The Future of Children, Princeton - Brookings: Providing research and analysis to promote effective policies and programs for children.

The bottom line is that very little is actually known about how children and pregnant women fare under managed care arrangements. Managed care appears to offer both advantages and disadvantages. Available research sheds little light on the degree to which the hypothesized advantages and disadvantages are actually realized. The little that can be gleaned from the literature suggests that, on average, children and pregnant women are neither hurt nor helped by managed care. Similarly, there is not strong evidence that the costs of care are less under managed care than they would be in a more traditional system or that managed care can effectively control the growth in health care costs while maintaining health care quality. Many proponents of health reform argue for a managed competition system which will rely heavily on managed care entities to deliver services.⁵⁸ Unfortunately, because the managed care market is in a state of flux, it is difficult to project what types of arrangements would pertain under such a managed competition system and, thus, how vulnerable populations would fare. Indeed even evidence gathered in 1993 may not be a good predictor of what would pertain in a substantially reorganized health care system. But clearly as the market penetration of managed care has exploded, plan arrangements keep differentiating so that it is extraordinarily difficult to assess the effect of managed care on costs, on the quality of health care received by children and pregnant women, and on their access to care.

Thus, to conclude, it is appropriate to discuss the areas where more research is needed to monitor and perhaps further refine the evolving system of managed care plans and to highlight areas that may require particular vigilance to protect the interests of vulnerable populations in managed care arrangements.

There is now tremendous variation in the types of managed care plans available. Plans differ in the degree of restrictions placed on patients, but in general, each plan requires that a prospective enrollee trade off an unlimited choice of providers, particularly specialists, for greater access to primary care. Because it is likely that individuals differ in their willingness to trade primary care for specialist access, it is likely that no single form of managed care will dominate. How children or pregnant women with special needs become distributed among different plans is poorly understood, as is the capacity of different plan configurations to care for such individuals. A pluralistic system which preserves a range of options may be beneficial for most consumers but still may not safeguard access to highly specialized and expensive care for those children and others with unusual health problems. In any event, the ongoing selection of individuals into some plans and out of others (that is, the selection bias problem) makes determining the true net effect of managed care arrangements difficult.

In addition, the fate in a pluralistic managed care environment of individuals residing in more sparsely populated areas or in areas suffering urban blight is unknown because managed care is unlikely to flourish unattended in such areas. Under ideal conditions, managed care plans should do the shopping for enrollees by striking deals with providers for lower prices, improved quality, or enhanced services. For this process to work well, plans must have reasonable alternative choices among providers. The right mix of plans and providers in individual markets is unresolved as yet, but clearly this process will not work as hypothesized in medically underserved areas where the choice of providers is limited.⁶⁵ For example, Medicaid managed care plans in poverty neighborhoods have frequently had to raise fees to providers to encourage them to participate in the plans in contrast to the discounts many private plans are able to negotiate with providers in more affluent areas, where there are many providers.⁶⁶ In addition, in medically underserved areas, it may not be possible to offer enrollees a reasonable choice of providers or plans or the ability to switch plans and providers, but these options are key elements to promote accountable health plans and consumer satisfaction. Accordingly, plans in rural areas and urban poverty neighborhoods may require different sanctions and incentives both to control costs and to assure quality.

As has been discussed throughout this review, an important theme of much of the research on managed care has been the effects of patient selection on outcomes. Very little attention has focused, however, on an important related issue, the self-selection of physicians into practices that do or do not accept managed care patients. Yet, most managed care models rely on physicians to play a key role in providing cost-effective care, so there is good reason to believe that much more attention should be paid to the effects of physician self-selection.

The issue of physician self-selection is critical from several perspectives. First, it is well documented that there is considerable variation in medical practice across different geographic areas and even among physicians within local medical markets. There are disagreements within the medical profession about the utility of various medical procedures, and these disagreements are reflected in substantial variation in rates of hospitalization and the performance of costly diagnostic and therapeutic procedures. These variations have important consequences for the cost of care, but their impact on health outcomes is less well established.⁶³

Doctors may also develop styles of practice that are reflected in the amenities they offer their patients. For example, some physicians may prefer a busy practice with standard amenities; others, a high-volume, low-amenities practice (typical of many Medicaid practices); still others, a low-volume, high-amenities practice with extended time for each patient visit.⁶⁷ Clearly, each practice style will primarily attract patients who prefer that style, and each style will be more or less compatible with managed care. Accordingly, managed care plans that attract physicians whose style is compatible with the goals of the plan may function well but will have little actual impact on medical practice. In fact, favorable selection of physicians may be an important, and largely unstudied, form of “cherry picking” by apparently successful plans. This practice may be enhanced by the enrollment practices of some newer forms of managed care which concentrate on developing networks of physicians with established practices. By signing up physicians whose practice style is compatible with managed care, the plans may also obtain a compatible population of enrollees.

