Can new payment methods in health system affect quality of care?


By Francisco Londono, August 10, 1998



    For the purpose of this article I will make some generalizations and I will define some words to facilitate the reading and understanding. These definitions are not intended to be exhaustive and must be understood only in the context of this article.

    In America health has had its own evolution, passing from the personal relationship between a physician and a patient to a complex system with many actors. As technology developed, on one hand, the costs increased and patients or their families weren’t able to pay by themselves. As a consequence, new payers, such as Government and employers appeared in the health industry. But once again, one treatment could be so expensive, that the resources of a small employer wouldn’t be enough to cover it, and his business could get in financial risk. Consequently, the typical insurers began to play their own role: The affiliation of large number of people paying a fixed premium per person and period of time, regardless the cost of the treatments needed by each of their affiliates.  A patient could choose the provider, pay the treatment by itself, and later the insurer reimbursed him the cost of it. If the number of affiliates is high, the probability of a high cost treatment becomes more standard or predictable and the excess of money the insurer earns with people who pay and don’t get sick can absorb its costs. This is known as the “big numbers” law.

    On the other hand, physicians became more specialized, and needed more technology not affordable on an individual basis. Now we have physicians, nurses, hospices, clinics, hospitals and complex systems joining all them in order to provide the care needed by patients. For the purpose of this article I’ll call all of them providers.

    Cost continued increasing, the relations between these actors continued changing, and the characteristics of each of them too. In the side of the insurers, in the 1980’s, the Health Maintenance Organizations or HMO’s appeared. Despite their differences, in the beginning most had similar characteristics: they were non-profit organizations providing care to their affiliates with a selected net of providers and special rules and procedures that patients and providers should follow in order to accept the service and pay the provider.
    Recently, in the search to achieve the key objective of cost containment, the payment systems used have also changed. At the beginning, the predominant way used by payers and insurers to pay providers was a fee for each service they provided to a patient (fee for service – FFS). The fees were defined sometimes by the provider, sometimes by the buyer and other times by a negotiation between them. FFS payment system had implied an economic incentive to providers in the form of being paid more as more services were delivered to patients. Different efforts were made to discourage the excessive cost. One of the best known, and still in use, is the Diagnostic Related Groups (DRG's), where the provider received a fixed amount of money for the complete care of a patient with certain diagnose, despite the resources used in the care of the patient.  This helped, but was not enough.
    Recent changes show that Medicare, to reduce its expenditures, is shifting toward the affiliation of more people to health care organizations (HMO’s). Medicare is paying HMO’s with a fixed amount of money per person and period of time (per capita), equivalent to 95% of the average cost of a Medicaid patient, instead of paying providers directly. In this way Medicare could decrease its expenditures by 5%.

    At the same time, in order to keep their costs down, HMO’s are using new payment methods too. In addition to the traditional FFS and DRG systems, they are sometimes paying providers in a per capita mode, assigning a group of providers a group of affiliates and paying them a percentage of the premiums collected from the affiliates. Sometimes they only make contracts for some services so the per capita payment is rated according to the scope of the services contracted. Other times, HMO’s give economic incentives to reduce referrals, prescriptions or expensive treatments to patients. If the total expenses of one group of affiliates exceed an HMO established quantity, the providers are penalized and receive less money or if they expend less than expected, they are better paid. The principal arguments of the HMO’s to do so are: first, costs must be reduced, which is generally accepted; second, the physicians are the critical decision makers in term of costs, so they should have strong (economic) incentives to reduce them; and third, it has been proven that many services are not always necessary or are not cost-effective. Even though not all HMO’s are the same, and not all of them use these kind of payment systems, I’ll use the term of HMO’s in this article to refer only those who have these practices.
    Next you will find a simplified diagram that depicts how all this actors are integrated in a real more complex system than that old one of the patient and his physician:

Principal questions about costs and quality implications

The purpose of this article is to analyze if these changes can affect the costs and quality of care delivered to patients. Not all aspects of costs and quality will be covered but some of the principal questions that have emerged, some of the findings that have been done in the USA by some researchers, and some conclusions will be given. The principal questions are:
      According to the importance of each question, all of them will be addressed in turn.

