HEALTH
COST CONTROL, DOCTORS' ETHICS, AND PATIENT CARE
PROLOGUE: The costs of medical care depend critically on doctors. They order the tests, prescribe the drugs, and decide whether patients are admitted to hospitals and when they are released. However, the major new Medicare reform, designed to curb hospital expenses, applies only to hospitals, not physicians. Under the prospective payment system (PPS), hospitals are now being paid a fixed price for each Medicare patient, determined on a disease-by-disease basis. Yet doctors are still reimbursed retrospectively on the basis of “usual and customary” fees.
In this reflective essay, Arnold S. Relman, a physician and editor of the New England Journal of Medicine, points out that under this fee-for-service arrangement, the longer the patient remains in the hospital and the more services he receives, the more the doctor is paid. If hospital expenditures are to be significantly curtailed under any program, Relman argues, physicians will have to be given incentives for frugality. Yet extending the prospective payment system to doctors, as is now being considered, may have serious implications for the quality of patient care, he warns. In particular, Relman asks whether doctors can remain advocates and agents for their patients if they have an economic stake in reducing hospital services,
In this essay, I will consider how the various ways of paying for medical care affect the attitudes and behavior of physicians. This is a matter of interest not only to physicians; the public is greatly concerned about the quality and cost of medical care, and these depend critically on doctors. Although physicians receive only about 20 percent of all the money spent on personal health care in this country, their decisions determine most of the remaining expense. They order the diagnostic tests and prescribe the drugs; they recommend the surgery and perform it; they determine whether patients need to be in the hospital and for how long; and their orders largely determine how the resources of hospitals are deployed for patient care.
Furthermore, the standards set by physicians largely govern the ethics and quality of our health care system. That is not because physicians are more moral or concerned than other health care workers. Not at all. As I will remind readers again in this article, physicians are inherently no different from other citizens. They have the same strengths and weaknesses of character and are susceptible to the same economic temptations as anyone else. What gives physicians special influence is the trust reposed in them by the public and the responsibility for the care of their patients invested in them by law. This trust and responsibility, in turn, are based on the assumption that the medical profession has unique technical competence, which it is committed to employ primarily for the benefit of the sick.
Until now, physicians as well as hospitals have been paid largely through a charge reimbursement system based on insurance. The cost of the premiums has been borne mainly by government (for example, Medicare and Medicaid) or by employers. Only a small fraction of the premiums is paid by the individuals who actually receive the services. Although hospitals have increasingly been forced to accept discounted rates, physicians have usually been reimbursed by third parties on the basis of “usual and customary” fees,
The fee-for-service system has at least three disadvantages for the public, First, the fee scales in most states reflect the interests of specialists, particularly surgical specialists, As a result, they are often too high for surgical operations and invasive diagnostic procedures but not high enough for primary care. This skewing of relative values accounts for large differences in income among different specialists and between specialists and primary care physicians. The current undersupply of primary care physicians and the growing oversupply of certain types of specialists cannot be unrelated to the incentives of the reimbursement system. Second, fee-for-service is piecework reimbursement and, like any piecework system, it provides powerful economic incentives to increase output—that is, to increase the number of services provided by physicians. The net effect is to encourage the performance of expensive technical procedures in borderline situations in which more conservative, less expensive options might be just as acceptable. Third, the fee-for-service arrangement creates a conflict of interest for the physician. Although doctors are supposed to be agents and trustees for their patients, the economic rewards of fee-for-service provide incentives for them to recommend services that may not be necessary or cost effective.
Before World War II, in an earlier and simpler era, the conflict of interest inherent in the fee-for-service system was in some ways more clearly drawn than it is now, but it was also less troublesome. Most physicians were generalises, and most patients paid what they could afford for their own care. Because the patient usually had to pay for medical services himself, the physician's economic incentives were in a sense more sharply in conflict with his professional duty to act in the patient's interest. On the other hand, the conflict was softened by the widely acknowledged obligation of the physician to adjust his fees to the patient's ability to pay and by a personal relationship between doctor and patient, which discouraged the doctor from exploiting his position. Furthermore, because the ratio of physicians to patients was much lower than it is now and most physicians had as many patients as they could handle, there was little economic reason to do more for any one patient than was clearly necessary. And even if they had wanted to improve their economic position, practicing physicians had limited means to do so because there were few specialists and because sophisticated medical technology was not widely available. Until fairly recently, primary care physicians had little more than their time and advice to offer.
