Managed Care Research Paper

1. Historical Context

The many aspects of managed care—and the debates over it—can only be understood in the context of the US health care system. Although some aspects of managed care may be applicable outside the US, attempts to transfer specific lessons need to be made with care. The United States is an outlier among highly developed countries because of its very high expenditures per capita, substantial proportions of the population without insurance coverage, and less than optimal health status (Anderson and Poullier 1999, National Center for Health Statistics 1999, Vistnes and Zuvekas 1997), reflecting the absence of a national health care delivery system or policy. To examine the development of, and the reaction to, managed care, it is necessary to go back to the period shortly after World War I. At that time, medical care was fee-forservice in its literal sense. Billing was uncommon, and there was no ‘third party’ between physician and patient. If the patient could not pay, care might not be obtained, unless the physician voluntarily provided care. This fit well the economist’s description of medical care as a cottage industry. Physicians tended to practice independently, or with a partner or two.

The development of health insurance plans during the Depression brought a third party into the picture. Indemnity insurance plans reimbursed the patient a fixed amount for services received. Any fees in excess of the indemnification, as well as the paperwork, were the responsibility of the patient. Blue Cross and Blue Shield plans—not-for-profit organizations sponsored by the state hospital and medical associations— provided what is known as a service benefit. They contracted with physicians and hospitals to accept their reimbursement as ‘payment in full.’ For hospitals, these were literally reimbursements—the Blue plan reimbursed them for their share of total costs. For physicians, the payments were generally based on usual, customary and reasonable fees. Nonetheless, physicians were often uneasy about the involvement of a third party in the doctor–patient relationship.

Enrollment in these health insurance plans was voluntary, and in the classic insurance model, services that could be anticipated, such as maternity care, wellbaby checkups, and physical exams, were not covered. Deductibles and coinsurance provided some brake on the use of services, but there was often an out-ofpocket cap to prevent a patient from being bankrupted by hospitalization or a major illness (Starr 1982).

The rapid expansion of employer-sponsored health insurance after World War II brought the entrance of a fourth party—the employer—into this picture. In most instances, the employer merely paid part or all of the premiums for the health plan it chose to offer, and health plans varied largely in terms of the breadth of coverage and the extent of deductibles and coinsurance. Employers continued to be passive payers of premiums until recently, seeing their ‘contributions’ to health insurance benefits as just another part of the compensation package.

1.1 Group Practice and Prepayment

These features largely characterized the medical care environment until the early 1970s, but there were a few notable exceptions. As early as the late nineteenth century, some physicians organized themselves into group practices (Trauner 1977). Their goal was usually to share clinical responsibility and coverage, and to have ready access to consultations. While most charged their patients on a fee-for-service basis, and many compensated their physicians on a similar basis, there was often substantial opposition to group practice from independent physicians, probably because the ‘in-house’ referrals threatened their access to patients.

Prepayment for medical care actually had its origins in the mid-1800s with immigrant societies that hired physicians to take care of their members. By paying a fixed amount per month, members could receive medical care when they needed it without any further out-of-pocket payment (Trauner 1977). This was not just ‘insurance for very expensive events,’ but true prepayment. This prepayment to a small set of physicians was seen by independent physicians to threaten their livelihood.

During the Depression and shortly after World War II, several new organizations married the concepts of group practice and prepayment. These included the Ross-Loos Clinic in Los Angeles, Kaiser Foundation Health Plan, Group Health Cooperative of Puget Sound, and Group Health Association of Washington DC. These plans used group practice, with its shared medical records and internal physician quality review, in combination with prepayment for all needed services to remove the financial barriers to care.

The early founders of these plans wanted a better way to organize and deliver care. Many were not-forprofit, and some were consumer controlled. They often had a strong public health focus, and offered complete coverage of prenatal care, maternity care, well-baby visits, and immunizations, services excluded by conventional insurance. Some of the plans, notably Group Health Cooperative, Kaiser in Portland, Oregon, and Northern California, the Health Insurance Plan of Greater New York and later, Harvard Community Health plan developed highly-regarded health care research units (Luft and Greenlick 1996).