Are these new payment systems really helping to reduce costs?

    As the key objective of the new changes is cost control, we must ask if the objective is being reached. Some findings directly related to these new trends, show that they can really do it. A few of these findings are:  

Whose costs are they helping to reduce?

    As we saw in the last question, in many cases employers and self-payers are finding savings or at least slow increments in their premiums. Some researchers have found that the 5% cost reduction expected by Medicaid is not going to happen. The principal reason for that conclusion is that in some cases, HMOs are doing favorable selection to enroll healthier people (Policy Information Exchange Online-PIE, 1996). In other terms, even though is forbidden to reject the admission of a Medicaid beneficiary in an HMO plan, they are targeting their marketing strategies toward healthier people, such as younger people with good income. Bruce C. Vladeck, the Health Care Financing Administration's  (HFCA) Administrator commenting on the study ‘Health Status of Medicare HMO Enrollees in 1994’ done by the HFCA, said “This study clearly shows there still may be a problem of risk HMO’s having healthier Medicare beneficiaries.” Later he also said “These findings confirm again the need for a payment method that adjust for the health status of Medicare beneficiaries who enroll in HMO’s.” (PIE, 1996). So, is Medicare going to save money this way? Probably not, because the care for sicker people, and thus people more expensive to care for, is still being paid directly to providers in a fee for service (FFS) way, and their particular average cost must be higher. So, healthier people costing Medicare 95% of actual expending average, and sicker people costing more than the average, could make Medicare total cost rise. As Salins says, HMO’s are “skimming the cream”. In this way, the HMO’s can keep costs down the average and make profits. It is also known that some HMO’s have been changed from no-profit to for-profit organizations.

Do they affect the quality? How can they affect it?

    A primary question is how can these new ways of relationships, with economic incentives to providers to control costs, affect the quality of care? Before answering, a new question must be asked: What is quality of care? There is no unique definition, but most authors recently included in their definition aspects that not only concern the medical act itself, but also the patient ease of access to the services needed and the general satisfaction expressed by the patient itself. For the purpose of these article I’ll only analyze these two factors.

    Research has been done recently on this topic and there is not enough evidence to have definitive conclusions (Hellinger, 1998:840). But some clues have appeared. As one critical aspect of quality is ease of access, we must answer these questions: Is accessing in HMO’s easier than that in typical insurers? Are the HMO’s accepting or denying treatments recommended from providers to their patients? Can patients use their preferred provider? In his revision of recent studies about ease of access to care, Hellinger (1998:838) shows these findings:

    Simultaneously, in recent years, as a response to the increasing dissatisfaction and public claims about bad practice, more regulation than ever has been made at state and federal levels, trying to protect the people’s rights in access and quality of care (Dickerson, 1998). Laws have been made to disclose the financial incentives that HMO’s offer physicians to control costs, to give consumers the rights to a full appeal process if denied treatment, to allow access to emergency-room care without previous acceptance by the HMO, to grant minimum lengths of stay in hospital for some treatments, to grant the continuity of plans from one HMO to another, and so on.
    As another clear answer to the permanent claims, and a way to help people to choose between the different options, the Consumer Assessment of Health Plans Survey (CAPHPS) is being developed. This study is sponsored by the Agency for Health Care Policy and Research (AHCPR), the Harvard Medical School, RAND, and the Research Triangle Institute. In this survey nearly 130,000 people are being asked about their own experience with different HMO’s or FFS plans. People are going to score their service in many ways, but the research will take time and the results are expected in November 1998 (AHCPR, 1998).

Are these new payment systems shifting the risk from the HMO’s to the providers?