All of this has changed over the past few decades as the technology and organization of medical practice have become increasingly specialized, complex, and expensive. Most doctors are now specialists, and the practice of medicine, even by nonspecialists, commonly involves tests, procedures, and many technical personnel, thus weakening the personal bonds between doctor and patient. In an era of specialists and high technology, the doctor-patient relationship tends to become perfunctory and transient. Patients now see multiple specialists or organized groups of physicians rather than one primary care given And this encourages physicians to take a more businesslike and technically oriented approach to their duties.
The diagnostic and therapeutic technology now available has vastly expanded the economic dimensions of medicine. Physicians now have many choices to make and many opportunities to generate income through the use of new technology in the office and hospital In addition, medical indications for the use of much of the new technology are not precisely defined, and that makes choices about its use more problematic and susceptible to influence by economic factors. The range of acceptable options in a given case is often wide enough to give the physician considerable latitude in his choice of procedures. It is in this gray zone that economic incentives have their greatest effect on medical behavior. I am not suggesting that economic considerations persuade physicians to do things that they know to be useless or unwise. I am convinced that the vast majority of physicians genuinely want to do the right thing for their patients. Often in medical practice, however, “the right thing” is simply a matter of opinion because many tests, procedures, and operations have not yet been fully evaluated or scientifically compared with other available measures for cost effectiveness.
In parallel with the growth of medical technology, and in response to the increasing expense of health care, insurance coverage has been greatly expanded. Employers now pay the premiums for their employees as a fringe benefit, not taxable to the employees, and government insures the elderly and the destitute through Medicare and Medicaid. As a result, neither patients nor physicians have been much constrained by medical costs.
Also important is the rapid growth in the number of practicing physicians, which has resulted from a near doubling of medical school capacity during the past 10 or 15 years. Within a short time, we have gone from a short-age to a growing surplus of practitioners. This has begun to increase the competition for patients and to threaten the livelihood of many physicians, who are now looking for ways to protect and maintain their professional income.
In this new climate, the conflict of interest in the fee-for-service system has taken on new dimensions. On the one hand, physicians have felt the economic imperatives and incentives more keenly than ever before. On the other, patients have been much better protected from the economic impact of their physician's decisions. Insurance has shifted the effects of the physician's economic conflict from the patient to a broader arena. The American people as a whole, rather than individual patients, have felt the consequences of the rapid rise in health costs during the past two decades. Insured patients have acted more like claimants than prudent buyers of services. They have often encouraged, sometimes even demanded, the application of sophisticated tests and procedures and have been quite willing to be hospitalized for this purpose—as long as their insurance paid the bill. Physicians, for their part, have often been unable to resist the powerful economic incentives of the fee-for-service system, particularly when their patients did not have to pay.
It is hard to know exactly how much these attitudes and incentives have contributed to the overuse of health care resources, but the effect has surely been major. I have long held the opinion, based on wide experience as a consultant and teacher in internal medicine, that more prudent choices by physicians could probably reduce expenditures for drop, tests, procedures, and the use of hospital facilities by at least 15 to 20 percent—without any loss of medical effectiveness. Lack of available information about the relative effectiveness of new technology and inadequate education of practitioners are partly to blame for this overutilization of medical resources, but the economic inducements of an insurance-based, fee-for-service reimbursement system surely play an important role.
In any event, the steady rise in health care costs over the past 10 or 15 years has finally produced an inevitable reaction. The major payers of health insurance premiums—government and private employers—have now initiated steps to change payment mechanisms and to eliminate the incentives for excessive utilisation. As recently described in these pages (Enthoven and Noll, Vol. I, No, I), Medicare began a phased transition last year to a new payment system for hospitals based on uniform payments according to diagnosis-related groups (DRGs). Instead of being reimbursed for Medicare patients on the of charges or costs, hospitals will now be paid fixed sums for each type of patient, by diagnosis. Businesses, for their part, are encouraging their employees to consider Health Maintenance Organisations (HMOs) as options to conventional insurance plans. HMOs, like the DRG system, are a form of prospective payment. For a fixed per capita fee, HMOs undertake to provide all the medical care their subscribers need, whether inside or outside the hospital Businesses are also shopping for health care bargains by seeking contractual arrangements with so-called PPOs (Preferred Provider Organizations), which are hospitals or groups of physicians offering their services at a fixed price in exchange for a guaranteed supply of patients. Such arrangements, not yet as widespread as HMOs, are still another alternative to the traditional reimbursement-of-charges method of payment.