Again the issue from the public policy perspective is what happens as the market for managed care expands to include an increasing number of physicians whose practice style is less compatible with managed care procedures. Socializing providers into highly structured programs may be difficult because it may involve changing their habits and challenging their prerogatives.¹⁹ Little is known about how to do this successfully, and the substantial ground swell of discontent among the medical profession which has accompanied increased efforts to monitor and influence clinical practice suggests that the process of socialization will require a lot of attention if managed care is to function well on a large scale.⁶⁸

A not unrelated concern pertains to the possible impact of managed care on socalled safety net providers, public and private hospitals, neighborhood health centers, and other facilities that have traditionally cared for uninsured children and others, often on an uncompensated basis. The ability of this sector of providers to function or even exist in a managed care environment is in doubt. In fact, the study of the Jackson County MDC site showed that public hospitals may find coping with the organizational changes necessary for managed care particularly challenging.⁶⁹ Moreover, the ability of private and public hospitals to provide care for those unable to pay depends on whether these institutions can recoup their costs through higher charges to insured patients and direct subsidies from the government and private charities. If price competition increases, the ability of hospitals to cross-subsidize their operations will likely decline. For example, a recent study of the effects of competition among managed care plans in California, the state with the greatest penetration of HMOs, reported a decrease in the level of uncompensated care offered by hospitals of $.50 worth of care for every $1.00 in price discounts obtained by managed care from participating hospitals.⁷⁰

Another phenomenon worth exploring is the potential impact of Medicaid managed care on the finances of hospital emergency departments. The literature shows unequivocally that managed care blunts emergency department use. At present, however, many Medicaid children receive primary care in emergency departments. If hospitals are now reimbursed for more than the actual costs of providing this primary care, a reduction in emergency room use that would accompany managed care could cause hospitals to curtail the operations of their emergency departments and perhaps reduce access to care for true emergencies. To the extent that large numbers of poor and low-income children, whether covered by Medicaid, uninsured, or underinsured, rely heavily on safety net providers for necessary health care, they will be threatened by health care reforms that may undermine the solvency of these providers. Accordingly, it may be appropriate to attempt to include safety net providers in managed care systems and to go slowly with health care reforms that may jeopardize these providers.

Serious concerns also arise about how well very sick children and others needing expensive specialty care over an extended period of time may be treated in managed care plans. Because most children are basically healthy, the distribution of health care costs among children is highly skewed. In 1987, 5% of children accounted for 59% of health care expenditures,⁷¹ and mean health care expenditures for the most costly 5% of children were more than 27 times greater than mean expenditures for the least costly 95% of children. Moreover, because many of these high-cost children suffer from chronic illnesses which will persist over a long period of time, it is likely that the concentration of costs in a small number of children over a long period of time may be even more extreme.⁵⁹ Not only will care of these children be costly, but they may require special services which may not be readily available nor efficiently provided by a managed care plan's network of providers.

At present, there is no evidence that care is withheld from high-cost users. However, plans face considerable incentives to skimp on their care, and the finding that managed care is associated with decreased utilization of medical specialists suggests that concern about the possible undertreatment of high-cost children in managed care may be justified. Moreover, the competitive option of disenrolling from a particular health plan because of inadequate care is of little value in this situation, for other health plans are unlikely to be willing to enroll a high-cost child except at an appropriately high premium. Regulation may be necessary to protect access to beneficial services by very sick children. Alternatively, this problem may be addressed through point-of-service options with high-cost case management which will help finance and coordinate the care of children with high health care costs. In any event, research is clearly needed to determine the extent, if any, of undertreatment of high-cost children in managed care plans and to examine the ability of alternative policies to help remedy the problem should it exist.

The skewness of the distribution of health care costs has implications not only for children and pregnant women enrolled in managed care plans but also for the success of managed care plans themselves. Adverse selection (a disproportionate number of high-cost cases) can have a disastrous financial result and cause otherwise well-run plans to become insolvent, leaving enrollees without access to contracted health care services. The risk of plan failure may, however, be reduced by requiring that plans have adequate reserves and by providing a mechanism for plans to spread their risk.

