Assessing Health Care Reform (1993)

For many policymakers and citizens, the defining objective of health care reform is to create more uniform, secure, and effective access to health care and health insurance for the people of the United States. This focus is understandable. Millions of Americans are uninsured or otherwise lack reasonable access to health care. Millions more fear that an illness, an employer's decision to cut health benefits, or some other event beyond their control could deprive them of coverage. Even voluntary actions, for example, a job change, are increasingly constrained by concerns about continued health coverage. What was once perceived as largely a problem for low-income people is now a growing concern for middle-income people as well.

Just as the cost of medical care is a major problem for many Americans, high and increasing health care costs are a major obstacle to health care reform. It has, for instance, been argued that health care costs must be contained before major steps to extend health coverage can be undertaken. The committee believes that such a two-step reform strategy underestimates the foreseeable costs of not extending access, in particular, the extent to which the lack of universal health coverage produces costly distortions in the way care is provided and financed and in the decisions made by employers, employees, and others. These distortions include patient delays in getting needed care, ''job lock'' (which occurs when a worker refuses an otherwise better job because it does not include any or acceptable health benefits), and to cost-shifting (when providers try to recover costs related to charity care and underpayments from public payers through higher prices to private payers).

The committee, nonetheless, expects that expanding health coverage will add to total health care spending—both public and private—in the near term. Although a phased-in strategy for broad reform—including a phased-in approach to improved access—is reasonable, such an approach should include early steps to achieve broader and more predictable health coverage. Expanded access should not become a second-stage contingency.

The primary barriers to access stressed in most reform proposals are financial—in particular, absent, inadequate, or unreliable health insurance. Proposals should also be grounded in a realistic understanding that access to effective health services is more than a matter of money. Other barriers to access also require attention.

Many obstacles can stand in the way of the timely use of effective health services. These obstacles include

• financial barriers, as noted above;

• distance from primary, secondary, and tertiary medical services, a problem exacerbated by transportation limitations;

• lack of education, language barriers, unsafe neighborhoods, and other nonmedical problems;

• deficiencies in the structure and functioning of the health care system, including the general lack of coordinated approaches to service delivery and provider payment levels that in some programs (notably Medicaid) may be too low to secure adequate access to care for beneficiaries; and

• shortfalls in our knowledge of how clinical care, the community and workplace environments, and individual social, cultural, and biological characteristics interact to affect health status and the use of health services.

To the extent that health care reform helps to reduce these barriers, it can allow more people to improve their health and well-being by using the right medical services at the right time. No one, however, should

expect such reform either to solve all health-related problems or to address the sources of these problems equally. For some problems, such as cocaine addiction in pregnant women, no effective medical therapies exist. For other problems, such as injuries and deaths due to domestic and neighborhood violence, nonmedical policies and programs that focus on issues such as job creation, education, and crime control may do more than medical programs to improve health and well-being. One reason for bringing the growth in health care spending more in line—even if only marginally—with growth in the rest of the economy is to free resources for these kinds of programs. In general, a long-term commitment to the objectives of health care reform requires a prudent appreciation of both the promise and limits of health care reform for achieving better health outcomes.

On both philosophical and practical grounds, the committee believes that an attack on financial barriers to health care should work from the basic, interrelated principles outlined below (IOM 1991f, 1993a,b; NRC/IOM, 1992). Transforming these principles into legislation—and translating legislation into improved access to effective health services—will be exceedingly difficult for both political and technical reasons. Current opinion polls and hard experience point to a gap between the generous values that this nation broadly espouses and the narrower interests that we actively protect. These realities notwithstanding, the following principles are, in the committee's view, an appropriate foundation for health care reform. Some important practical implications and difficulties raised by these principles are discussed in the next section.

• All or virtually all persons—whether employed or not, whether ill or well, whether old or young—must participate in a health benefits plan. The plan may be a single nationwide program or one of several alternative public or private plans, but no one should need

to depend on charity care or be allowed to stay uninsured while they are healthy.

• Whether a single health plan or multiple plans are envisioned, a uniform package of core or basic health benefits must be defined and periodically updated. The package should include an array of services that are thought to be valuable in improving health and well-being. The committee also believes that individuals should be able—consistent with equity and other policy objectives—to purchase coverage to supplement the core package on a non-tax-favored basis.