In all of these new payment systems, the incentives for the providers are exactly the opposite of those in the traditional reimbursement method. Providers who succeed in holding costs below the level of the prospective payments are rewarded by being allowed to keep the difference. Those who cannot reduce their costs below the level of payment are penalized by being required to absorb the loss. My purpose in the remainder of this essay is to consider how physicans are likely to be affected by these new developments— in particular, how the attitudes and standards of the profession, upon which so much depends, will be influenced by these changing economic incentives.
II
Until the advent of DRGs, physicians and hospitals had compatible economic incentives. Both received higher reimbursements from Medicare and other insurers when they provided more services to hospitalized patients. In a very real sense, hospitals were the workshops for physicians—a convenient place to observe patients, conduct diagnostic studies, and provide treatment with little or no cost to patients as long as they were insured, and with little or no overhead expense to the physician. Many insurance plans, including Medicare, reimbursed for certain services provided in the hospital but not in the doctor's office. No wonder, then, that physicians were quick to hospitalize their patients and were in no particular rush to discharge them, and that patients rarely objected when hospitalization was recommended. No wonder, also, that hospital managers rarely hesitated to provide new services and facilities requested by physicians, as long as they were assured that the hospital charges would be covered by third-party payments.
Under a charge reimbursement system, hospitals have had little incentive to control costs, reduce length of stay, or limit new capital investments; therefore, they have had no reason to discourage the expenditures ordered by physicians. Investor-owned, for-profit hospitals, in contrast to not-for-profit voluntary or public hospitals, have had much stronger reasons to encourage the use of hospital resources by physicians because more utilization has meant greater profit for investors. In a sense, physicians practicing in for-profit hospitals have been the hospitals' business partners. Not unexpectedly, this arrangement led to greater income per admission for the investor-owned hospitals. All published studies of this question have shown that for-profit hospitals have collected 10 to 15 percent more revenue per admission than have comparable not-for-profit hospitals. This difference has been due entirely to charges for diagnostic tests or for special services and supplies—that is, to the services ordered by physicians and not to the basic overhead costs of room and board.
Hospital management, of course, has no direct control over physician behavior, but management can create an ambience that encourages a technology-rich style of practice, particularly among new physicians who may have been recruited to the hospital staff with special inducements such as free office rent or guaranteed income. Such favors are tacitly contingent upon continued, active use of hospital resources. I mention this here not to point an accusing finger at the aggressive marketing tactics of the for-profit hospitals— many voluntary hospitals have lately become just as aggressive in their struggle to survive the competition—but to illustrate how physician behavior can be affected by the climate by hospital management. When both physicians and hospital management have strong economic incentives to maximize the use of expensive (and profitable) technology, as has been the case in investor-owned hospitals, it is only to be expected that hospital charges will be higher.
Prospective payment of hospitals, but not doctors, drastically changes the nature of the economic relationship between these two providers of medical care. Hospitals paid a fixed sum per patient by Medicare are no longer indifferent to the resources expended in the care of patients, as has been the case with not-for-profit hospitals. Nor are they eager to increase those expenditures, as was the case with investor-owned hospitals. Prospective payment mandates a reduction in hospital input per patient—a shortening of average length of stay and a reduction in the tests, special procedures, supplies, and equipment used—for only by controlling costs per patient can hospitals expect to balance their or show a profit. (Of course, increased admissions continue to be an economic desideratum, as long as payment remains linked to the number of patients treated.) Furthermore, as the hospital surveys the prices Medicare will pay for each type of illness, it perceives that certain diagnoses are likely to be profitable and others not. Clearly, it is in the hospital's interest to encourage the admission of certain types of patients and to discourage others. Most important of all, however, the hospital must now be to control its expenditures per patient. Because more than half of the hospital bill the cost of those special items ordered by physicians, and because physicians must decide when to discharge a patient, any significant reduction in hospital expenditures will require the cooperation of the medical staff.
