What is a disease?
The recent furor over Viagra for sex offenders also generated heated debates over whether Medicaid should cover Viagra for anyone. Pro- and con-arguments seemed to be trading on core assumptions about the nature of disease and the goals of medicine. So, I thought it might be fun to examine some of these conceptual questions in greater detail.
The debate about public insurance coverage for Viagra unfolds in the following context: Healthcare resources are finite. We can't afford to cover every effective treatment for everyone. So, how do we decide which treatments should be covered by public insurance?
A good rule of thumb is that medical insurance should cover the prevention and treatment of disease, but not recreational or cosmetic procedures. This model treats health insurance like a car warrantee. If something breaks, they'll fix it. If your warrantee includes a service plan, you can also take the car in for tune-ups, oil changes, and other preventative car care. But if you want to pimp your ride, you'll have to buy your own neon lighting kits and rear windshield decals because your warrantee doesn't extend to the car equivalent of cosmetic surgery.
This model works well for cars because there's an objective standard for car "health," not to mention a strict expiry on the warrantee. The car is considered healthy if it performs to factory specifications. The warrantee will pay to bring your car back to health.
The problem with treating health insurance like an extended body warrantee is that there's no comparable standard of health for human beings. There are easy cases like broken bones and malignant tumors, but in many cases, the model offers no clear guidance. There's too much variability within the normal population. We recognize this in terms of height and weight. There's always going to be a range of weights and heights in a population. Being above or below average doesn't necessarily mean there's anything wrong with you--except at the very extreme ends of the distribution, and even then it's not necessarily clear cut. But we're often prepared to cover treatments for people at the very extreme ends of the distribution, even if we can't identify any specific pathology. For example, most people would support publicly-funded medical treatment for a child who would otherwise become a midget or giant. They wouldn't ask whether the child was suffering from a documented growth hormone secretion defect, or whether he was just "naturally" destined to be 3 feet tall.
The case of Viagra is complicated because there's a range of erectile function among normal, healthy men. Maybe the distribution looks like a bell curve with the porno woodsmen and the asexuals anchoring the tails of the distribution, everyone else somewhere in the middle.
Most people would consider total impotence in a very young man to be a serious medical problem. Any reasonable "body warrantee" would cover treatment for such a grave and unusual condition. We also assume that it's "normal" for erectile function to decline with age. There's no doubt that it's the statistical norm, but just because it happens to most people doesn't mean that it isn't a medical condition worthy of treatment. Cataracts are a normal part of the aging process--but we're not prepared to let the indigent elderly go blind because it's "normal" for the lens of the eye to develop faults with heavy use.
We used to assume that dementia was just a normal part of aging. Luckily, we now see age-related cognitive decline as a medical condition, not an inevitable part of the aging process. When an older person becomes confused, we don't just chalk it up to some nebulous "aging process", we actively seek out the pathology and try to correct it. Maybe a similar paradigm shift is overdue for age-related declines in erectile function.
If the goal of medicine is to fix what's broken, we need some standard for normality. Perhaps the correct standard is based the typical level of function in a population. In that case, it's the extreme outliers who deserve medical treatment. In that case, the question becomes population we should use for comparison. If we compare an older guy to the population at large, he might be an outlier, but not if we compare him to other people his age.
We also have to ask whether the norms for a given population are healthy or desirable. It begs the question to say that whatever is typical is healthy. Consider the heights and weights of North Korean children. In this population, a typical weight is not a healthy weight because malnutrition is so pervasive in North Korea. Likewise, it might turn out that the average IQ of the older population is lower than it should be because of treatable dementia. What looks like normal aging might actually be widespread pathology.
We could refine our definition of "disease" in various ways. One strategy is to define health in terms of functionality and construe disease as deviation from functional health. On this view, osteo-arthritis treatments should be covered even if wear and tear on the joints is paradigmatically normal for old people. It's quite reasonable to say that the joints are designed to flex painlessly and therefore that stiff or painful joints deserve medical attention.
The functional model isn't a free pass to objectivity or rigor, however. Functional idealizations are useful, but they don't eliminate the need for further controversial value judgements. These days, children who can't sit still in school are often diagnosed with attention deficit disorder (ADD). ADD-skeptics sometimes argue that the problem lies with a system that requires rambunctious kids to sit down and shut up for long periods. In other societies, energetic kids might have had more outlets in physical work and rough play. Be that as it may, in our society, the capacity to sit still and listen is very important to one's future prospects. If you want to use a functional criterion for insurance coverage, you have to ask whether to include this ability in your model. Personally, I think that if person is constitutionally incapable of paying enough attention to get by, even when they want to, then they've got something sufficiently disease-like to warrant insurance coverage.
Despite it merits, the functional model doesn't shed much light on the Viagra coverage controversy because it's not clear which function(s) the treatment is supposed to address. The most obvious biological function is reproduction--yet nobody's pretending that Viagra is primarily a fertility treatment. At this point someone will drag in evolutionary biology, but discovering which functions promoted fitness in a hunting and gathering society doesn't say much about what our medical priorities should be today.
Of course, we all know that the real function at stake is recreational sex. By itself, the functional model can't tell us how much priority we should assign to preserving the capacity for sexual pleasure vs. other biological functions. I'm inclined to assign that particular function a relatively high priority, but the functional model doesn't advance my case.
Debates about the true nature of disease seem to proceed in a circle. We want to find some objective standard for pathology. Someone proposes a plausible criterion. Skeptics ask why we should care about that standard, and we're back to where we started.
I think it's a mistake to base health care allocation on a concept as ill-defined as "disease." The fundamental problem is that we're looking for a bright line that doesn't exist. We started with a pre-scientific concept and we're making heroic efforts to reconcile it with modern science and medicine as it's practiced in our culture. The fit isn't always perfect, nor should we expect it to be.
The decision about whether to cover Viagra shouldn't depend whether erectile dysfunction meets someone's definition of disease. What matters most is plain old cost-benefit analysis. Obviously, not every beneficial intervention is health care. However, separating medical treatment from non-medical treatment is not a serious practical problem. As long as we're talking about prescription drugs or procedures that can only be administered by physicians, we're talking about potential candidates for health insurance. (There are legitimate background questions about how medicine and medical power are delineated in our culture. We can argue about whether a drug ought to be available over the counter, or whether non-physicians should be allowed to deliver a certain type of care, but these issues are peripheral to this discussion.)
So, what of Viagra? Ultimately we're talking about a relatively expensive drug that provides symptomatic relief for a non-lethal condition. Advancing sexual health is a worthy goal, but if that's our objective, we'd do better to allocate more of our budget to preventing and treating high blood pressure and diabetes, which are among the leading causes of both impotence and premature death. I also support Viagra coverage for very severe cases of erectile dysfunction with a clear organic cause, such as prostate surgery, or advanced diabetes.
It's clear that intense promotion and high demand have resulted in more insurance coverage for Viagra than the drug deserves. It's equally clear that there are good reasons to cover some or all of the cost of the drug under certain circumstances. These decisions have little or nothing to do with whether erectile dysfunction is a "real disease."