Big Pharma: the Unseemly First Amendment Champion, Part One

Jane Bambauer

For the conference on Public Health in the Shadow of the First Amendment

Big Pharma: the Unseemly First Amendment Champion

The pharmaceutical industry is enjoying the sort of influence over First Amendment jurisprudence once reserved for the likes of Hustler Magazine. In Sorrell v. IMS, the Supreme Court protected the transfer and use of prescription data for pharmaceutical detailing (that is, the in-person sales pitches that pharmaceutical reps make to doctors.) A more significant free speech victory for Big Pharma was delivered by the Second Circuit in United States v. Caronia, which overturned the conviction of a pharmaceutical detailer for off-label marketing of a drug. Public health experts and constitutional law scholars are intensely critical of these developments, arguing that they are perversions of the First Amendment to favor large corporations at the expense of consumer protection. For reasons I sketch here, their reactions overstate the conflict between Big Pharma and public health and underestimate the First Amendment interests at stake.

Two Views of Caronia

The FDA approved the drug Xyrem in 2002 for the treatment of narcolepsy in patients over age 16. Alfred Caronia, an employee of the drug’s manufacturer, encouraged doctors to prescribe Xyrem for narcoleptics under age 16 (in lower doses) and for medical conditions other than narcolepsy. The FDA successfully prosecuted Caronia under the Federal Food, Drug, and Cosmetic Act (the FDCA) for conspiracy to introduce a “misbranded” drug into the market.

Caronia can be understood two ways, under the speech view, and under the modern regulatory state view. The speech view looks incisively at regulations that suppress speech and analyzes the censorship without respect for the larger context. The modern regulatory state view starts with the larger goal of regulating a product or industry, and regards the speech regulation as an incidental burden. 

The Second Circuit approached Caronia’s prosecution under using the speech view. Caronia was convicted for speech that the FDA never proved was false or misleading. Unless the court intervened, the FDA would be free to prosecute detailers for providing accurate information to doctors about off-label uses of a drug. Federal law leaves doctors utterly unconstrained to prescribe drugs for off-label use. It also permits anybody other than the drug manufacturer to describe and encourage off-label uses of drugs. Viewed as a regulation of speech, this system looks absurd. The only speaker barred from giving doctors information about off-label prescribing is the manufacturer and its agents—speakers with a profit motive to be sure, but also speakers likely to know the most relevant information about the benefits of off-label use. 

The other way to understand Caronia views the conviction in the larger context of a modern regulatory state. Under this view, Caronia’s speech played only a functional role. His promotion simply defined a product. By describing Xyrem as a drug appropriate for ailments other than narcolepsy, Xyrem’s manufacturer was essentially selling a new drug. And that new drug, like all new drugs, would need to go through the usual FDA procedures (in this case, the manufacturer would have to file a Supplemental New Drug Application with accompanying clinical trial evidence) before it could be promoted.

This legal treatment of product-defining speech is not unique to drugs. For example, I can sell sharp steak knives without running into product liability problems, but if I start promoting the knives a fun children’s toys, that speech will opt me into products liability regulation for the safe design not of knives, but of toys. Chris Robertson has argued that the FDA regulation should likewise be conceived not as a prohibition on speech, but as a product regulation that gets triggered only when the chemical is promoted as a new drug. By this logic, the Caronia decision threatens to interfere with the regulation of potentially dangerous products.

So, Caronia is either a modest protection of truthful and non-misleading speech, or it is a crippling disruption of consumer protection. Which is it?

In my view, the Second Circuit got it right. Caronia’s conviction was a regulation of speech, not drugs. For unapproved drugs, the FDCA works precisely as Robertson says. The FDA regulates the drug itself by conditioning its sale on proof of efficacy and safety. But for already-approved drugs, the FDCA does nothing to limit the sale of the drug for off-label purposes. Doctors can prescribe at will to anyone for any purpose within the (generous) bounds of state medical malpractice law. So for already-approved drugs, the FDCA affects the drug’s usage, if at all, only by limiting the information doctors hear about the approved drug. Even if consumer protection is the ultimate goal, the FDA achieves that end by suppressing pure speech.

The complexity of First Amendment coverage does not end there. Even if Caronia's conviction was a regulation of speech, an argument can be made that the FDA has wide latitude to regulate. Assuming pharmaceutical detailers are engaged in commercial speech, under Central Hudson that speech would receive First Amendment protection only if it is truthful. False and misleading speech falls outside the scope of speech protection and would receive only rational basis review. The Second Circuit glossed over this prerequisite and concluded that Caronia’s speech was not misleading. Perhaps the brevity in the court’s analysis was the result of the government’s failure to allege that Caronia’s speech was misleading since the FDCA, on its face, did not require that allegation to be made or proved. The FDA’s next move, then, is to argue that promotion of a drug for off-label purposes is inherently misleading if the drug has not been FDA-approved for that purpose. Scholars have begun to make this very argument.

Who Controls the Meaning of “Misleading”?

Today, the FDA and FTC act as gatekeepers for information that is permitted to reach the doctor and the public by defining the meaning of “false” or “misleading” speech. To some extent, this role is unavoidable. The Supreme Court was responsible for creating this fault line between protected and unprotected commercial speech in Central Hudson. If the government wishes to clamp down on false or misleading speech at all, as it surely does in the context of drug advertising, it will have to choose which types of claims are true enough, and which aren’t.

