|
|
| |||||||||||||||||||||||||||||||||||
A government-run system promises administrative simplicity but lacks the tools to slow the rate of growth in costs. Chronically underfunded public agencies are not equipped to assess rapidly changing medical technology in a timely manner. Manufacturers and other special-interest groups will continue to dominate the decision-making process. Since Medicare has been unable to slow the rate of growth in costs, how would a single-payer system be more effective, given our political system, especially when the stakes for special-interest groups are even higher?
Encouraging competition among health plans, even if one of them is "public," will also fail to solve the cost problem. With the exception of highly integrated organizations, such as Kaiser Permanente, health plans have only two tools to control costs: financial disincentives for patients and fee reductions for providers. Acceptable out-of-pocket maximums, however, vitiate economic incentives to restrain use, particularly for expensive care such as inpatient care. Unable to alter provider behavior, health plans primarily try to avoid enrolling people who are likely to need costly care.
The Universal Coverage Pool
An alternative approach is to combine universal coverage for care that requires insurance (i.e., hospital care and ongoing care for chronic illness) with maximal freedom for both patients and providers through a reorganization of economic incentives in ambulatory care.3 Under this plan, everyone would be enrolled in a universal coverage pool.
Hospital care (and similar outpatient-based interventional services) combined with the management of chronic illnesses accounts for more than 60% of all medical care expenditures. The proposed pool would pay directly for inpatient episodes — that is, some preadmission and postdischarge care and all associated physician services — in ways that would create stronger incentives for providing efficient and high-quality care, as described below (Table 1). In contrast, ambulatory care would be economically organized around primary care providers. After monthly payments were made from the pool to cover the costs of managing chronic illness, the premium for each patient would reflect the practice styles and fees of the primary care provider chosen and of the other providers to whom the primary care provider usually refers patients. Financing could be fully tax-based or could build on the existing employer-based system, recognizing that employer contributions are part of overall compensation. Mandating coverage by employers is politically difficult and economically inefficient. Individuals could be required to enroll in the pool and to pay age- and sex-based premiums, with income-based subsidies to address the issue of affordability. Employers could continue to offer their own plans with similar benefits, or those plans could buy into the universal coverage pool for those benefits (Figure 1).
|
|
Instead of paying individual providers for each specific service, payments should be bundled to make clinical and economic sense, with different approaches for inpatient services and ambulatory care.
Inpatient Care
Physicians involved in inpatient care could collaborate with their hospitals to create new entities, called care delivery teams, and to determine their membership and governance. In the early days of Medicare, hospital-based physicians such as radiologists and pathologists strongly resisted having their fees bundled into diagnosis-related groups (DRGs), fearing they would have little bargaining power vis-à-vis the hospital administration. Patients, however, seek out surgeons and medical specialists, not hospitals, so in creating these new teams, physicians will have significant power. The care delivery team would receive from the pool a bundled payment to cover preadmission workup costs, facility and professional services during the hospital stay, and some postdischarge care. (Ambulatory care physicians could provide services for the team at their usual office-based rates.)
Responsibility for the postdischarge period could be tailored to the specific condition (e.g., including care in a skilled nursing facility for some conditions). It would also include the costs of readmission shortly after discharge (e.g., within 7 or 30 days). Cases involving complications that require readmission are relatively infrequent but can be very expensive, presenting a risk too great for individual hospitals to bear. Medicare currently offers outlier payments for extremely expensive cases and pays separately for readmission in such cases — a practice that fails to add incentives to improve the quality of care. In contrast, care delivery teams would use part of their bundled payment to buy reinsurance against such events, and teams that learned how to reduce complications would benefit from lower premiums.
Unlike current Medicare DRGs, which reflect the average performance of providers, this expanded DRG payment would be set to cover the resources used by teams that achieved superior patient outcomes. Teams would decide how to allocate funds to the participating hospital, to physicians, and to any subcontracting providers. With a fixed payment per case, the team has incentives to demand better information about the resources and processes being used, thus improving outcomes at lower cost. Firms offering reinsurance coverage will seek to learn the best processes and offer that information to their clients.