But what incentives do doctors now have to cooperate with hospitals in this effort to reduce per-patient expenditures? Most physicians are still paid on a fee-for-service, piecework basis. The more they do for hospitalized patients and the longer patients stay in the hospital, the more doctors are paid. The economic interests of the hospital and the doctor are now in opposition rather than in concert, at with respect to Medicare patients. Physicians have always tended to be a little suspicious of hospital administrators, often regarding them as bureaucratic impediments to the achievement of optimal care. Such suspicions can only be exacerbated under the new payment system, which requires hospital management to press the medical staff for restraint in ordering elective and services and for earlier discharge of patients. Appeals for staff cooperation on the of institutional loyalty are likely to be unavailing with fee-for-service physicians if they have no special ties to the hospital other than their staff appointments—and particularly if they have appointments at other hospitals as well, which is often the case. pressure through quality assurance review committees might be effective when a physician's practice is clearly outside established norms, but it is unlikely that such norms could be effectively applied in more than a minority of cases. Furthermore, I suspect that physicians who object to being pressured to curtail services to hospitalized patients when there is no clear evidence that the services in question are of absolutely no value could initiate legal action that might effectively discourage further pressures of this kind. The courts have often shown themselves willing to defend the rights of patients to receive, and doctors to prescribe, whatever medical services might be of help—without regard to cost.
Some observers believe that under present conditions, the most effective way to get fee-for-service physicians to cooperate voluntarily with hospitals in controlling expenditures is to offer them a share of the cost savings that hospitals may realize through the DRG system. Joint ventures between hospitals and medical staff—new corporations that allow physicians and hospitals jointly to share the risks and profits of the new prospective payment system—have been advocated for this purpose. Proponents argue that this kind of arrangement would not only motivate physicians to work together with management to reduce hospital costs, but would also, in effect, give the medical staff a share of managerial responsibility and thus help to maintain the quality of care. Such arrangements between hospitals and their medical staffs are being explored in many hospitals around the country and have already been established in a few.
III
It is too early to know how successful these joint ventures will be. Nonetheless, they raise a disturbing ethical question about the role of the medical profession. Physicians have an ethical and legal responsibility to do the best they can to ensure that their patients receive all necessary medical care. When medical needs are not clear, patients should receive the benefit of the doubt. Is it in the patient's interest for his physician to have an economic stake in reducing his hospital services? When both the hospital's and the physician's financial interests are arrayed against him, can the patient expect to receive all the hospital care he really needs? Similar questions arise when physicians working in an HMO setting receive a bonus or share in profits earned by the HMO management through reductions in expenditures on patient care. I would like to believe that a physician's professional commitment to his patients would make him oblivious to such financial considerations, but that is unrealistic. Economic incentives do count in medicine as in any other field; if we want to have physicians act as conscientious agents for their patients, we must see that the incentives are appropriate for that purpose.
Many services provided by physicians to hospitalized patients can also be provided outside the hospital. One important consequence of the change to prospective hospital payment is an acceleration in the trend by physicians to provide more services outside the hospital, where they do not have to deal with hospital managers or the constraints imposed by new forms of hospital reimbursement. Even before the advent of DRGs, outpatient and home services were increasingly seen as ways to reduce hospital utilization and make health care available in more convenient settings. Ambulatory surgery centers, walk-in clinics, out-of-hospital diagnostic laboratories and radiologic centers, and all types of home care have been expanding rapidly, particularly under the entrepreneurship of investor-owned businesses, Seeking to protect their income in an increasingly competitive environment, more and more physicians have been participating in these ventures, either as proprietors, partners, or investors. Physicians have also been adding more technical equipment to their office to enable them to conduct many tests and diagnostic procedures that might formerly have been done in the hospital Hospitals have also become more interested in diversification, and many of them (both nonprofit and for-profit) have begun to “unbundle” their services, splitting off certain profitable activities into outpatient settings or establishing new subsidiaries to deliver various types of services to ambulatory patients. They, too, have been motivated by mounting concerns that prospective payment and an increasingly price-sensitive market may threaten their ability to generate surpluses and accumulate the capital they need to maintain and modernize their facilities.
