Printer Friendly

Health policy guided by five questions.

The Congress now has before it, and in the next few years will continue to consider, various policy options for restructuring the U.S. health care system. These options range from mandating employer-provided health insurance (e.g., the Pepper Commission approach), to federalizing Medicaid, to covering the entire population under Medicare, to establishing full-blown national health insurance with the government as the provider. The answers to five questions are particularly important to the evaluation of these (or any other) policy options.

Question 1. How does this option

affect existing and potential


Overutilization is the central dilemma of our existing system and the primary problem any proposed reform needs to address. [1] "Overutilization" is the consumption of resources, the cost of which is in excess of the value or benefit generated by that consumption. The evidence of overutilization of health care resources in the United States is our higher level of per capita consumption than in any other country, generally without commensurate results in population health status indicators or individual outcomes. [2]

Our fundamental overutilization is characterized by an excessive base (volume) of resources devoted to health care more than by excessive year-to-year rates of increase. Rate setting (price controls) such as in the Prospective Payment System, budgets negotiated relative to the previous year (e.g., the state of Maryland and the Canadian system), and various all-payer systems simply don't address the aggregate base. Further, there is no reason to expect the after-the-fact approach of Volume Performance Standards within the physician payment reform mechanisms (resource-based relative value scales) to produce any significant reduction of the Medicare Part B base; it may well cause quite the opposite result. [3]

Any successful policy must first address shrinking the existing base of excessive resource use. Constraining existing overutilization will require a multidimensional strategy that would reasonably include achieving at least the following four objectives:

* Consensus managed care--At the delivery level, we must encourage the development of both protocols and individual care plans in which services are tailored to meet only identified medical needs through even-handed consultations between care givers and managed care professionals. Correctly done, managed care needs not conflict with the basic principles embodied in medical tradition and the Hippocratic Oath.

* Constraining "defensive" care--We should remove economic incentives for care givers to attempt to limit tort liability by engaging in defensive care, defined here as the application of all remotely relevant resources, independent of the expected direct benefit of those resources to the care recipient. Obviously, achievement of this objective requires significant reform of tort law.

* Ensuring care recipients "vote their own dollars" for routine care--A means-tested approach will have to be used to maintain individual deductibles and copayments at levels that ensure care recipients' choices involving routine care are sensitive to price. "First dollar coverage" causes care to appear to be a free good, insulating the consumer from the cost of the resources consumed and encouraging the care recipient to overconsume resources of marginal benefit. One lesson of the RAND Health Insurance Study was that nonpoverty-level individuals spending their own dollars (e.g., through high copayments) consumed fewer resources without adverse effect on health status or treatment outcomes. [4] Whatever the legal status of providers waiving copayments and deductibles, doing so from a system perspective is counterproductive, just as the across-the-board availability of standard Medigap insurance coverage is bad system policy. All of us, within our fiscal capacities, need to stop spending someone else's dollars and to start spending our own.

* Lowering public expectations--Recognizing the political value in promising constituencies more rather than less, we need to find a mechanism for dispelling the myth initiated a quarter century ago that each citizen has a "right" to all health and medical care that is technologically possible, regardless of cost. The Oregon Medicaid experiment in establishing care priorities, incorporating cost/benefit tradeoffs in full political daylight may provide an important model for building efficient social safety nets.

Question 2. Will this option actually

result in increased real (physical)

access to care for those lacking

such access?

Providing fiscal access to care (stimulating demand for services) should not be equated to providing physical access to care (supply of services). Defining a service as "covered" does not guarantee that any provider will step forward to deliver it. This is especially true in rural areas and many central-city enviroments. While the definition of covered services will obviously influence to some degree the supply of services, the way in which delivery of care will be organized, the mix or composition of institutional providers, and the supply of health manpower, mere influence is often not enough.

Obviously, paying too little for services (e.g., Most Medicaid payment) effectively denies access, but even extending significant financial carrots may be insufficient to stimulate the supply of services in some settings. Medicaid in Chicago pays less than one-half the private-pay fee for a routine office visit, and the fee is typically paid months after a private fee would have been collected. Not surprisingly, Medicaid-eligible patients are finding it increasingly difficult to locate a physician who will see them. As in New York City, these patients often wait eight hours to be seen in a hospital emergency department for nonemergency care because no physician will see them. Yet even larger financial incentives may fail. In Louisiana, fifty-seven areas with total population of about 1.2 million have fewer than physician per 3,500 persons, and one-third of that population has no primary care physician within 40 miles, all notwithstanding the state's program forgiving substantial medical school student loans for physicians who work in deprived rural areas. And New Mexico, North Dakota, and Mississippi all rank lower than Louisiana in primary care physician supply.