Care delivery teams resemble physician-owned specialty hospitals with respect to incentives for efficiency.4 To avoid the current system's potential for self-referral, however, the physicians on the team would not be able to recommend admissions but they could decline admissions. A community-based cardiologist, for example, would recommend admission for cardiac care, and the interventional cardiologist, on behalf of the team, would decide what specific interventions might be necessary. National payments for specific, expanded DRGs would float up or down on the basis of the resources used by the best teams. The relative margin for each type of episode would be similar, eliminating the current need for cross-subsidies among DRGs. Teaching hospitals will claim they have sicker patients and may put forward evidence for better risk adjustment and the variables to do so. Free from congressional politics, the pool would readjust payments within categories on the basis of such evidence. Individual mandates and income-based subsidies would yield universal coverage, thus eliminating bad debt. Graduate medical education would be financed through direct grants to training programs rather than through subsidies to hospitals.
Payment for care delivery teams would come from the pool. The incentives for efficiency and quality would come from episode-based payments. This would eliminate the need for the external rules and guidelines that counter the problematic incentives of the current payment system, which encourage excessive use of services and hinder coordination across providers. The pool would establish advisory committees comprising clinicians and members of the public to identify the desired outcomes for each expanded DRG, such as survival 90 days after acute myocardial infarction or improvement in mobility 90 days after hip replacement. The pool would gather these outcome measures, adjust for coexisting conditions, and give each team its risk-adjusted scores. Payment for each expanded DRG would reflect the average resources used by teams with risk-adjusted scores above the median. As their professionalism leads teams to seek better outcomes, economics will lead them to become more efficient. If better outcomes are more expensive, the pool will pay more for them. Process indicators are important in showing clinicians how to change their practices, but an emphasis on outcomes will lead to continually improved processes. Inpatient care routinely brings together a group of physicians who share in a patient's care during an acute episode of illness. With the rise of the hospitalist, these groups have become smaller, making them well suited to the concept of a care delivery team that receives a bundled payment.
Ambulatory Care
The organization of ambulatory care is quite different from that of inpatient care. Within a geographic area, patients use widely varying networks of both primary care and specialty physicians.5 The complexity of these implicit referral relationships has been responsible for the failure of most managed-care organizations. A new payment system for ambulatory care can be centered on the individual patient's primary care physician. Medicare data suggest that patients often avail themselves of a wide range of physicians,6 and this can continue to be the case; however, patients would be asked to identify a primary care physician as a "medical home." Each primary care physician would select one of several payment intermediaries to handle billing. The intermediary would offer a "health credit card" to all patients of that primary care physician, and the card would be accepted for all medical services by any provider. Through the claims-paying process, the intermediary would note which patients had chronic illnesses and would receive monthly payments from the pool for the average monthly costs of ongoing management of such conditions. (Hospitalization costs, however, would be paid directly by the pool and would not be included in this payment.) This concurrent risk adjustment would eliminate the problems inherent in the prospective risk-adjustment approaches, which are widely recognized.7
Most patients would purchase some coverage for ambulatory care from the payment intermediary to average out their monthly costs for care. Overall, minor acute care accounts for about one third of all costs, and preventive services cost an additional 4 to 5%. Together, these add up to less than the total now spent on deductibles, copayments, and services that are not covered. The intermediary would have no need to determine what fraction of an office visit was for a minor acute problem as opposed to managing a chronic illness. The monthly premium is simply the age- and sex-adjusted average for patients of that primary care provider, after subtraction of receipts from the pool for management of a chronic illness. With the costs of hospitalization and chronic illness removed, the remaining costs vary relatively little according to the patient's age. Some patients will choose high deductibles with lower premiums for their ambulatory coverage, whereas others will prefer only nominal copayments but with higher premiums. However, premiums for the patients of a specific primary care provider will reflect the services they collectively use and the fees charged; it is not the patients' illnesses that will drive premiums but rather how their illnesses are managed.
Economic incentives for patients would therefore focus largely on the differences among physicians in terms of the premiums for the ambulatory care that these practitioners provide and recommend. Physicians can charge what they want for different services at rates they individually determine. Some patients prefer specialists and high-technology testing and imaging; others prefer office visits with their primary care physician and extensive communication. It is not clear which of these two options is more expensive or offers better care. If patients think that another physician can provide better value — that is, the desired quality at a lower net premium — they will switch providers.
The payment intermediary has no reason to control fees or deny services; it simply passes costs on to the patients of each primary care physician, charging a nominal claims-processing fee similar to that imposed by credit-card companies. The intermediaries, however, would compete for the primary care physician's business by offering information on how to provide the kind of care patients value while increasing the physician's income. This may mean substituting longer visits for quick but expensive referrals, identifying specialists who offer their time and expertise instead of expensive imaging and tests, and easy-to-use information on what drugs are most cost-effective for specific indications. Intermediaries are likely to offer (either directly or through a contractor) electronic clinical-decision support services for independent practitioners.