• If multiple health benefit plans are permitted, policies should minimize barriers to initial and continued health coverage (e.g., waiting periods and restrictions on coverage for preexisting health problems) for those who move, change jobs, become ill, start or stop receiving public assistance, or face similar changed circumstances. Otherwise, such barriers will undermine both access to appropriate health care and labor market mobility.

• Requirements that individuals share in the cost of health coverage and health services should not create barriers to needed care for low income individuals, although administrative practicality will limit the degree to which deductibles, copayments, and similar cost-sharing mechanisms can vary by income. Certain services, such as selected preventive services, might be exempted altogether from cost-sharing provisions.

• To reduce incentives for health benefit plans to compete for healthy individuals and avoid the ill, the payments received by health plans (from governments, employers, or other sources) should be adjusted to reflect important differences in the distribution of low-risk and high-risk individuals across health plans. Some plans may also receive additional support for intensive outreach programs for special populations such as low-income mothers and children.

• Correspondingly, what individuals pay for health coverage should not be linked to their health status (past or anticipated), age, gender, occupation, or similar factors. Thus, what an individual pays into the system for health coverage may differ from what is paid out to a health plan for enrolling that individual. Some argue that it is fairer and more efficient to link individuals' payments for coverage to their risk of incurring expenses, but, on balance, the committee believes that such

discrimination in the cost of coverage is a divisive and highly imperfect way of achieving greater efficiency in the financing, use, or provision of health services.

This discussion highlights only a few of the more prominent implications of the above principles. First, if coverage is to be universal, the healthy and the well off must share the cost of covering the ill and the poor. This is not simply a matter of philanthropy but a reflection of our common and lifelong vulnerability to illness and disability—to moving from the category of well to ill. However, to the extent that people are divided into multiple, separate financing arrangements (or risk pools) such as employee groups, a broad sharing of the costs of protection against medical expenses becomes more administratively and politically complex. As recommended earlier, requirements that individuals share in the cost of health coverage and health services should provide for some adjustment by income, within the bounds of administrative feasibility. Again, the details of a benefit package may be left to administrative agencies or perhaps a special commission, but a reform proposal must set forth basic criteria and procedures to be followed.

Second, any proposal that involves multiple health benefit plans requires methods and mechanisms for (1) distributing payments to health plans and (2) augmenting any individual payments for coverage with contributions from employers, governments, or others that are adequate to cover appropriate, efficient care to the specific mix of more and less healthy individuals who select any particular plan. Without such adjustments, which may involve government, employers, or specially created organizations such as purchasing cooperatives, health plans will face strong incentives to market only to the well and to discourage enrollment by the less healthy. Unfortunately, strategies to adjust payments to reflect risk selection are not yet technically or practically adequate, although progress is being made. The more a reform proposal includes other means to manage or compensate for risk selection, however, the lighter the burden will be on techniques for adjusting health

plan payments. Standardized benefit packages, some reinsurance provisions, and monitoring for abusive health plan marketing and management practices are examples of such means (IOM, 1993b).

Third, a provision for core or standardized benefits based on effectiveness implies the need for explicit processes and criteria for defining and updating such benefits that build on the best scientific evidence and professional judgment available. Some of the most politically sensitive questions any reform proposal must answer are: who will define and update the basic package and determine any restrictions on supplemental coverage, what measures of effectiveness will be employed to include, exclude, or discontinue coverage for specific services, and what other criteria for inclusion will be employed? The stakes involved in benefit design are very high for specific categories of health care providers and patients. Formidable difficulties face any effort to establish a structure that is reasonably accountable to the public, reasonably insulated against pressures from narrow interest groups to expand excessively the core package of benefits, and reasonably feasible to initiate and sustain over time. Points four through seven below and the discussion of core and supplemental benefits in the cost containment section of this report all underscore this point. Again, defining basic benefits will present policymakers with many difficult problems.

Fourth, at this stage, evidence for the effectiveness of a great many services—including many that are widely believed to be effective and that are generally covered by public and private health plans—simply does not exist (IOM, 1992b). Many (perhaps most) judgments about what services are appropriate or—more stringently—essential will have a considerable subjective component that should be acknowledged. One element of subjectivity relates to expert judgment in the absence of evidence; another involves differences between experts and patients (or among patients) in judgments about what outcomes are important. It will be expensive and time-consuming to increase our base of scientific knowledge and bring professional judgment systematically to bear on the question of what care is effective for the great range of individual health problems. Thus, linking a core benefit package to services of demon-

strated effectiveness and value to patients will be an incremental process.¹ Initially, many services of undocumented effectiveness would need to be covered to reflect patient, practitioner, and community perceptions of what care is appropriate. As our knowledge base expands (as discussed in the section on infrastructure), that should change.