Even where adding a particular service or benefit to a package of coverage would result in a supply-side response, providing targeted delivery of care within a specific budget rather than an open-ended entitlement may be the better approach in some settings. A second tier of care isn't bad per se if it is done right: narrowly targeted, adequately funded, and continuation made contingent on achieving measurable outcomes. [5]

Question 3. How will this option

address the issue of the uninsured

(and, perhaps, the underinsured)?

The Employee Benefit Research Institute's most recent estimate of the number of elderly persons uninsured is 34.4 million in 1989, more than 85 percent of whom are employed or dependents of an employed head of household. Indeed, of the 34.4 million, more than 5 million live in households with annual incomes in excess of $40,000, and another 9+ million are in the $20,000-40,000 bracket. In addition to the excellent demographic detail of the uninsureds EBRI has given us, this population has other characteristics we are just beginning to appreciate.

Of special importance is preliminary evidence that an extremely high rate of turnover exists among the uninsured. If we count 34 million uninsured today and again 12 months from now, our policy approaches to that situation should differ substantially under two alternative scenarios: Scenario I--each count includes 29 million people who are the same people each time; Scenario II--each count includes only 5 million of the same people. Research under way by Swartz and McBride at the Urban Institute indicates that the second scenario may be closer to reality. They suggest that about half of all uninsured spells end in four months or less, and only 15 percent of those spells last longer than 24 months. A recent Bureau of the Census study by Nelson and Short reports that, during a 28-month period, more than 63 million citizens had at least one month without health insurance, evidence wholly consistent with the work of Swartz and McBride.

Short uninsured spells are hardly surprising in a mobile society where changes in employment are common. Most people do not work forever at low-paying jobs having minimal fringe benefits or for employers not providing health insurance. And employees with coverage also change employers in a world of employment at will, layoffs, and upward mobility. Front-end periods in new jobs are typically probationary and are often without fringes, and front-end insurance policy limitations for pre-existing conditions may also produce coverage interruptions.

Mandating that all employers provide coverage identical to existing coverage would fail to capture many such uninsureds, who will resume coverage soon anyway, but would significantly increase burdens on many small businesses and labor-intensive low-wage businesses. And mandating employer coverage that would do away with the short-term reasons for existing coverage interruption would impose an even greater labor cost burden on many more employers.

As labor cost increase, businesses' choices are to respond by doing one or more of the following:

* Pass the cost on by raising prices, jeopardizing domestic and international market share, employment, and survival of the business.

* Offset the cost with a lower cash wage (usually proscribed by the minimum wage effect) or a reduction in other benefits, leaving the worker with fewer economic degrees of freedom than before.

* Substitute capital for labor, jeopardizing employment.

* Accept a lower return on investment, making it more difficult to attract money in the capital markets and jeopardizing the long-run survival of the business.

* Recognize that necessity is the mother of invention an conceive of heretofore undreamt ways to become even more efficient. (I wish we could count on this last one, but it is probably the least likely of the five to occur as a response.)

A. Foster Higgins and Co., Inc., has recently reported that the average total cost of health plans per employee (including dependent coverage) in 1990 was $3,217, or about $1.55 per hour. Assuming some of that is overutilization and assuming further that we don't want to provide existing uninsureds with anywhere near the average of existing employer-provided coverage, we might be able to get by with a $0.50 per hour cost, and at least raise some uninsureds (if not their dependents) to an underinsured status. For a service industry employer with a near-minimum-wage work force representing only half of total costs and with a 2.5 percent margin on sales, that $0.50 per hour would be a 10 percent increase in labor costs, a 5 percent increase in total cost, and a 195 percent reduction in margin (i.e., instead of making $25.00 on each $1,000 if sales as it did prior to mandatory insurance, it will now lose $23.75 on that same $1,000 of sales), absent one of the responses listed above.

Are these difficulties with mandated employee coverage an argument in favor of full-blown national health insurance? They are only if one is willing to turn the existing system upside down for about 220 million people in order to provide minimal insurance for:

* 20 or 30 million persons who are receiving and paying for their own care now and who will end their uninsured spell in a few months anyway.

* Another five or so million, many of whom rationally choose to purchase their own care directly and many of whom, while genuinely in need of access, may be helped most effectively by direct service targeting.

Question 4. Will this policy remove

involuntary cross-subsidization from

the private sector and restore it to

being an explicit governmental


One of the fundamental functions of modern government is wealth redistribution, typically for the purpose of achieving various social goals. (This redistribution is typically done between groups on a current basis, though recently it has also been done on an intergenerational basis.) Traditionally, the private sector engages in wealth redistribution on a purely voluntary basis. Only in the event that private philanthropy and eleemosynary activity are insufficient for a particular purpose and the political process establishes that purpose as a public policy objective does government exert its power to require the wealth redistribution to take place via taxation and subsidization.

