Medical Decentralization

States are not waiting for action on the part of the federal government to provide solutions to medical problems. For example, Oregon ranked medical procedures in order of effectiveness and then rationed medical care to Medicaid patients. For example, Oregon will not fund expensive risky procedures such as liver transplants. By doing this Oregon was able to extend Medicaid coverage to 100,000 near poor. Utah has created a market-oriented health plan that, if it succeeds, will provide affordable health care coverage to all state residents. Florida's Health Care Reform Act of 1992 provided premium discounts to the near poor to make health care more affordable. New York has passed legislation to regulate HMOs. Many states have created pools to make medical insurance more affordable to small business.

In this book we shall focus on medical information policy and those aspects of medical policy which can not be separated from medical information policy. States face four major information problems associated with medical policy. First, as medical records shift to the social nervous system the issues of access, security and accuracy must be addressed. Second, Conservative and Liberal states differ markedly on whether medical insurance should be pooled or based on risk. Information policy to support their decision must be created. Third, states must establish in their information policy provisions to promote medical efficiency. Finally, states must address the problem of self-insured firms' economic incentives to hire only healthy workers.

For medical decentralization to be effective the federal government should decentralize Medicaid, Medicare, and control over self-insured firms to the states. States currently finance one half of Medicaid. Because the poor do not vote, the federal government is likely to shift responsibility for Medicaid to the states in order to reduce its budget problems. Shifting Medicare to the states is unlikely because the elderly do vote and any possible diminution in their services will result in lost votes in the next election. The most difficult aspect of decentralization is control over the medical policies of self-insuring firms. Because they are currently exempt from state mandates, firms will lobby hard to maintain their privileges. Yet decentralization of control over self-insuring firms is the most important to achieve variation in policies and empirical knowledge.

In any computer system there is a tradeoff between security and the ease of access. For example, if patients medical records are made too secure, then such a patient may not be able to receive proper care in an emergency if emergency care givers either can not access or take too long to access the patients records. There are many ways this problem can be handled. For example, in a hospital with medical records accessible at workstations, there access could be restricted by passwords. While this would control access, it could be catastrophic in an emergency. Another possibility is having an audit trail of everyone who accessed a record. Accessing records without a justifiable reason is grounds for dismissal. The second involves more risk, but is much better in emergency situations. While the federal government is proposing a medical privacy bill, it should leave the issue to the states until there has been enough empirical evidence of what works and what does not.

States could require medical insurance firms to maintain all medical insurance records of state residents within the state so that their care would be subject to state law. This would include the records maintained by the medical insurance trade association in Massachusetts for the purpose of insurability of future requests for insurance. By doing this the states could impose alternative penalties for errors in the records. For example, some states might simply let the matter be resolved under common law for which the insurance firms would have to take more care in ensuring that the records were accurate or suffer the possibility of a tort law suit.

A major difference between Conservatives and Liberals in medical policy is whether individuals should be responsible for their own medical payments or whether medical risk should be pooled. Let us consider the Conservative position first. From the Conservative perspective it is fair that in buying medical insurance individuals should pay premiums based on their risk factors and the healthy should not have to subsidize the sick.

Thus for medical insurance Conservatives believe that the young should have lower premiums than the middle aged. Indeed since the middle aged earn more on average than the young this is only social justice. Premiums should also reflect differences in medical costs between the sexes. In addition, premiums should reflect lifestyle choices such as smoking or drinking alcohol. If we take the conservative position to its logical conclusion, then medical insurance should be based on all information which makes a difference. If social justice means that the healthy should not have to subsidize the less health, then individuals with good genes should not have to subsidize the medical costs of those with poor genes. Thus if insurance is really to be based on risk, there is no privacy and all information which can be used to predict expected medical costs should be used.

The important point about having insurance based on risk is that any information which remains private means that the risk for that information is being pooled. One method to preserve privacy and to have medical expenses based on risk is not to have insurance and have each individual pay for his or her medical expenses as they arise. With payment for services rendered out of the patients pockets, there is no need for any disclosure of genetic or any other type of medical information to have medical payments based on pure risk.

Most individuals especially most conservatives would be aghast at the prospect of having to give up genetic information in order to obtain insurance based on pure risk. A compromise proposed by conservatives in the concept of a medical savings account. Individuals have a savings account with a specified amount of savings. What they do not spend on medical expenses the individual gets to keep. A medical savings account maintains individual privacy because there are no required disclosures. This type of an account does create a conflict with the conservative's ideal of no medical subsidization between the healthy and sick. Because the concept of a medical savings account provides individuals with a tax savings, it provides the healthy with more benefits than the less healthy. Thus it implicitly is a subsidy from the sick to the healthy.

Since a medical savings account does provide enough money to cover major medical expenses, a medical savings account would be coupled with catastrophe medical insurance. For this type of insurance most conservatives would base risk on age, sex and lifestyles. This means, of course, that such insurance is pooling the risk for nondisclosed information. Because the expense of catastrophe medical insurance can be made inexpensive by raising the amount in the medical savings account, it should be affordable to all.

The conservative approach to improving medical efficiency is by having the individual be very concerned with the cost of each medical procedure. With third party insurance, patients did not concern themselves with the cost of each procedure as long as the insurance covered the cost. Now with a medical savings account individuals will be very concerned. They certainly will vehemently oppose hospitals trying to distribute costs for indigent patients or new medical technology to them. Definitely medical savings account will make medical care givers alert to the costs of alternative procedures and drugs to treat disease. With medical savings accounts, states to make the choice mechanism effective will have to have as a provision of their information policy that medical care providers fully disclose all costs and provide a detailed breakdown on request so that concerned patients can challenge the costs. Such an information policy would require a fundamental reassessment of how to pay for the medical expenses of the near poor who can not afford medical treatment or catastrophe medical insurance.

The problem which as we will point out that neither conservatives nor liberals are prepared to deal with is the separation of medical decision from work decisions in self-insuring firms. Suppose a conservative state (or federal government) creates a provision in its information policy that all genetic information is private and permission is required for any type of testing. As molecular biologists interpret the human genome they will discover more and more genetic information related to disease. For example, researchers have already discovered the genetic code for Huntington's disease. Now when a firm self-insures its medical expenses directly reduce profits. Such firms have powerful economic incentives to have a healthy workforce. They have strong incentives to improve lifestyles which add to medical expenses. For example, firms can test for drugs and not hire individuals who are on drugs. They can prohibit smoking in the workplace and have corporate gyms to promote fitness among their employees.

The extent to which firms will be tempted to genetically screen employees depends on a variety of factors. One is the cost of genetic screening equipment. If genetic screening equipment becomes inexpensive and automatic in that it does not require a molecular biologist to interpret the result, then self-insuring firms will be tempted to use it to obtain a healthy workforce. This temptation would be greater if states do not enforce laws prohibiting secret genetic testing. Given the current method of finance of elections, state and federal representatives would not have strong incentives to enforce the law.

I speculate that employees in self-insured firms are already significantly healthier than the population in general and than this difference will continue to grow in the next century. When a self-insuring firm interviews a large number of potential employees, they have strong incentives to consider the expected medical costs of a prospective employee. If two candidates have equal ability, the firm has a strong economic incentive to pick the candidate who has the lower expected medical costs. The same factors influence promotion and all other work related decisions.

To insist by law that self-insuring firms not take medical expenses into consideration when making work related decisions is difficult and expensive to enforce under the present method of self-insuring medical expenses. With a large number of prospective employees for a job the task of proving that a firm took medical factors into consideration is not easy to prove. For example, the investigators might send a series of equally qualified candidates with very different expected medical costs to seek employment. But this is very expensive and can not be employed for every firm. What could happen is that there could be infrequent scandals which promote period active enforcement followed by inactive until the next scandal.

Try to separate medical and work decision purely of the basis of information policy is likely to be ineffective. A much better approach would be to change the finance of medical expenditures. For example, if medical expenses had to be contracted to outside insurance and the insurance pool were large, firms would have little concern for medical expenses as part of work decisions. Denial of medical insurance would be observable, whereas denial of a job because of high expected medical costs is not readily observable.

States with medical savings accounts would would not have much resistance for self-insuring firms for making catastrophe insurance an outside purchase because the expense is so much smaller than the current medical coverage and the fact that states might mandate special provisions would have much less of an impact on the overall cost.

To prevent firms from having incentives to determine potential employees and current workers genetics the best approach is to reduce their incentives to acquire such information.

Liberal states would take a very different approach to medical and medical information policy. For liberals the proper approach to medicine is to pool the risks under a community rating. This can be justified under a Rawls criterion of assuming individuals had to choose a desirable medical plan before they were born for a random age. They would not know their age, sex or genetic characteristics. Under such a selection process individuals might well choose to pool medical risk because they would not know whether their own medical risk would be higher or lower than average.

Liberal states might well implement a variation of the national plan proposed by the Clinton administration. Medical insurance would have a community rating which means that there would be variations in medical insurance among communities based on risk factors, but that everyone in a particular community would pay the same rate. Liberal states would also have a tendency to impose the costs of medical insurance on business.