Fifth, whether the package of core benefits should be relatively narrow or broad will be a major point of contention as the objective of expanding access competes with policymakers' efforts to limit the cost of making the basic package available to those now uninsured. To limit inequities in access, the committee believes that the core package or standard plan needs to be reasonably comprehensive.

Sixth, if the core package is defined narrowly (e.g., it omits coverage for outpatient prescription drugs and mental health services) to reduce the cost to government and employers of subsidizing it, then supplemental benefits will become more important to many individuals. Purchase of such benefits, however, will be difficult for lower income individuals, and even employer-paid supplemental coverage may be a limited blessing if it increases individual tax liability. If the core package of benefits is a "bare bones" one, then the committee

recommends that supplemental benefits for selected services be tax-favored.

Seventh, reforms that provide for basic and supplemental benefits imply additional procedures and methods for relating the two kinds of benefits and monitoring both. The aim is to ensure that the availability and marketing of the supplemental benefits do not seriously subvert the equity and effectiveness objectives sought through the definition of a basic benefit package and do not unduly complicate the evaluation of competitive health plans with different supplemental coverage. This may require that supplemental benefits be priced separately and purchasable either from the plan providing the basic benefit package or separately. Depending on the way core and supplemental benefits are structured in the context of a comprehensive plan for health care reform, adverse selection might make supplemental benefits unworkable except perhaps for those provided across-the-board by employers to all their employees.

Eighth, a provision for standardized benefits implies a process for defining benefits that applies to health maintain organizations and other network health plans as well as fee-for-service settings, assuming that a reform proposal accommodates both. A related issue is how patients should be informed about health plan policies that constrain access to certain generally covered services, for example, formularies that limit access to many drugs or staffing criteria that restrict the numbers and kinds of practitioners available to provide mental health services. The varied opportunities for health plans to restrict access may, in turn, imply the need for some system to monitor that covered services are reasonably available from health plans and not just pro forma promises in health plan brochures. For example, if patients face long delays in getting a particular service in the basic benefit package because their network health plan has deliberately hired or contracted with too few providers of that service, does the plan at some point fall out of compliance with basic benefit requirements? The design of a reform proposal should include some guidance on such questions, even if most details are left to implementers.

Ninth, because reforms, once adopted, cannot just be assumed to be successful in meeting their objectives, policymakers need to monitor changes in access over time. A recent IOM report (IOM, 1993a) contains specific recommendations on this point, and the

discussion of information and research issues in this report is also relevant.

Tenth, if a reform proposal would continue employment-based health benefits, the committee believes that the Employee Retirement Income Security Act must be significantly amended and strengthened. The discretion now accorded self-insured health plans, on the one hand, and state insurance regulators, on the other, should be rationalized and, in a number of areas, circumscribed. Critical coverage, funding, and other health plan features should be made more consistent across health plans, more reliable and predictable over time, and less of a barrier to the continuity of health care and job mobility.

More generally, the current relationships between public and private programs and state and federal policy with respect to health benefits will require considerable rethinking and realignment. Health care reform proposals that aim to provide universal coverage and control costs should define the role of existing public sector health care coverage or service programs such as Medicare, Medicaid, the Department of Veterans Affairs health system, public health services, and various federal block grants to states and communities. Even if reform proposals do not directly encompass the beneficiaries of such programs, they should be clear about how those massive public programs ought to tackle their own serious access and expenditures issues. Moreover, those parts of any reform effort intended to extend insurance coverage to the uninsured must also address the problems of the poor and medically needy who may (or may not), on any given day, be eligible for or enrolled in Medicaid.

The current structure of the federal-state Medicaid program is already the subject of much dissatisfaction independent of the agitation for health care reform. Several options exist for Medicaid. They include retaining Medicaid's focus on the medically indigent but improving program policymaking and operations; expanding it to enroll individuals not covered by employment-based plans; modifying the program by moving the elderly long-term care portions to Medicare; limiting the program to long-term care; or abolishing it altogether and covering everyone not eligible for Medicare or employment-based coverage under a single health program. Neither this committee nor

other IOM groups have systematically examined these approaches; thus, the committee takes no position on them. The committee emphasizes, however, the importance of thoroughly rethinking the program's role and structure.