However, courts cannot give public health agencies unconstrained discretion to define “misleading speech” however they please. If the government is too aggressive in its role as the arbiter of truthful speech, First Amendment values will be badly compromised.

Let’s consider the spectrum. Some cases are easy. If a pharmaceutical agent says, with intent to deceive, that a drug is FDA-approved for a purpose for which it is not, the statement is provably false. Moreover, when a company promotes the human ingestion of a chemical, factual statements that are utterly devoid of support should be well within the meaning of “misleading” even if the government does not know for sure that they are false. Chances are, the drug will not be effective and could very well cause harm. But on the other end of the spectrum, it is equally obvious that an exacting standard for non-misleading speech would hinder too much speech. If the FDA required all drug promotion statements to be based on ten years of data from clinical trials of a nationally representative sample of patients, and if the FDA interpreted statements without this support to be “misleading,” the standard would effectively foreclose pharmaceutical speech—even speech that is quite likely to be accurate. At some point along the spectrum, the First Amendment’s commercial speech protections must intervene. Otherwise, the government can exploit ambiguities in the term “misleading” to avoid scrutiny altogether. We have little guidance on when this intervention should occur. Scholars who have weighed in usually argue that government should receive wide latitude (see, e.g., Rebecca Tushnet's article on the definitional question of "misleading".)

Today, the FDA and, to a lesser extent, the FTC (through its substantiation doctrine) apply very stringent standards to commercial speech. Under the current FDCA rules, in order to promote an off-label use, a pharmaceutical manufacturer is restricted either to sharing results published in academic journals (which themselves apply rigorous publication standards) or to going through the procedures to get the use *on* label (including costly clinical trials.) The FTC appears to enforce similarstandards by treating a health claim as “substantiated” only if the statement is consistent with the standards of peer-reviewed science journals.

This approach is bad policy and will probably draw the agencies into constitutional turmoil. It places too much faith in scientific conventions.

First, the FDA’s and FTC’s standards reject too much useful information. Consider the information and prescribing environment as the FDCA creates it:

	What doctors can prescribe	What pharmaceutical companies can say	What everybody else can say
Before Approval/ First Use	Only after strong proof	Unrestricted	Unrestricted
After Approval/ Off-label Use	Unrestricted	Only with strong proof	Unrestricted

Placebo-controlled double-blind experiments might be an appropriate standard for a scientific journal, but those journals have the luxury of avoiding the constant stream of hard decisions that every doctor and consumer must make. At a critical point of deliberation, bad science on a drug’s risks and benefits could be better than no science at all.

Moreover, the agencies also risk training the public to think that scientific conventions are tantamount to “truth” which is, ironically, antithetical to the scientific process. Science is messy, tentative, and error-prone. We can get a sense of this just by considering the (arbitrarily chosen) 5% standard for statistical significance. A result that is statistically significant at the 5% level deserves a lot of confidence when it is viewed alone. However, as soon as we pull together fifteen statistically significant findings, the chance that at least one of the findings is actually random noise exceeds 50%. If we pull together a hundred studies, which is often what our body of public health knowledge consists of, the chance that every statement is true falls to under 1%. Add to this the fact that most public health research cannot be replicated and we get a humbling picture. Even when evidence-based claims live up to the current standards of scientific rigor, we are fumbling in the dark. (I don’t mean to be too pessimistic here; observational studies and controlled experiments are the only thing that takes us from wrong to very slightly less wrong, and that's a good thing.)

The FDA’s regulations rely on an assumption that less-than-rigorous scientific statements in favor of a drug’s adoption on balance not only alter doctor behavior, but also lead to patient harm. This assumption may be correct, but at present it lacks a basis in evidence. Profits and good advice are not always at odds. When they pull in the same direction, skepticism and regulation can get in the way of health improvements. For example, there are some drugs, such as tPA for the treatment of stroke, that are so effective that the lag among ER doctors to use it is simply tragic. The public health community has not been able to show that pharmaceutical detailing causes more harm than good, which is why the health-related justifications in Sorrell were unconvincing. In the absence of evidence that the free exchange of information causes harm, the FDA’s approach is in trouble. It flies in the teeth of the Brandeisian assumption that bad information is best countered by more information rather than censorship.

I suspect that, in the wake of Caronia, Sorrell, and other cases, courts will apply intermediate scrutiny to any restrictions on commercial speech that is not provably false. Courts will not defer to agencies on the definition of misleading, so if an agency regulates speech that is technically true but arguably misleading and distortive, it will have to come to court with evidence of the potential for both misinterpretation and consequent harm. 

This shift will not be a welcome one to critics who think that the commercial speech doctrine was wrongheaded from the start. But these critics tend to undervalue consumer and listener interests in advertising, and to overlook other, better, alternatives for promoting good research. (To be continued in Part Two.)  

Jane Yakowitz Bambauer is Associate Professor of Law at the University of Arizona James E. Rogers College of Law. You can reach her by e-mail at janebambauer@email.arizona.edu