In health care, however, we have strayed from this path. Existing entitlement programs' payment mechanisms, almost from inception, have created significant economic incentives for so-called "cost shifting," the collection of more monies from private payers than would have been required but for the lower payments received from the entitlement programs. Absent an all-payer mechanism, government has defaulted on its role as tax collector and subsidizer to the extent that it forces health care providers to "tax" some recipients of care (or their insurers, which ultimately means the businesses that pay most of those insurers' premiums) and "subsidize" entitlement program beneficiaries and enrollees. This is an inappropriate role for the private sector and implements involuntary wealth redistribution ineptly, inconsistently, and inefficiently. Part of any new policy should be to limit wealth redistribution to true social safety net objectives that have risen in the political process to the level of public policy and to take private sector health providers out of the business of doing government's taxation/subsidy work for it.

Question 5. For any policy that

seeks to emulate Canada's (or

any other country's) system, is

Canadian grass really greener?

Aside from the details of covered services, payment mechanics, phasein rules, etc., there are three main characteristics that are central to comparisons between the health systems of the United States and other countries:

* What share of the economy is represented by health care (and its companion inquiry, what are the health status or outcome indicators)?

* What is the rate of expansion or growth of the health care share of the economy?

* What is the public/private mix of financing of the health care share of the economy?

For the United States of Canada, the raw data tell us that, for 1989, health care as a percentage of total economic activity was 11.6 in the United States, compared to 8.7 for Canada. Like all ratios, of course, percentages have both numerators and denominators. The denominators reflect the fact that Canada's nonhealth economy has been growing significantly faster than that of the United States. If since 1975, three years after the start date of the Canadian Health Plan, the Canadian nonhealth economy had experienced growth identical to that of the United States, with no change in health economy growth, Canadian health would have been 10.6 percent of its economy in 1989 rather than 8.7%, in other words, only one rather than three percentage points below that of the United States.

Further, Canada's population is slightly less affluent and slightly less old than ours. Younger, less affluent populations consume fewer health care resources than older, more affluent populations. [6] The age difference may account for half of the remaining difference between the two countries' percentages, and I estimate the affluence factor would account for all of the rest of the differential.

As to growth in the two health economies for this same period (1975-1989), the average annual rate of nominal (real plus inflation) increase was 11.68 percent in Canada versus 11.43 percent in the United States.

Thus, even if the United States were miraculously to remove all overutilization from its health care system, and even if the access and coverage characteristics of the Canadian system were politically palatable to a majority of U.S. voters, [7] why would we desire a system in which costs as a proportion of the economy may not be fundamentally different from ours and in which costs are rising faster than in the system we now have? In addition, were we to immediately drop our total health expenditures per capita to the level of Canada while adopting the aggregate Canadian financing mechanisms, public sector spending for health in the United States would jump 25.8 percent, or about $65.4 billion per year using 1989 levels of expenditure. (The public sector finances 75 percent of all health care expenditures in Canada versus a 42% level in the United States.) There may be substantive reasons for adopting aspects of the Canadian system (though I've not discerned what they are), but doing so because it is expected to save money or lower rates of cost increases are not among them.

In sum, the primary focus of reform must be rebasing, the elimination of overutilization. This task begins at home and is achieveable in the context of a predominantly privately financed and administered health delivery system that is sensitive to the concerns raised in this article.


[1] The irony in our system, of course, is that, simultaneously, a significant minority of the population, rather than too much care (as is the case with overutilization), receives too little care. This minority is a subset of the segment addressed by Questions 2 and 3: those with constrained physical and/or fiscal access.

[2] William L. Roper, MD, of the Centers for Disease Control cautioned in testimony before the House Ways and Means Committee on April 16, 1991, that "simplistic comparison of health statistics [between nations] can be extremely misleading." Yet the same day, George J. Schieber, PhD, of HCFA noted i his Ways and Means testimony that "[w]hile... [outcome] measures of infant mortality and life expectancies at birth are too gross for making assessments about the performance of our health care system, they are quite discouraging when taken in the context of what we spend."

I concur with Dr. Roper in that I wonder what such statistics would look like if they were controlled for intercountry differences in births to chemically dependent central city teenagers, young male black-on-black violence, etc. My suspicion is that much of the U.S. outcome differentials or shortfalls reflect uniquely U.S. factors relating to drug abuse and handguns. I am in no way suggesting that these social problems be ignored or minimized, but it is grossly unfair to accuse our health care system of poor performance because of situations it did not cause and that are not problems in the countries being used for health comparisons. Once the crack baby reaches the neonatal intensive care unit or the gunshot/knifing victim reaches the emergency department, it's a bit late for the health system to compensate for our inability to control illegal drug use or our failure to adopt meaningful gun control.