1.2 Health Maintenance Organizations

The growth of prepaid group practices (PGPs) in the 1950s and 1960s led to increasing evidence that, even though they avoided copayments and deductibles, offered more extensive benefits, had arguably com- parable quality of care, (although there were few measures of this), and seemed to keep most of their patients satisfied, their most impressive achievement was lower premiums. By the end of the 1960s, the expansion of insurance coverage by Medicare and Medicaid led to rapid growth in medical care expenditures. The Nixon Administration sought a new, market-based approach that would control costs, but the concept of prepaid group practice was anathema to the American Medical Association.

To address this problem, Lewis Butler and Paul Ellwood coined the term ‘Health Maintenance Organization,’ or HMO, in 1971. It encompassed both the concept of the prepaid group practice and the San Joaquin-style foundation for medical care (now relabeled the Individual Practice Association, or IPA) in which physicians in their independent practices collectively took responsibility for the costs associated with an enrolled population (Harrington 1971). The IPA paid the physicians on a fee-for-service basis, and there was no need for physicians to work together in group practices. While these two types of HMOs differed significantly in how medical care was organized and delivered, they shared the notion of integrating the insuring function with the responsibility for delivering services to a defined population.

There could be many different organizational and legal structures for these HMOs, often in response to the vagaries of state regulatory requirements. However, the nature of the classic PGP and IPA forced physicians to recognize that the economic implications of the decisions they made would be borne by their organization, rather than some distant insurer or government payer. The PGP, however, typically involved physicians who saw only HMO patients, and on whom the HMO relied to provide nearly all its medical care. This meant that they could develop implicit styles of practice consistent with clinical, professional, and organizational needs. In the IPA, by contrast, the vast majority of patients seen by its physicians were covered by other insurance plans, and because physicians did not practice together, informal approaches to developing consistent practice styles were impossible.

Aside from the usual policy tools of grants and loans to develop new organizations, and federal preemption of state laws prohibiting HMOs, a mechanism was developed to facilitate market entry by HMOs. If an employer of 25 or more workers offered a health insurance plan, HMOs could use the mandate authority in the legislation to require that they also be offered to employees at terms no less favorable than the fee-for-service plan.

The mandate option, however, was not unlimited. An employer needed only offer one plan of each type—PGP and IPA. This led to questions about how one should define the two types of plans. The commonsense notion was that the PGP combined prepayment for a set of benefits with the delivery of medical services through a group practice, while the IPA did not rely on the group, or clinic-like practice setting. However, suppose that a large PGP was already wellestablished in an area. A new PGP could not gain entry to employers wanting to avoid having to deal with multiple plans. If classified as an IPA, however, such a plan could gain entry in the marketplace, and a few IPAs emerged with salaried physicians who practiced in clinic settings.

Even without such subterfuge, the PGP and IPA definitions were inadequate. New HMOs began that contracted primarily with networks of such clinics, much like the Kaiser model, but unlike Kaiser, the clinic physicians continued to see fee-for-service patients. Some called these plans network model HMOs, but they were classified under the IPA rubric for mandating purposes.

An early review of the evidence on HMO performance (Luft 1981), based on data through the 1970s, focused largely on comparisons of HMOs versus fee-for-service insurance (FFS), but recognized the distinction between IPAs and PGPs. In fact, the cost-containing effects of HMOs were most apparent for PGPs. This was not surprising given the economic incentives. The more extensive coverage offered by HMOs increased demand, so some other mechanism, such as salaried payment to physicians or organizationally influenced practice patterns, was needed to constrain supply. Although IPAs may have had a fixed overall budget, they paid their physicians fee-forservice. Predictably, enrollees’ assessments of IPAs and FFS were comparable, but differences appeared between PGPs and FFS.