Maintaining the financial incentives for plans to deliver care efficiently while sheltering them from devastating losses for high-cost cases can be tricky. Payment for pregnancy, child birth, and infant care provides a good example of the difficulties that can be faced in attempting to balance incentives and risk. There are few medical events more costly than the care of a very low birth weight infant. Managed care plans that experience a disproportionate number of such births would likely face financial disaster. Accordingly, it might be appropriate to provide plans with back-up risk-spreading mechanisms (such as reinsurance) to protect them from such large losses. However, the prospect of large losses can also provide a strong incentive for plans to provide pregnant women with a full complement of prenatal care which may reduce, but not eliminate, the likelihood of high-cost infants. Accordingly, a key yet-to-be-resolved issue is how to design payment schemes that encourage plans to manage pregnancies (and all care) efficiently while protecting them and their enrollees from the risk of untoward events.⁷²

There are a number of areas of concern about the health care system which take on particular importance in the context of managed care with its emphasis on cost containment and restrictions on access to providers. Marketing is one such area of concern. To the extent that managed care plans contain benefits and restrictions which are quite different from either traditional indemnity insurance plans or fee-for-service Medicaid, it may be appropriate to inform potential enrollees of the features of the plans prior to enrollment and to educate them regarding how to navigate the organization successfully before the need for service arises.

Many plans currently rely on a directory of participating providers, a listing of enhanced services, and sometimes promises of reduced cost sharing and administrative hassles to market their plans. Anecdotal evidence, from both the Medicaid and privately insured populations, suggests, however, that enrollees may not fully appreciate the fact that enrollment in a plan may restrict their access to convenient sources of care or to high-cost specialty care.^{73, 74} As the number of plans proliferate and differentiate, it may be even more difficult for consumers to evaluate plans prior to enrollment. Just as the federal government found it necessary to regulate and standardize offerings in the market for Medicare supplementary insurance, so too it may be helpful for either the government or private voluntary organizations to educate consumers about the benefits and drawbacks of individual managed care plan features and provide some oversight of plan marketing activities.⁷⁵

Assuring the quality of care within managed care plans and providing information on quality to enrollees, potential enrollees, and regulators are important aspects of managed care which have hardly been explored. Quality assurance, like so many features of managed care, is also an important and controversial area of activity in traditional, nonmanaged health care systems. It takes on particular importance in managed care, however, because it has been hypothesized that managed care has the potential to improve the quality of care; but incentives also exist in managed care to undertreat costly patients and to control information by restricting access to alternative providers, which may make the detection of inappropriate treatment more difficult. In addition, the greater reliance of managed care on capitation payments and global budgeting rather than on fee-for-service payments may make measurement of the quality and even the quantity of the care delivered problematic. Not only has the quality of care received by children and pregnant women in all care settings not been well studied, but even less attention has been paid to developing mechanisms to monitor and assure quality in the many different types of plans currently being offered or in development.

Recently, 11 leading HMOs joined with staff from RAND Corporation and the UCLA Department of Medicine in a consortium to develop an assessment system to measure quantitatively the quality of care delivered in various types of prepaid health plans. This activity is intended to produce a set of scientifically validated quality assessment measures and instructions on how to implement the system.⁷⁶ Adoption of this quality assessment system, however, would be the responsibility of the industry, government, or private sector groups who are interested in quality of care. Areas of particular interest to children and pregnant women targeted by this activity include prenatal care and rates of low birth weight births, childhood immunizations, and the process of care for otitis media and childhood asthma. Results of the consortium's research are not yet available, and whether its activities will yield useful tools to monitor quality in managed care plans is unknown. The consortium's work, however, is an example of the type of activity that will be necessary to add the dimension of quality to the health care purchasing decisions consumers and government agencies are making.

In recognition of the demand for objective information that employees and consumers can use to compare the performance of competing health plans, United Healthcare Corporation, a member of the consortium, released a report on the quality of care it provides in March 1993.⁷⁷ United Healthcare used many of the indices being investigated by the consortium in compiling its report card. If this activity is well received, it may encourage other plans to follow suit and spur development of quality assurance indices that can be applied objectively across different plans.

Clearly, there is a long way to go before we can determine whether managed care can realize its potential of providing children and pregnant women with high quality care at reasonable cost or is proven inadequate to this daunting task. For more than 70 years, managed care organizations have functioned side by side with the traditional fee-for-service delivery system, but that they have flourished in a small, self-selected niche of the health care system is no guarantee that they will be able to address successfully the many problems faced by the U.S. health care system today. Because the current situation is so fluid and the track record of managed care, particularly in many of its current forms, so uncertain, the situation warrants careful watching by those concerned with the health and well-being of children but also provides fertile ground for health services researchers, regulators, and innovators.