Availability of Data
Because information is what economists consider a public good and the system receives public money through taxes or tax subsidies, all entities receiving the pool's funds (care delivery teams and payment intermediaries) must agree to have their transactional patient care data included in a national data set. Initially, only claims data would be included, but as clinicians request better risk adjustment for outcomes assessment and payment, they will voluntarily add laboratory values and other measures to the database. For analytic purposes, the pool would link these data and provide files from which patient and provider identifiers had been removed. Confidentiality provisions similar to those of the Health Insurance Portability and Accountability Act (HIPAA) would apply to recipients of such data, easing the burdens on providers of the data. Those requesting the data would have to make public the methods used to obtain their findings, eliminating "black box" decision models. With the data no longer controlled by health plans and industry, independent researchers, professional groups, patient advocates, and others would be able to identify what works best. People may differ on how they define "best" — for example, they may have different views on the value of extending life versus the value of improving the quality of life — but such debates will take place in public.
Not More of the Same
Proposals to simply extend coverage through a single-payer system or to have people choose among competing health plans will do nothing to change the underlying incentives. Patients do have important roles in their health and medical care. Patient behavior is critical for preventing illness; adherence to appropriate regimens is necessary to achieve the best outcomes. One-size-fits-all economic incentives for patients, however, are ineffective in shaping care and may keep people from seeking care when it is most appropriate.
Insurance is supposed to pool risk, to have the healthy among us share in the costs of caring for the sick, and this system is particularly necessary for those who require expensive care for acute and chronic illnesses.8 A universal coverage pool accomplishes this goal of spreading the risk. When premiums are set by health plans (or a single payer) and payments are averaged across all providers, there are no incentives for providers to improve the efficiency or quality of care. This effect can be seen in the geographic variation in Medicare expenditures, which differ by a factor of three across geographic areas, with no evidence of better quality in the higher-use areas.9 In contrast, bundling payments to care delivery teams for inpatient care and having ambulatory care premiums reflect the practice styles of primary care physicians and the specialists to whom they refer their patients will create incentives for both efficiency and effectiveness. A commitment to pay for the resources needed by those who achieve above-average outcomes offsets simple pressures for cost reductions. Pooling and making publicly available the vast quantities of data routinely generated by the health care system, combined with incentives for providers to provide high-quality, efficient care, will foster a demand for converting such data into information that is useful for improving clinical practice.
Clearly, it is not enough to envision a future system without addressing transition issues, but a more complete discussion is beyond the scope of this article. Although the changes proposed here would transform the current delivery system, initial steps can be taken by Medicare to expand the DRG payment system in order to include physicians and hospitals that are willing to form care delivery teams. Early voluntary participants should be assured that if Medicare fee-for-service and DRG payments fall, their payments will not also fall, but that if conventional payments rise, their payments will also rise. Teaching hospitals that join this effort may be relieved of many costly compliance regulations.10 Medicare should make its fee-for-service data available to a new administrative entity created to release data to researchers (without patient or provider identifiers), thus facilitating the development of analytic tools. Health plans that contribute their data to this entity will also have access to it in order to develop alternative payment systems. Physicians can then view those data to see whether they are low-cost providers and can identify themselves as such to potential intermediaries to be eligible for the alternative payments.
Physicians chafe at managed care, complain about arbitrary fee constraints, and resist externally imposed practice guidelines. But these tools will continue to be used if physicians, either individually or in small groups, do not take on more financial responsibility for the choices they make on behalf of their patients. The system outlined here (and in more detail elsewhere3) absorbs the risks of unpredictable events as well as the variability in patients' needs and preferences. It can also deliver the information needed by clinicians to accept such responsibility and, in turn, allow them the freedom to provide high-quality care at sustainable costs.
No potential conflict of interest relevant to this article was reported.
Source Information
From the Palo Alto Medical Foundation Research Institute and the Philip R. Lee Institute for Health Policy Studies, University of California San Francisco, San Francisco.
Address reprint requests to Dr. Luft at Research Institute, Palo Alto Medical Foundation, 795 El Camino Real, Palo Alto, CA 94301, or at lufth{at}pamfri.org.
References
| |||||||||||||||||||||||||||||||||||
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