1.3 Managing Costs and Managing Care

The 1980s saw a slow but steady growth, both in the number of HMOs and their enrollment. Much of the growth was in forms other than PGPs. Nevertheless, the policy debate referred to the research on HMOs (generically) even when most of it was on PGPs. Furthermore, much of the research focused on a few highly visible plans, most of which were longestablished, not-for-profit, plans with internal research groups. Generalizing from these plans to all PGPs, let alone all HMOs, was clearly unwarranted. But, these studies provided strong evidence that alternatives to FFS could provide high quality care with low cost and reasonable enrollee satisfaction.

Nixon’s HMO strategy did not lead to the rapid growth of HMOs. Instead, government price controls were implemented and then private sector approaches to controlling utilization of services, such as ‘second opinion programs’ in which an insurer would pay for the procedure only if its necessity was confirmed by a second physician. The effectiveness of these programs was questionable, but it set the stage for the refusal by payers to simply reimburse every service rendered by a duly licensed physician (McCarthy and Finkel 1978). Instead, some services would be covered only if approved by the payer. These payment decisions could be quite problematic because the insurer became involved only after the services were rendered and the patient submitted a claim. Not surprisingly, these ‘retroactive denials’ resulted in substantial patient dissatisfaction.

By the early 1980s, HMOs were becoming increasingly visible outside of the geographic areas in which they had long existed, and employers were asking conventional insurance carriers why they couldn’t manage costs as well as did the HMOs. Large employers were self-insuring, thereby capturing the interest earnings on premiums, and avoiding statemandated insured benefits, such as mental health coverage and chiropractic. Utilization review firms that determined whether benefits should be paid were capturing the cost-control market, and third-party administrators were processing claims. In fact, the very survival of conventional health insurance plans was being threatened because their products were not unique, and others seemed to perform the separate functions better.

California legislation that allowed both its Medicaid program for the poor and private insurers to selectively contract with certain providers changed this situation (Bergthold 1984). Prior to this time, plans controlled the benefits they offered and the financial incentives placed on patients, but patients were free to go to any licensed health care provider. While the Blue plans had contracts with providers, the historical connections between these plans and the local hospital and medical associations meant that all providers interested in participating were accepted.

The intent of the California legislation for its Medicaid program was quite clear. Hospitals were encouraged to bid low in order to increase their market share at the expense of their competitors. Companion legislation allowed the development of what are now called Preferred Provider Organizations (PPOs), which could selectively negotiate contracts with providers for the ‘private’ market. With fees negotiated, and with a known deductible and coinsurance rate, PPOs were also able to offer their enrollees the advantage of direct billing by the provider. From the patient’s perspective, this is markedly simpler, but the physician may now have to deal with hundreds of different plans, often with differing claims forms and coverage rules.

To further improve the attractiveness of their plans to employers, insurers offering PPOs included the option of various utilization management approaches, such as prior authorization, concurrent review of hospital stays, and the like. Because the PPO now had a contractual relationship with the providers, it could require the provider to agree to these utilization management efforts as part of the contract. Undertaking them in advance would avoid embarrassing retroactive denials.

In most parts of the country there is an oversupply of hospitals and physicians, particularly specialists. Thus, when California’s state Medicaid program, or health plans more generally, presented providers with contracts in order to continue to participate, there were two reactions. The first was the hope that as a ‘contracting or participating provider,’ one would get more patients, even if the fee paid per patient might be somewhat lower. The second was fear that if one did not sign, a large fraction of one’s patients would be lost.

PPOs were seen by policy analysts as just a method to constrain payment levels—‘fee-for-service in drag.’ However, with a unique network of providers, in combination with various utilization management approaches, including using primary care physicians as gatekeepers, a new ‘product’ was defined. Conventional insurers began using the term ‘managed care’ in the mid-1980s to distinguish this new approach from the utilization management offered by small stand-alone firms, and from simple PPOs that merely negotiated discounts. Not surprisingly, the large insurers who were attempting to transform themselves in order to survive quickly pointed to the evidence on HMO performance as the rationale for managed care organizations.