    Another problem that providers complain about, is that with the per capita payment system and with the reimbursement based in an inverse relation to the cost of the attention provided (economic incentives to reduce costs), HMO’s are trying to shift the risk directly to the provider. As explained in the introduction, one of the key principles of insurance is the general law of  “big numbers”. A patient’s treatment can be so expensive that it can consume the resources of many patients. If the group of patients is small, all patients’ premiums can be consumed and the HMO or the provider loses money. If the group is large enough, then the looses are much smaller. In general, providers joined in nets or alone don’t have the same number of patients, and their risk is higher than that of the HMO’s. As Salins says, shifting the risk to the providers, HMO’s play a zero-sum game. They can assure profits and minimize almost to zero the risk of looses. In his article, Salins also suggests that as providers are taking risks, they are also beginning to do favorable selection. Some teaching hospitals that typically have searched for the sickest people in order to learn more are beginning to express their worries, because if they continue doing the same way, and are being paid with the new systems, they will no longer be able to cover their costs. So, who is going to care for the sicker people? Once a provider or group of providers has over passed in cost the reimbursement, who is going to pay for the services still needed by patients? Can the provider absorb looses or are they going to refuse treatments or services in order to survive? If the risk is going to be shifted to the provider why do we need insurers, and of course why do we need more actors to be paid in the chain of the health care industry?

Must we ask and answer ethical questions about incentives and similar issues?

    Another key issue, very difficult to be studied with numbers is the ethic question that physicians and providers have to ask themselves frequently. What is first, the health of the patient or my income? The answer seems very clear in the perspective of ethics, but what about this other related questions: Is the prescription really needed? Are there better options? Are better options more expensive? Are there other cheaper options? Do they have the same risks? Are cheaper options well tolerated by the patient? What level of treatment tolerance by the patient is desirable? What is the right definition of cost effectiveness? What is the tolerable level of effectiveness that a treatment must have in order to be accepted? Is the use of pain control treatment worth it on a terminal person? and so on. As medicine and health care are still not exact sciences, and the concept of health is not unique and universal, many questions do not have a clear answer, but providers have to make decisions on a day to day basis, now.

Do different approaches exist to find a solution that helps to reduce costs without affecting quality of care?

    One of the best contributions that the HMO’s, the insurers and the providers can make is just in progress, and all of them are working hard to find “best practices” on health care. The information that can be gathered today is not still completely used as it should be. Best practices, if managed well, can help simultaneously to improve quality and reduce costs. Health is not the exception to one of the principles of quality: Quality costs less. We must not expect an exact answer to treat each different illness in each patient, but Pareto’s law also applies to health. Most costs are consumed delivering care to patients with a few common diseases (Pareto’s law says < 20%). If we learn to treat well these few diseases, we can increase quality and reduce costs. New studies of what is called Evidence Based Medicine are in progress and they are going to help to decide what treatments are the most indicated for the most important diseases.

    Simultaneously with the support of the HCFA, new approaches combining the FFS payment system with the Group-Specific Volume Performance Standards (GVPS) are in progress. This new approach takes advantage of giving economic incentives to providers based on quality of care and the use of standards instead of using direct cost reduction incentives. This approach would permit patients the freedom to choose the provider. Also the providers could be paid on a FFS basis and be encouraged to use cost-effective service delivery patterns (Tompkins et al, 1998).


    According to the findings, accepting that the evidence is not as complete as anyone should desire, and taking into account the questions that are being made, some conclusions are:
      Finally it’s clear that the evidence available it is still not enough to make definitive conclusions. However life is going on, decisions are being made and must be made. Just with the scarce information available today we have to make the right decisions, trying to decrease the risks involved, and at the same time answering, in the most acceptable way, the questions that have emerged.



Agency for Health Care Policy and Research (AHCPR). (1997).
Consumer assessment of health plans (CAHPS): Fact sheet.
Available: [1998, 16 July]

Dickerson, John. (1998). Lets play doctor. Time, 152 (2), 28-32.

Hellinger, Fred. (1998). The effect of managed care on quality: A review of recent evidence.
Archieves of  Internal Medicine, 158, 833-841.

PIE Online. (1996). Medicare risk HMOs experiencing favorable selection.
Available: htto:// [1998,7 July].

Tomkins, C; Wallace, S.; Bhalotra, S.; Chilingerian, J.; Glavin, P.; Ritter, G &
Hodgkin, D. (1996). Bringing managed care incentives to Medicare’s fee-for-service
sector. Health Care Financing Review, 17 (4), 43-63.

Salins, Craig. The new faces of “Managed Care.” Available: [1998, 2 July].