Whatever their specific philosophy and approach, reform proposals should be sufficiently detailed that policymakers and others can understand and evaluate six key dimensions of the reform strategy. These dimensions are:

• The processes and criteria that would be used to define the core benefit package, including how they would take into account health outcomes, financial constraints, community values, patient preferences, and scientific evidence; how the package would be revised to take changing technology or other factors into account; how the benefit package would relate to prepaid or capitated delivery systems; and what kind of role would be envisioned for coverage that individuals or their employers could purchase to supplement the core benefits.

• The processes and policies that would be adopted to discourage or overcome biased risk selection and discrimination against higher-risk and higher-cost individuals.

• The restrictions, if any, that would be placed on the free choice by individuals of health practitioners and providers.

• The provisions that would be needed to minimize disruptions in continuity of care that could arise from employers, governments, or other purchasers changing the health plans they sponsor, from individuals changing jobs, or from turnover in the practitioners associated with particular health plans.

• The degree to which geographical or employer-based variations in the cost of coverage would be allowed or limited.

• The ways special categories of individuals such as foreign tourists and undocumented foreign workers would be treated.

Because improved access and health status are key goals of health care reform, and because impediments to the achievement of these goals are not just financial, proposals to extend health insurance coverage should define where coordination is needed with other public and private programs that target these nonfinancial barriers. Such programs will include:

• broad public health and health education initiatives that help people understand how to take care of their health, use health care services appropriately, and seek healthful environments in the home, the workplace, and the community;

• efforts to structure health care services, systems, and financing to more effectively reach special populations such as residents of inner city and rural areas, high-risk mothers and children, frail elderly persons, homeless and migrant groups, and those with certain health problems such as AIDS, substance abuse, and severe mental illness;

• programs to recruit and train (or retrain) health care practitioners to support expanded access to primary and preventive services, especially in areas where such services are already in short supply;

• clinical and health services research that provides a better knowledge base on which to construct education, health promotion, and special outreach activities; and

• vigorous, well-financed programs of quality assurance and health services research to help protect against potential unwanted side effects of health system bureaucracies (both public and private) and cost containment efforts.

These programs, in turn, will require bridges to nonmedical programs and policies that focus directly on the many social and other problems that particularly burden some populations. Although health programs are not the main avenue for tackling such issues, health workers may involve themselves with problems as diverse as homelessness, inner city and rural transportation deficits, and domestic violence.

For no special population is concern more acute than for children and pregnant women, especially those who are at high risk either medically or socioeconomically. Reform proposals should include specific provisions for these groups; that is, they should cover all or virtually all the cost of services that are critical to the health and well-being of these groups. For children, these services include routine immunizations, well-child care, and routine dental care. For pregnant women and women of childbearing age, benefits should include prenatal diagnosis and care and family planning services. For those at highest risk, provision for nutritional support and maternity outreach will be necessary (IOM, 1985, 1992d,e), in addition to services covered under the basic benefit package.

Public sector health and social networks for disadvantaged children and high-risk pregnant women are, for the foreseeable future, essential to ensure their access to these kinds of services. Under some circumstances, comprehensive, multidisciplinary, community-based centers can be more effective in serving these populations than freestanding private practices, because the sociomedical problems of these individuals are often greater than a single office-based practice can accommodate (IOM 1992d,e). The special resource needs of such comprehensive centers for disadvantaged children and pregnant women should be recognized if they are to offer effective services.

Also, health plan features designed with the middle class in mind may prove to be significant obstacles to obtaining appropriate, effective, and timely care for those with low levels of income and education. For example, some managed care plans now decline to pay for more than 24 hours of hospital care following normal childbirth, thus reducing the opportunity to monitor newborns in the critical first one or two days after birth and to teach new mothers how to care for their babies.

Such policies, as well as requirements for advance approval of hospitalization and similar cost containment policies in private and public health plans, should be studied to determine whether they impede the use of appropriate or effective health services, particularly by special populations. The same call for research also applies to the complicated administrative structures that cut across or link coverage under public and private programs for the elderly, the poor, and other groups. Even for the college educated, health plan rules and procedures can be

confusing and discouraging, and they also demoralize health care practitioners.

A challenge for health care reform is to minimize bureaucratic complexity and to assist people in negotiating the structural and procedural complexity that remains. Such efforts will supplement quality assurance and research programs and help to identify and remedy both financial and nonfinancial barriers to appropriate care.