1.4 The Managed Care Market En ironment

The Clinton health care reform proposals of 1993 added further terminology. In addition to promising close to universal coverage through a variety of funding mechanisms, there was a carefully designed effort to structure ‘managed competition’ among health care delivery systems. While fee-for-service could be one of these systems, others would be similar to HMOs. Managed competition, however, was supposed to deal with the flow of funds to health plans, to adjust for risk differences among plans, to monitor quality, and address consumer and patient service. Even if all managed care plans were well-designed, well-meaning prepaid group practices, a market organizing structure would still be necessary. Unfortunately, in the wake of the Clinton reform failure there is no such market organizing structure. The decade of the 1990s was one of intense market-based pressures on employers to become more efficient. Corporate restructuring and cost reductions were commonplace in all industries. Whereas wages and salaries were increasing at 1 percent or less per year, health insurance premiums were increasing at over 13 percent per year at the end of the 1980s. Employers began to pressure health plans to reduce that rate of increase. Aggressive cost containment became the watchword for health plan survival and the rate of premium growth fell steadily until by 1995 it was actually negative (Levitt and Lundy 1998).

These external pressures also led to a transformation of the health insurance and health plan industry. New forms of health plans developed and some not-forprofit Blue Cross and Blue Shield plans turned forprofit. Physicians sold their practices to practice management firms and became employees. Pharmaceutical companies began advertising directly to the consumer. These and other changes further eroded the traditional physician–patient relationship.

One core feature of the managed care organization is that it has a network of physicians, regardless of whether they practice together in a group setting such as in the old PGP model, or independently with contractual relationships to the health plan. This means that if a person switches from one plan to another, the physicians available may change. If the switch results from an employer changing health plan options in search of lower premiums, then the anger at having a longstanding relationship broken can be substantial. Reflecting this concern, there has been a shift to broader networks, including nearly all physicians, but this makes it more difficult for plans to concentrate patients and control costs by altering practice patterns.

Cost containment efforts by plans led to constraints on physician fees, both by the federal Medicare plan for the elderly, by state Medicaid plans for the poor, and by health plans. To maintain their incomes, physicians were under pressure to see more patients in less time. Interestingly, there is little evidence that the average patient visit has gotten shorter, or that physician incomes have fallen, but that is a common perception, as is the attribution of the problem to managed care (Luft 1999).

Outpatient drug coverage has become an increasingly common benefit during the 1990s, but the marketing of new, and much more expensive drugs has led to rapidly increasing costs for this benefit. Health plans have limited their formularies, but since coverage may vary across patients in a plan, let alone across plans, patients may find out their new drug is not covered only when arriving at the pharmacy.

The rapid growth of some plans in the mid-1990s, the conversion of some plans to for-profit, and the resulting fortunes made by some industry leaders attracted enormous attention in the press. Numerous studies pointed to administrative waste and high profits in the industry (Woolhandler and Himmelstein 1997). Alternatively, government data on health expenditures indicate that insurance administration and profit was falling between 1994 and 1997 (Health Care Financing Administration website).

At the same time, the cost pressures of managed care contracts has made it more difficult for hospitals to charge their paying patients more in order to subsidize the uninsured. Premiums for nonmanaged care plans have increased very rapidly, and there has been a general reduction in the prevalence of employersponsored insurance, leading to a rising number of uninsured. Thus, while managed care plans may be containing costs for their enrollees, the perception is that this market-focused approach has worsened the situation for many outside the plans, and even for those in the plans, many expectations are not being met.

2. Evidence on Managed Care Performance

Has managed care totally supplanted the promise of prepaid group practice? In contrast to earlier reviews, ‘current’ reviews found that both IPAs and PGPs demonstrated lower costs and utilization of services. As expected, consumers were often less satisfied with the nonfinancial aspects of HMOs, particularly when they felt they did not have a choice of plans. The evidence on quality of care was of particular interest. In contrast to the beliefs of either the advocates or detractors of HMOs, there were as many studies showing better quality in HMOs as worse quality. This balance, however, was not without significant findings—in fact, there were several studies with clinically important and statistically significant differences, but the balance of evidence was maintained. It also means that advocates and detractors can easily point to selected studies to buttress their case (Miller and Luft 1997).

