonistic toward managed care organizations. Their regulations generally reflect this antagonism, often in part because of the interests of local medical societies. For example, in Texas, protections against insolvency are much more thoroughgoing (some would say oppressive) than those in other states [32, 33].

Texas also has a strong “any-willing-provider” law. The prototypical any-willing-provider statute requires that managed care organizations allow any individual or institutional providers who are willing to meet a plan's financial and educational criteria into their provider networks. Such laws make it more difficult for managed care organizations to maintain the cohesive panel of practitioners that is central to reducing utilization, and they reduce the organization's bargaining power in negotiations with providers.

At least 27 states have now passed some species of the any-willing-provider statute [34]. Some statutes require notice to all providers of the opportunity to become network participants, creating large expenses for managed care organizations [35, 36]. Other statutes create due process-based appeals structures. Some statutes insist that patients be able to see out-of-network providers.

Any state regulation of managed care organizations has two major weaknesses. First, such regulation generally applies only to licensed health maintenance organizations and not to all integrated delivery systems. Most states fail to recognize that managed care is spreading beyond the traditional managed care organization. Second, in many circumstances, state regulations may be preempted by the federal Employees Retirement Income Security Act (ERISA), which creates immunity from state oversight for beneficiaries of self-insured health plans. Recent cases brought by managed care organizations suggest that any-willing-provider laws and other state regulation of managed care organizations may be illegal under ERISA [37].

Federal Regulation of Managed Care

The American Medical Association has endorsed the federal Patient Protection Act [38]. This proposed legislation incorporates much of what is found in any-willing-provider laws at the state level; it would require contracting between managed care organizations and any licensed provider who meets reasonable predetermined standards [11]. Further, under this legislation, managed care organizations would be required to provide members with all terms and conditions of the health plan in easily understood, truthful, and objective terms. Physician credentialing within the health plans, particularly credentialing issues based on economic factors, such as the physician's past expenditure per patient, would also be subject to oversight. Case mix, severity of patient illness, and patient age adjustments must be incorporated into such credentialing. Finally, the proposed Patient Protection Act states that all enrollees of health benefit plans are to be offered a selection of plans, including at least one traditional fee-for-service plan. The Patient Protection Act and the any-willing-provider laws of various states can be characterized as important efforts to ensure that cost-quality tradeoffs under managed care programs are not detrimental to patient care. They can also be characterized as carefully crafted efforts by the medical profession to retain certain economic prerogatives at the expense of managed care organizations.

The federal government now limitedly oversees managed care techniques [39]. For example, under its authority to police the use of incentives in the Medicare program, the Inspector General's Office in the Department of Health and Human Services [40] recently proposed rules about bonuses. Unfortunately, this proposed regulation is vague, perhaps indicating the problem with governmental oversight of care. Other federal laws that affect managed care include the tax code for nonprofit organizations, anti-kickback provisions in antitrust law, and the Stark self-referral provisions [33]. These laws are especially important in affecting the way vertically integrated organizations are structured [34]. However, little coordinated strategy currently exists at the federal level, as is best shown by the rush to approve waivers for Medicaid managed care [41]. Moreover, the preemptive effect of ERISA could hamper state-based regulation.

Self-Regulation of Quality by the Managed Care Industry

Managed care organizations are now creating a self-regulation framework similar to that undertaken for hospitals by the Joint Commission for Accreditation of Health Care Organizations. In 1979, the Group Health Association of America and the American Managed Care Review Association, the major trade associations of the managed care industry, founded the National Committee for Quality Assurance. In 1990, this committee was spun off as an independent nonprofit organization to operate an accreditation program that reviews all types of managed care organizations [42].

The National Committee for Quality Assurance has formulated a series of accreditation standards that focuses on six areas: preventive health services, medical records, utilization management, members' rights and responsibilities, credentialing, and quality improvement. The quality-improving aspects of accreditation are based on a list of performance measures known as the Health Plan and Employer Data and Information Set (HEDIS) [43]. The specific measures were selected on the basis of three criteria: the value and relevance of the measure to employers; the ability of health plans to provide the requested data; and the potential the measure has for improving quality of care.

The specific components of HEDIS are quality, access, patient satisfaction, membership, utilization, and finance. Each of these components contains specific subcriteria. For quality, HEDIS concentrates on preventive medicine, prenatal care, acute and chronic disease, mental health, and substance abuse. Under preventive medicine, for example, health plans must report rates of childhood immunization, rates of cholesterol screening for particular populations, rates of mammography screening for women older than 40 years of age, and rates of Papanicolaou tests for women 15 to 70 years of age. For each measure, the health plan must calculate rates by using the specific denominators prescribed by HEDIS.

The National Committee for Quality Assurance is committed to publicly sharing at least some information. In the fall of 1993, it launched a Report Card Pilot Project involving 21 health plans. The Committee's intention is to produce a report card for each participant on the basis of selected HEDIS measures. Independently, some plans have already begun to share their scores on report cards [44]. The commitment to public disclosure is laudable, but some question whether the HEDIS measures are sufficiently rigorous.

