The Oregon Rationing Experiment
An early attempt to ration health care was the state of Oregon's Medicaid program that went into effect in 1994. Oregon officials had determined that many of their citizens were going without needed medical care because they were uninsured and not eligible for Medicaid. They also knew that it would be impossible to provide every medical service for every uninsured person in the state, so Oregon policymakers proposed a plan that offered health coverage to those without insurance by enrolling them in a revised Medicaid program. The only way to pay for this new group of people was to limit, or ration, care.
In 1989, the Oregon Health Services Commission was formed to create a list of approximately 700 medical conditions and their treatments. The list categorized the medical services as either "essential services" that prevented death with full recovery (for example, appendectomy), or as "very important services" where the treatment improved quality of life (for example, migraine), or as "services valuable to certain individuals" (for example, infertility services), or, lastly, whether the treatment had only "minimal effect" (for example, end-stage cancer). Oregon Medicaid was then going to pay for treatment beginning at the top of the list covering essential services (such as the appendectomy) and continue funding medical care on down the list until the state Medicaid money ran out at the end of each year.
We have a lot to learn from what went on in this Oregon rationing experiment. First of all, no one could agree on how to do the ranking. They began using straightforward cost-effectiveness analysis (discussed in our March column), but they could not agree on what should be considered "essential" and "very important." To "fix" the problem, they used focus groups made up from the general public to decide which services would be on the list. Interestingly, this selection process caused problems because two-thirds of the members of the focus groups were health care practitioners who often thought that the most important conditions to pay for were those most related to their area of practice. In time, and to make up for a lack of a broad consensus, the commission produced the final list by doing what it thought was right. However, when the Oregon plan with its list of prioritized covered benefits was submitted to the federal government for approval in the summer of 1992, the plan was rejected on the grounds that it discriminated against people with disabilities (for example, highly expensive treatments for incurable cancer, the final stages of AIDS, or extremely premature infants with virtually no chance of survival) and violated the federal Americans with Disabilities Act. Nonetheless, two years later, under a different administration in Washington, the plan was approved.
The lesson of the Oregon experiment is that true rationing is easier in theory than in practice, given politics, patients and physicians who want to make the system work only for them. Shortly after the program began, the system began to unravel. This failure resulted, in part, from the federal government requiring coverage of almost all services for certain groups (such as those with disabilities) as well as practitioners figuring how to get around the "rules," such as coding treatments for patients with multiple medical problems under categories that were covered even though they all might not have been covered individually. As a result, Oregon's rationing system did less rationing than expected, and in turn, provided less cost savings than projected. Net savings were only 2 percent after five years.
Rationing as a Lack of Access and Coverage
But, in truth, the United States does ration. The ultimate rationing tools in the U.S. are lack of health insurance coverage and lack of access to needed medical care. As we discussed in our January column, "coverage" is defined as having health insurance, whether paid for by the employer, the government (Medicare, Medicaid, state or county hospital) or individually. We ration health care in this country by allowing 46 million people to go without any insurance coverage.
We are not alone; every country in the world rations, but each in a different way. Many countries have a limited list of drugs that government insurance will pay for; if the drug is not on that list, there is no payment. Some countries limit these drugs "because their effectiveness has not yet been proven" -- in other words, they are still experimental. In England and Australia, mathematical cost-effectiveness analysis is actually used, just like the design of the Oregon rationing experiment, and some drugs are not available because their effectiveness does not justify the cost. The most open form of rationing has been used in certain European countries where, for example, dialysis was not available to anyone over the age of 65. This rationing practice has been replaced by more quiet agreements between families and physicians that consider the patient's quality of life. Such discussions seem to be less common in the United States because our national values require a great deal of medical care, even at the end of life.
What Can We Do?
How will we deal with the ever-increasing availability of new diagnostic tests and treatments that are increasingly expensive? In the short term, we will make every attempt to avoid explicit rationing.
We believe that eventually there will be guidelines to decide what should be covered. Ideally, the services that are covered should provide a reasonable baseline for every American, one that we would be proud to tell our friends around the world about, but one that is not excessive and with reasonable time to see a doctor and have procedures. Appropriate medical services would cover those treatments for which there is evidence that the service is beneficial (so called "evidence-based benefits"). We must do a better job of getting the evidence. Fortunately, the stimulus package has provided $1 billion for comparative effectiveness research.
However, even the list of truly beneficial services is likely to be unaffordable at some point, and then the really hard work will begin. We will need to consider not only effectiveness but also cost. Because it is important to take such decisions out of the realm of politics, a completely independent group, like the Federal Reserve, may need to be established. Broad decisions regarding the insurance coverage of essential medical services cannot be reached by individuals, either by the physicians treating patients or by the patients themselves, since, at the individual level, everyone will always want everything. However, this doesn't mean that physician and patient input is not important: in determining what is "essential," we will need to gather information from many types of health practitioners as well as patients, the general public, ethicists and social scientists to help develop the methodology and conduct the research to assess the medical evidence. Whether it is more important to maximize the benefit to a single identified patient who is ill (e.g., "my son") or whether it is more important to maximize the good of a larger number of people (e.g., "the community"), with the same number of life years saved, are examples of the kind of questions and trade-offs that will have to be discussed.
Because this is America, those who can pay will be able to buy more care than those who cannot pay. This disparity is not unethical as long as we define an "adequate" baseline of care and provide it to everyone. Unfortunately, the definition of "adequate" depends upon where you stand. That is why future rationing decisions require good science, professional judgment, highly ethical approaches and public input. No matter what, we will eventually need to come to grips with the fact that we cannot all have it all.