This research has several implications. First, labels matter very little. Some HMOs have very good quality; some very bad. There is not enough data to determine whether ‘plan type’—PGP versus IPA—is a distinguishing factor, but the ways in which labels are applied suggest that plan type will not be very informative. While not-for-profits may have certain desirable characteristics, not all for-profit plans are ‘bad guys.’ Furthermore, it is almost impossible to separate (in a statistical sense) the effects of not-for-profit status from PGP model from ‘old established plan.’ It may be best to forgo a belief that types and categories truly determine performance, and focus instead on how to evaluate each organization, give it the appropriate incentives and resources to perform well, and then monitor that performance continuously.

An important conceptual limitation is that the standard for comparison matters. Some of the more recent evidence suggests that HMOs may have a smaller cost advantage relative to their competitors than had previously been the case. It also appears that areas with rapid HMO growth have a slower rate of growth in costs for FFS-covered enrollees (Chernew 1995, Robinson 1996, Wickizer and Feldstein 1995). These two findings are consistent and suggest that the spread of HMOs creates externalities, or ‘spillover effects’ outside the plans themselves. This will result in the incorrect observation that HMOs have no impact, when in fact, their impact may be pervasive, and this cannot be determined merely by comparison with FFS. Instead, external measures are needed for what is good care, what utilization of services is appropriate, and what costs are reasonable. Setting such ‘gold standard’ measures without reference to an existing medical care system may be impossible, and is certainly well beyond the scope of studies focusing only on managed care.

Furthermore, there is important evidence that choice, and the lack thereof, matters in the assessment of plan performance. Americans, especially those with sufficient incomes to see themselves as consumers, rather than as supplicants, value choice highly. The greatest dissatisfaction is expressed by people who have been given no choice but to be in an HMO (Davis et al. 1995). All types of health care systems and coverage imply some tradeoffs to contain costs. HMOs limit the choice of physicians, require prior approval before certain services can be rendered, or encourage physicians to sometimes say ‘no’ to patient requests. FFS plans use less visible means—restricting the benefit package, having premiums so high only the rich can afford them, or imposing substantial deductibles and copayments. The early PGPs, such as Kaiser, required that a FFS plan be offered as an alternative. This guaranteed that their enrollees were there by choice; if they were dissatisfied, they could, and did, leave. Choice allows people to get what they want, but it causes problems because not all patients are equally costly. Unfortunately, plans have no incentive to provide the best quality care because they will attract the sickest enrollees.

3. Lessons from Managed Care

In some ways, it is far too soon to offer lessons because the US is still in the process of developing a health policy, and managed care is but one of many tools being used. However, some speculation may be useful while awaiting more evidence. The early forms of managed care, PGPs, were designed as integrated delivery systems—integrated in the sense that both the financing and delivery components were linked. While some of these organizations owned their hospitals, this was not universal, and others were able to develop long-term arrangements with their hospital providers. Allowing funds to be flexibly allocated among inpatient and outpatient settings, physicians and other staff and pharmaceuticals seems to be one of the lessons of PGPs that may translate well across boundaries.

It is also important to note that these plans were, and still are, minorities in the context of an oversupplied, overly specialized, and unorganized fee-forservice system. Furthermore, the people who began these PGPs, and to a large extent, those still attracted to working in them, were seen as unusual, if not suspect. In turn, the organizations could design staffing patterns that fit their concept of integrated care, and not have to passively accept the specialty mix preferred by medical training programs. Being able to operate almost as islands in a FFS sea, with voluntary recruitment of staff and enrollees, allowed these plans to develop clinical practice patterns that were mutually acceptable, without having to force change on anyone. The US political system makes massive, uniform change impossible, but the lurches back and forth in other nations suggest that in the absence of knowing precisely what approach is best, experimentation on a smaller scale may be desirable. At the very least, it allows one to attract those most willing to adapt to a new system.

