The Volokh Conspiracy

Mostly law professors | Sometimes contrarian | Often libertarian | Always independent


Why We Really Keep Seemingly Beneficial Drugs Off the Market


President Trump has received a non-FDA approved drug under a "compassionate use" exemption. Can a legal regime that allows the President access to a medicine but generally prohibits the public from purchasing it be justified?

The most obvious justification is that the President should receive special treatment. Sure, it is worthwhile for the country to spend a great deal of resources protecting the people's electoral choice by giving the best medical care possible. But that does not explain why the government forbids voluntary transactions between others and the drug maker. If the bulk of the evidence right now is that the drug is helpful, then why not approve the drug at least provisionally? Tyler Cowen advances this line of argument, suggesting that there should be more individual freedom to buy drugs.

One potential counterargument is that drugs might be misused. Doctors might have conflicts of interest, and anyway they might make bad decisions. We'll allow the President to use the drug because we can be fairly confident in his medical team. But if we allowed just any doctor to prescribe the drug, then on net we would do more harm than good. But it doesn't sound like President Trump's doctors have performed an especially complicated calculation here. It's worth it for the President to take the drug given his age and weight. Why then can't the drug be approved for everyone meeting certain conditions? One answer is that our system doesn't work that way; once a drug is approved, it can be used off-label. But that will not work as a defense of the status quo. It suggests that we should approve some drugs earlier with off-label use prohibited.

There is, however, another way to defend a system that insists on very high confidence in the safety and effectiveness of drugs but makes exceptions for VIPs or other select cases. The argument is based on the value of information. The candid way to make the argument is like this: Based on current information, this drug would improve patient welfare. But we don't know for sure. The population of future patients will benefit if we have reliable data. Once we approve the drug, it will be much harder to perform randomized controlled trials, both because the pharmaceutical company's incentives to perform such trials are reduced and because patients' incentives to participate in trials are reduced. We keep drugs off the market even after we think their benefits exceed their costs because we need guinea pigs. Only by keeping the drug off the market can we coerce people into participating in trials. But we can make a few exceptions without harming any ongoing trials, if sufficiently great interests are at stake. The President's health is one of those exceptions.

This argument makes many people uncomfortable. Medical ethics condemns experimenting on individuals with treatments that we think are likely to harm them, even with their consent. We can justify medical experimentation only when we do not know whether treatments will be harmful, and in this state of epistemic equipoise, we can run randomized controlled trials. That same logic might suggest that once a drug seems more likely to be beneficial than harmful, we should cease experimentation. Refusing someone an apparently beneficial drug to obtain more information harms that person. But status quo bias can avoid cognitive dissonance. There is not an easily definable point where our best estimate shifts from net harm to net benefit. And so we allow experimentation in some broad range of uncertainty, and we comfort ourselves in the thought that we are denying people these drugs for their own good. But when the President takes an experimental drug, we are forced to concede that this isn't always so. Sometimes, we are denying a drug not because of the interest of current patients, but because we want to make sure that the drug is safe before we give the drug to other patients.

Perhaps we would be better off if we were more honest about this. Sometimes we do not allow drugs on the market for genuinely paternalistic reasons (for better or worse), but at other times we insist on more randomized tests because we need the information that will result from experiments. The problem with admitting this, of course, is that the charade may end. The public would not let the government continue to keep a drug off the market if the drug admitted that the only reason it was keeping the drug off the market was that it wanted test subjects. But I am not convinced that randomized controlled trials require dishonesty. Some patients will genuinely be in equipoise about whether to take a drug, and trials can continue. Maybe drug companies will need to provide benefits (if not cash, then medical benefits) to those who agree to participate in trials instead of receiving drugs that are thought more likely than not to be beneficial. This introduces new ethical challenges. But it is hard for me to avoid the conclusion that the medical ethics of human trials is sometimes an elaborate analytical fiction designed to disguise the fact that we coerce people into participating in trials by harnessing the power of the state to keep seemingly beneficial treatments off the market. That seems more ethically troubling to me. If we admit that the public benefits from randomization, then we can confront both the ethical and informational challenges more honestly and directly.