In this new climate, physicians and hospitals are becoming competitors, acting like businessmen struggling to guarantee their place in a fast-changing market. In this competition, physicians have had the advantage so far because they control the patients. This advantage may not persist for very long if the trend toward HMOs, PPOs, and other contractual arrangements continues to grow Hospitals and hospital companies may soon be making their own arrangements with third-party payers, under which they assume responsibility for the medical care of a large group of subscribers and then contract with a group of physicians to provide the care.
IV
Unless contractual arrangements of this kind come to dominate our health care system, however, any reform of hospital payment mechanisms will sooner or later have to deal with the question of physician payments. For the near future, physicians will continue to be the major influence determining the utilization of hospitals and all other health care facilities. That is why the law establishing the DRG system for hospitals also stipulated that the secretary of the Department of Health and Human Services study whether a similar system could be applied to the reimbursement of physicians for services furnished to hospitalized Medicare patients. By congressional mandate, the secretary is required to report to the Congress by July 1, 1985, on the advisability and feasibility of this proposal. This would be a piecemeal approach, of course, because outpatient services and patients covered by other insurers (as well as patients without insurance) need to be considered as well. But this study probably represents the first step in a process that has long been expected—an overhaul of our methods of paying for medical care.
The application of the DRG principle to individual physicians poses formidable obstacles and has several seemingly major disadvantages. First, it provides doctors with powerful incentives to underserve their patients. As discussed above, this antipatient incentive is also built into the hospital DRG system, but there it is counterbalanced by the traditional fee-for-service system of paying physicians. When practicing physicians become joint partners with hospitals, or are employed by them, the balance is threatened. Patients' interests are even in greater jeopardy, it seems to me, when both physicians and hospitals are paid according to the same DRG principle.
A second major problem would be the complexity of a physician-based DRG system and the difficulty of implementing and monitoring it. Hospital DRGs have problems enough, but I can foresee endless complications with any system designed to handle systematically the almost endless variety of services provided to patients by physicians. Furthermore, the DRG system assumes that individual variations will be averaged out by the large number of each type of case seen in a hospital. That may be true for hospitals, but it won't be for the limited number of patients in a given physician's practice. If inequities are to be avoided, payments would have to be made to groups of physicians, which would require considerable sacrifice of autonomy and independence by the majority of doctors who now practice in solo or in small partnership arrangements.
Finally, DRGs for physicians would even further accelerate the trend toward providing more medical services outside the hospital as doctors sought to escape the constraints of prospective payment. To the extent that it eliminated unnecessary hospitalization, this development might be welcome. However, it would undoubtedly also lead to overuse of many ambulatory services, which might end up being more costly than the hospital services that were eliminated. There is also the danger that necessary hospitalization might be delayed or avoided.
An alternative to the DRG plan is a negotiated fee schedule for physicians' services. This would be more workable, in principle, and undoubtedly more acceptable to most physicians than a DRG system. A negotiated fee schedule, however, probably offers less opportunity for the government to control total Medicare costs because it deals only with prices and not with utilization.
The next act of this political drama is hard to predict. Organized medicine will probably resist the application of the DRG principle to physicians' services and instead offer to negotiate a fee schedule. Whether the administration will accept that arrangement is not clear. What does seem clear, however, is that we are in a period of transition. The present system of payment for medical care is unstable and bound to change. So far, physicians have been part of the problem. They have, of course, been responding in predictable fashion to the perverse economic incentives of an open-ended reimbursement system. But they have not yet been able to take initiatives to help solve the dilemma. Without such initiatives from the profession, government will probably be forced to take ever-stronger regulatory action to change physician behavior. Current discussion of plans to modify the payment of doctors for the care of hospitalized Medicare patients is only the first of many steps that may be anticipated.
Arnold S. Relman received his medical degree from Columbia University in 1946. He was a professor of medicine at Boston University and was chairman of the department of medicine and Frank Wister Thomas Professor of Medicine at the University of Pennsylvania School of Medicine. Since becoming editor of the New England Journal of Medicine in 1977, Relman has also been a professor of medicine at Harvard Medical School and a physician at Brigham and Women's Hospital. He has also been a visiting professor at some 25 medical schools and president of the Association of American Physicians.