In the late 1990s, many of the problems that people point to in the US health care system reflect the absence of universal coverage, and thus large numbers of uninsured individuals, combined with a surplus of medical resources for those with coverage and a strong desire by the people for the right to choose both health care systems and providers. The presence of a wide variety of plans meets consumer demands for choice, but results in high marketing and administrative costs, along with incentives to not emphasize high quality of care for fear of attracting high-risk enrollees. Efforts to control costs imply changes in the perceived, and possibly actual, control physicians in traditionally FFS practice have over their decisions and thus encounter substantial resistance. The fragmented employer-based system of coverage in the US is not recommended for export, but experimentation with some variations in delivery of care (in contrast to financing coverage) may have some merit.

The changing managed care terminology suggests that new labels will be developed in the future, sometimes to better describe reality, sometimes to expand an accepted concept into new territory. This highlights the importance focusing on the components of health care delivery systems, how (and why) they are organized, the context in which they operate, and the standards by which they are assessed. New types of delivery systems are being created, some will succeed, but most will fail. New technologies such as information systems, will have enormous implications, but exactly how these will develop is unknown. The variability in approaches, however, allows us to examine in more detail what works and what does not, as long as one focuses on fundamental aspects, rather than simply labels.

Bibliography:

1. Anderson G F, Poullier J P 1999 Health spending, access, and outcomes: trends in industrialized countries. Health Affairs 18(3): 178–92
2. Bergthold L 1984 Crabs in a bucket: The politics of health care reform in California. Journal of Health Politics, Policy and Law 9(2): 203
3. Chernew M 1995 The impact of non-IPA HMOs on the number of hospitals and hospital capacity. Inquiry Journal of Health Care Organization 32(2): 143–54
4. Davis K, Collins K S, Schoen C, Morris C 1995 Choice matters: Enrollees’ view of their health plans. Health Affairs 14(2): 99–112
5. Harrington D C 1971 San Joaquin Foundation for Medical Care. Hospitals 45(6): 67–8
6. Health Care Financing Administration website (www.hcfa.gov stats nhe-oact nhe.htm) 1997 National Health Expenditures, Table 2, National health expenditures aggregate amounts and average annual percent change, by type of expenditure: selected calendar years 1960–97
7. Hellinger F J 1998 The effect of managed care on quality: a review of recent evidence. Archi es of International Medicine 158(8): 833–41
8. Levitt L, Lundy J 1998 Trends and Indicators in the Changing Health Care Marketplace. The Henry J. Kaiser Family Foundation, Menlo Park, CA
9. Luft H S 1981 Health Maintenance Organizations: Dimensions of Performance. Wiley, New York
10. Luft H S 1999 Why are physicians so upset about managed care? Journal of Health Politics, Policy and Law 24(5): 957–66
11. Luft H S, Greenlick M R 1996 The contribution of groupand staff-model HMO to American medicine. Milbank Quarterly 74(4): 445–67
12. McCarthy E G, Finkel M L 1978 Second opinion elective surgery programs: Outcome status over time. Medical Care 16(12): 984–94
13. Miller R H, Luft H S 1997 Does managed care lead to better or worse quality of care? Health Affairs 16(5): 7–25
14. National Center for Health Statistics 1999 Health, United States, 1999. US Government Printing Office, Washington, DC
15. Robinson J C 1996 Decline in hospital utilization and cost inflation under managed care in California. Journal of the American Medical Association 276(13): 1060–4
16. Starr P 1982 The Social Transformation of American Medicine. Basic Books, New York
17. Trauner J B 1977 From Bene olence to Negotiation: Prepaid Health Care. in San Francisco, Ph.D. thesis, University of California, CA
18. Vistnes J P, Zuvekas S H 1997 Health insurance status of the civilian noninstitutionalized population: 1996. MEPS Research Findings No. 8. Agency for Health Care Policy and Research, Rockville, MD
19. Wickizer T M, Feldstein P J 1995 The impact of HMO competition on private health insurance premiums. Inquiry Journal of Health Care Organization 32(3): 241–51
20. Woolhandler S, Himmelstein D U 1997 Costs of care and administration at for-profit and other hospitals in the United States. New England Journal of Medicine 336(11): 769–74
21. Woolhandler S, Himmelstein D U 1997 Erratum. New England Journal of Medicine 337(24): 1783