Even though I believe our outcome statistics would look far better after adjustment for such nonhealth system societal factors, I am also convinced that on the input side we do have substantial overutilization, as evidenced by the RAND Health Insurance Study and Medicare's experience with hospital length of stay under prospective payment.

[3] Not only will volume performance standards encourage greater intensity generally, but, as a matter of the most basic economic principles, RBRVS rate changes across CPT codes and the attendant effect on enrollee copayments will tend to reduce fiscal access to primary care (where rates and copayments are being increased) while expanding fiscal access to procedural medicine by lowering rates and copayments. One of the earliest warnings of this phenomenon is found in Long, H., "Do These Guys Know Why They're Applauding?" Physician Executive 16(2):16, March-April 1990.

[4] As Robert H. Brook, MD, of RAND noted in testimony before the House Ways and Means Committee on April 17, 1991, the RAND Health Insurance Study gives us different conclusions and implies different policy approaches for persons both poor and sick.

[5] In many settings, merely offering care is insufficient. What is required is caring enough to reach out with care that is accepted by the patient. This means adapting to local customs, values, mores; overcoming language differences; seeking out persons who could benefit from care; and doing what is necessary to ensure ongoing provider influence or involvement after the face-to-face provider-patient encounter ends. A physician in an urban ghetto or in a clinic on Native American land is of little value if no one comes or if advice and prescriptions are ignored as soon as the office visit is over.

[6] Our own data support differential consumption by age group. For the relationship between affluence and health spending, see Exhibit 8 to George J. Schieber's testimony of April 16, 1991, before the House Ways and Means Committee.

[7] U.S. citizens will almost certainly prefer the freedom to choose from a wider range of services, conveniently available, even if at higher cost, rather than opt for a single, uniform, choiceless, publicly administered, Canadian-type system. The literature is replete with examples of the effects of full or partial nationalization of health care. Among these are long waits for cardiac surgery (Isher, J. "A Comparison of Expenditures for Physicians' Services in the United States and Canada." New England Journal of Medicine 324(8):566-7, Feb. 21, 1991, and Breo, D. "Universal Medicare: The Canadian Experience." American Medical News 32(10):3, March 10, 1989, for major joint surgery (Breo, op. cit.), and for surgery generally (Whitner, C. "British Health Service, Much Beloved but Inadequate, Is Facing Changes." New York Times, June 9, 1991, Section 1). In addition, extended lengths of stay producing excessive inpatient days per capita are also typical (Schieber, J., and Poullier, J. "Datawatch: International Health Spending and Utilization Trends." Health Affairs, 7(4):105-12, Fall 1988), and limited choices of services once admitted (Friedman, E. "Treatment of End-Stage Renal Disease in Canada." New England Journal of Medicine 324(15):1078-8, April 11, 1991). Such a system can result in private funding up to 17 percent of all elective surgery, 70 percent of hospice care, and virtually all nursing home beds (Lister, J. Proposals for Reform of the British National Health Service." New England Journal of Medicine 320(13):877-80, March 30, 1989). currently, 3,500 Ontario residents are choosing the United States for drug/alcohol treatment alone, and costs of the exodus from Ontario for all care increased 45 percent from 1990 to 1991 (Borsello, M. "U.S. Hospitals, Doctors Milking 'Canadian Cash Cows'". Medical Post, Jan. 29, 1991, p. 16).

Hugh W. Long, PhD, is Associate Professor of Health Care Management, A.B. Freeman School of Business, Tulane University, New Orleans, La., and a member of the Medicare Geographic Classification Review Board. This article is based in large part on testimony given April 23, 1991, before the full House Ways and Means Committee and represents Dr. Long's views and not necessarily those of Tulane University or MGCRB.
COPYRIGHT 1991 American College of Physician Executives
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1991, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:includes bibliography
Author:Long, Hugh W.
Publication:Physician Executive
Date:Sep 1, 1991
Previous Article:From quality assurance to continuous quality improvement.
Next Article:The Physicians Practice Profile: a piece of the quality puzzle.

Related Articles
How to prepare for the ABMM examination.
Controlling public health nuisances: guidelines for local health departments.
Environmental Hazards: Toxic Waste and Hazardous Material.
Research Paper Writer.
Bibliography as an interdisciplinary information service.
Zora Neale Hurston: An Annotated Bibliography and Reference Guide.
The Drug Problem: A New View Using the General Semantics Approach. (Books).
Research utilization web resources.
Serving Latino clients with addictions: using best practice models when providing assistance to Latinos suffering from addiction can ensure that...
Science literacy: a collaborative approach.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters