چه چیزی باید (عمومی) بیمه سلامت را پوشش دهد؟

In any system of health insurance, a decision must be made about what treatments the insurance should cover. One way to make this decision is to rank treatments by their ratios of health benefits to treatment costs. If treatments that are not offered by the health insurance can be purchased out of pocket, the socially optimal ranking of treatments to be included in the health insurance is different from this standard cost-effectiveness rule. It is no longer necessarily true that treatments should be ranked higher the lower are treatment costs (for given health benefits). Moreover, the larger are the costs per treatment for a given benefit–cost ratio, the higher priority should the treatment be given. If the health budget in a public health system does not exceed the socially optimal size, treatments with sufficiently low costs should not be performed by the public health system if treatment may be purchased privately out of pocket.

In any health insurance system, public or private, one must make a decision about what treatments the health insurance should cover. Health economists have often argued that cost-effectiveness analysis should play an important role in choosing what should be offered by health insurance. Cost-effectiveness is in this context usually defined as the minimum cost for a given health benefit, or equivalently, maximal health benefits for given expenditures on health care.1 There is a large literature that is critical to this type of analysis. One line of criticism is that cost-effectiveness analysis requires an aggregate measure of health benefits. Whether this measure is “quality adjusted life years” (QALYs) or some other measure, one needs severe restrictions on a general preference ordering over life years and health quality of each life year to be able to represent preferences by any simple aggregate measure.2 A second line of criticism has been that whatever aggregate health benefit measure one uses to represent preferences at the individual level, one might question the ethical or welfare theoretical basis for aggregating health benefits across individuals.3 The present paper ignores the above-mentioned problems with cost-effectiveness analyses of prioritization issues. The focus is instead on a different important issue: at least for public health insurance, most of the literature that discusses how a health budget should be allocated across potential medical interventions explicitly or implicitly assumes that the health interventions that are not funded by the public budget are not carried out. However, both under public and private health insurance it is often possible to purchase treatment out of pocket if treatment is not covered by the health insurance. Examples of treatments that typically may be purchased out of pocket are surgical sterilization, assisted fertilization, cataract surgery, dental care, prescription medicine. Comparing different insurance arrangements one will find that they differ with respect to what is covered and what is not. When treatments of the type above are not covered by the health insurance, they are nevertheless available for those who want to finance the treatment out of pocket. The paper discusses the use of cost-effectiveness analyses for prioritizing a health budget for a public health system or a private insurance company when an out of pocket option exists. It is shown that when there is an out of pocket option, a simple cost-effectiveness criterion of maximizing the sum of some aggregate measure of health benefits for a given budget is not necessarily the best way to allocate the health budget. In particular, such standard cost-effectiveness analysis does not maximize the sum of utility levels of the members of the health insurance. The reason for this is that the benefit of including a particular treatment in the insurance program can no longer be measured simply by the gross health improvement this treatment gives: some of the health care would otherwise have been performed in any case, so the net health increase is lower than the gross increase. On the other hand, by including a treatment in the health insurance, there are reduced personal costs of treatment financed out of pocket. This cost saving should be included in the benefit side of including the treatment in the health insurance. In order to add the benefits of improved health with the personal cost saving one is thus forced to make a monetary valuation of the net increase in health benefits. The paper shows that maximizing the sum of utility levels of the members of the health insurance (given the budget) gives a different outcome than simply maximizing gross or net health benefits for the given health budget. A comparison is also given between the ranking that maximizes the sum of utility levels and the standard cost-effective ranking of different treatments. While most of the content of the present paper applies both to public and private health insurance, it is perhaps most relevant for public health insurance. For public health insurance, the members of the health insurance will typically be exogenous, and often equal to the total population. It then makes good sense to allocate an exogenously given health budget to maximize the sum of the members’ utility levels (although this is not the only conceivable social objective). As a normative recommendation, this also makes sense if the health insurance is private. However, in the latter case health insurance will often be voluntary and there will typically be a choice between insurance companies. The present analysis has nothing to say about what is optimal from the point of view of an insurance company, or how the allocation of persons across insurance companies may depend on what treatments the insurance companies offer. Finally, we disregard the issue of a possible mix of private and public insurance, as discussed by e.g. Blomqvist and Johansson (1997), Allesandro (1999) and Hansen and Keiding (2002). The rest of the paper is organized as follows. In Section 2 the basic model is introduced, and the socially optimal ranking of different treatments is defined as the ranking that maximizes the sum of utility levels of the members. The standard cost-effective ranking (i.e. ranking according to ratios between health benefits and costs) is compared with the socially optimal ranking. When there is no out of pocket option these ranking are identical. With an out of pocket alternative, however, the rankings differ. An important result in this section is that it is not obvious how the costs of treatments should affect the ranking. It is shown that for identical health benefits, a treatment with higher costs may be ranked higher than a treatment with lower costs. Moreover, the larger are treatment costs for a given benefit–cost ratio, the higher priority should the treatment be given. While the health budget is assumed exogenous in Section 2, the socially optimal budget is derived in Section 3 for the case of public health insurance. It is shown that if the budget is not higher the optimal level, treatments with sufficiently low costs should not be covered by the health insurance, no matter how large the health benefits are. Section 4 discusses some extensions, and Section 5 concludes.

The preceding analysis has shown that the existence of an out of pocket option has important consequences for the ranking of treatments in a cost-effectiveness analysis of what health insurance should cover. The conclusions are summarized in Proposition 1, Proposition 2, Proposition 3, Proposition 4, Proposition 5 and Proposition 6. To be able to make an exact ranking over alternatives, one would in addition to the health benefits that enter a standard cost-effectiveness analysis also have to compute the second term of the ranking function given by (6). However, even without such an exact calculation of this term, our qualitative results may be of some use in how to rank different treatments. In particular, for treatments with roughly the same benefit–cost ratios (h/c) that are all close to the threshold level λ, our analysis suggests that the health insurance should cover the treatments with large costs per treatment but not those with relatively small costs per treatment. An example of treatment with low costs could be surgical sterilization (at least for men). This is a once in a lifetime treatment and the cost is only a small fraction of 1% of the life time income of most people. The present analysis therefore gives a good justification for not including this treatment among the treatments covered by the health insurance. Prescription medicines for chronic diseases are on the other hand often quite costly (over a life time), and there are thus good reasons for covering such expenses by the health insurance, even in cases where the health benefit to cost ratio is lower than for some treatments that are not covered by the insurance.