The HEDIS criteria have their critics [45]. None of the measures is risk adjusted, and only two measure true outcomes. The rest focus on administrative matters or relatively vague process issues. Their effect on quality of patient care is unknown.

Some Proposals

The potential for quality deficiencies in a heavily capitated health care system is real and must be addressed. We think it is appropriate for states to enact new, carefully targeted regulations that would complement existing controls and support the ethical commitment of physicians to patients without retarding the growth of managed care.

1. Broadened Oversight of Managed Care

At present, managed care oversight applies only to organizations that seek specific licenses from the state. The touchstone is usually formal risk underwriting, and many integrated delivery systems do not underwrite risk. Each state should ensure that its rules about financial integrity, clarity of contracts with beneficiaries, and reporting of quality assurance measures apply to all fully integrated plans. We also support any ERISA waivers necessary to ensure that state authority in these areas is not preempted. Regulators must come to grips with the fact that managed care mechanisms, including capitation, are becoming universal as we move toward oligopolistic competition between integrated delivery systems. Oversight cannot be limited to traditional managed care organizations.

2. Oversight Committees for Integrated Delivery System Care

These committees would consist of providers, administrators, and patients who would meet at least every other month to review new, cost-effective care measures proposed by the integrated delivery system. Any new incentives, care protocols, critical pathways, or guidelines would have to be studied by these committees. The main purpose of the committees would be to protect patient welfare, and each committee would be charged with developing specific criteria for use in judging protocols and guidelines. If a proposal did not pass review, it would not be allowed into practice. We would let each state determine how committees are to be constituted. Some might want to have elections, such as those for school board or judiciary positions; others might prefer selections made through local medical societies, hospital boards of trustees, and interested citizenry. Public membership and open meetings are critical to the concept of the care oversight committee.

This form of regulation is local and decentralized. The care oversight committee would, however, be charged with maintaining careful records and detailed minutes. These could be available to a public authority within the state without notice. In this regard, the care oversight committees would be similar to institutional review boards, which are local in their operation but which must be open to inspection by the Department of Health and Human Services or the Food and Drug Administration.

The internal review boards are a proven method for giving voice to ethical concerns in medicine while opening up deliberations to the public. We believe that professional ethics have an important role to play in promoting quality. The oversight committee is a vehicle through which the altruistic commitment of the provider to the patient can affect organizational policy.

3. Public Disclosure

Several states currently require hospitals to report health care data. For example, in the late 1980s, the New York State Department of Health began to report risk-adjusted, heart surgery-related mortality rates [46]. In addition to ranking hospitals, New York rates individual heart surgeons on adjusted mortality rates. There is reason to believe that the Cardiac Surgery Reporting System is having a positive effect on quality [47]. Forty-two of the 50 states either have implemented or plan to implement some sort of system for collecting and disseminating health care data. This effort should be extended to integrated delivery systems. Measures such as compliance with common measures such as the HEDIS criteria, patient satisfaction, and health status should be made available to the public after the best applicable clinical severity adjustments. More importantly, each delivery system should specify both the measures it uses for internal benchmarking and its full panoply of incentives (some states already require this).

4. Regulation of Capitation

We believe that it is prudent to control capitated risk. Small units of providers should not be at full risk for care of patients. If a group were to be impoverished by one very sick patient, the care of many other patients could be affected. Thus, we favor limiting the dollars of care per patient for which a care unit is responsible. We would lodge the authority to define such a limit with the office of each state insurance commissioner.

We prefer a limit that is tied to the types of care for which a care unit is at risk, the number of primary care physicians in a care unit, and the number of members who have chosen the unit as their source of primary care. Thus, a large group of physicians (say, 100 or more physicians in a multispecialty group) with a combined responsibility for perhaps 15 000 member lives has a risk pool that is large enough to sufficiently dilute the risk that is posed if any one patient has excessive medical care expenditures. But a group of 4 physicians with a combined panel of 1500 member lives should not be responsible for the same level of risk. Limits could be handled by stop-loss or reinsurance mechanisms; some managed care organizations already use these methods.

5. Avoidance of Radical Reform

We favor neither the passage of sweeping any-willing-provider laws that would poison the entire concept of managed care nor the reinstitution of fee-for-service insurance. Integrated delivery systems can provide ethically sound care, especially if some of the mechanisms listed above are put into place. We do not believe that government efforts to regulate managed care directly (for instance, through oversight of practice guidelines) can possibly work. The task is simply too large. The mix of limited state oversight, such as we have suggested, and quality competition for the business of employers with knowledgeable benefits managers, is appropriate.

Conclusions

We do not believe that professional ethics alone can cure the moral conflicts that threaten quality of care in an integrated health care market. Judicious use of regulation that blunts some of the incentives in capitated financing; that gives consumers information; and that allows physician, nurse, and patient oversight of care protocols is appropriate. The need to develop these regulatory mechanisms is urgent, because the integration of health care is proceeding quickly, and the good of the patient is not always the central concern in today's health care market.

Acknowledgments: The authors thank Joseph Dorsey, Richard Frank, Nancy Kane, Joseph Newhouse, Lynn Peterson, and Lois Snyder for helpful advice.