Liberals have a very different idea about how to achieve medical efficiency than conservatives. Liberals assume that medical procedures are too complicated for the average person to understand and be able to choose between the alternatives. Experts would judge the quality and cost of alternative medical programs and restrict the offerings of the best programs to the community residents. The information policy provision to make this possible is that the expert must have access to the details of medical care provided by each alternative provider in order to judge the costs and benefits. This also means that to judge effectiveness, the experts would have to be able to examine patient records. However, since the experts have no interest in the patients name they could limit themselves to a representative scientific sample.

Liberals like conservatives would have a difficult time dealing with the economic incentives of self-insuring firms to mix medical decisions with work decision. Self-insured firms would vigorously resist losing their exemption from state mandates in such liberal states as such a loss would definitely drive up their medical costs. As is the case in conservative states, trying to separate medical and work decisions solely on the basis of information policy, such as prohibitions on collecting medical information, is likely to be both expensive and ineffective.

The goals of decentralization of medical and medical information policy is to obtain a large number of empirical approaches to a very difficult social problem of trying to provide universal health coverage at a reasonable cost.

Let us first consider medical records. Over the next several decades the technology of medical records will be in constant flux. Because of this constant change a single federal solution concerning medical record security, access and accuracy is likely to be inadequate. With decentralization some states will fail miserably, but some will achieve a good balance of security and ease of access in emergencies which over time will be imitated by other states. The difficulty is this regard is what medical information should third parties be able to obtain.

Consider first medical research. Medical research need representative samples without labels identifying individual subjects in order to determine the effectiveness of medical procedures in the field. Followup on cases can also be done without the researchers knowing the identity of the subject. This information is also useful to define standard practice. Even though medical researchers do not need labels identifying individuals, many individuals would insist that medical researchers have their permission before obtaining the medical data. This will not lead to representative samples and greatly decrease the value of such data for research purposes. Some states are likely to allow research data without permission and others not. The same problem would occur in liberal states who opt for community-based pooling and need data to evaluate the cost and benefits of alternative medical plans.

Now consider third party insurance. Increasingly insurance firms are increasing their demands for medical information from care givers in order to control costs by making medical decisions as to whether to authorize or deny medical procedures. By expanding the network through which medical information flows it becomes less secure. Also their is a fundamental issue as to whether medical care givers or accountants should be making decisions on medical care. States are likely to vary considerably on what medical information should be allowed to flow to third party insurance.

The issue of what medical information self-insuring firms should be able to obtain when they pay the bills is likely to be even more controversial than the case of third party insurance. One was to limit the ability of self-insuring firms to mix medical and work decisions is to severely limit the flow of medical information to such third party firms.

When we consider the issue of risk-based insurance even the most conservative states are unlikely to push the risk classes far beyond age, sex and lifestyles.

The problem states will have promoting medical efficiency is distinguishing between medical efficiency and medical rationing. To lower costs HMOs may decide to limit care. How states might best deal with this problem is an open question. Currently states are formulating patient rights at HMOs. Such rights include direct access to specialists, no gag rules on medical care givers on restricting care, easier access in emergency care, improved grievance procedures, and quality control assurance. There is currently great variation among the states as to how many provisions they have enacted. States should also making information policy rules specifying HMOs clearly specify what procedures and tests they authorize and which they do not.

Like it or not economic considerations will influence medical decisions for all time. Society is simply not prepared to spend the amount of money required to provided everyone with every conceivable treatment medical research has created. Is this issue to be determined solely by what medical procedures individuals can afford through individual purchase or insurance. If public or private medical insurance is to be limited, most states may be forced to follow the Oregon plan to limit medical procedures to cost effective procedures. States which authorize representative samples to medical research will have an easier time determining the effectiveness of alternative medical procedures and creating good guidelines.

The federal government should fund research into which type of state plan leads to the greatest medical efficiency. States will have to choose between medical savings account and community pooling of medical insurance because the two approaches can not coexist in the same place at the same time. Healthy individuals will always select medical savings accounts driving up the costs of pooled insurance and forcing the almost healthy to opt for medical savings accounts. Careful study of the two approaches over an extended period is necessary to determine a clear cut social superiority.

Medical savings account may well meant that the less health are subsidizing the healthy. Individuals may skimp on preventative medicine to save money which in the long run will result in higher medical costs. Individuals may or may not understand medicine will enough to be able to select between alternative procedures. On the other hand community-based pooling may result in large administrative overhead to determining the best plans and compensating insurance firms who accept customers with higher medical risks.

The most difficult tasks states will have is trying to separate medical decisions from work decisions in self-insuring firms. While it would be more effective to end the practice of self-insuring such firms now immune from state mandates will be very unlikely to give them up. Trying to rely solely on information policy in this regard is likely to be expensive and ineffectual.