Many see Vioxx and Avandia as clear signs that the drug safety system has failed. As soon as reports of these drugs' adverse events began to flood the media, consumers—and Congress—demanded to know: "Why didn't we know sooner?"
Indeed, as we enter into the new Culture of Drug Safety, "knowing sooner" will be the call to arms. Companies, regulators, and others will tap existing and new surveillance systems to generate more safety signals. It is likely that such systems will improve the speed at which potential safety issues are spotted. However, even with these improved systems, understanding what truly qualifies as a safety signal—and what's background noise—will take time, likely cause confusion, and in sum, not be a magic bullet to solve the safety problem.
To move the safety system forward, we need to realize that it's not enough to know about drugs' risks—you also have to communicate them. After all, for both Vioxx and Avandia, the drug surveillance system worked exactly as designed. With Vioxx, a clear safety signal emerged after the VIGOR study, a 5,000-patient randomized active-controlled clinical trial that showed Vioxx to be riskier than the control drug naproxen. But from that trial, Merck still needed to "qualify the signal" to understand if Vioxx was increasing the risk of myocardial infarction or if naproxen was suppressing the risk. This required a placebo-controlled study of about 2,000 patients over a two-year period, the APPROVe trial. Once that trial started, the difference between Vioxx and placebo was not apparent until after 18 months. With Avandia, the cardiovascular safety data emerged only after both the company and a private investigator examined a very large number of clinical trials in different meta-analyses—nearly eight years after the drug was first approved.Doctors learn about these drugs' risks as the research unfolded. However, when it comes to patients, the pharma industry has done a poor job of explaining risk—and, particularly, unqualified risk. As more safety signals arise from the increasing number of safety and postmarketing studies conducted, we will need to give patients a better framework for processing the news about risk. There will be some hard questions to answer. For example, how do we tell the public "there may be something bad happening, but we aren't sure" in a way that won't cause panic? Regardless, we must begin moving forward. Here are some ideas to get the industry started.
Accutane. Thalomid. Tysabri. These are the drugs we associate with risk-management plans (RiskMAPs). But actually, many more drugs have risk-management plans, and many drugs without formal RiskMAPs use some risk-communication tools like "Dear Doctor" letters, patient package inserts, medication guides, and informed-consent sheets.
These materials must follow the law and provide a message that is consistent with the approved package insert. So, too, must other marketing communications, which can include anything from DTC ads to detail pieces. However, even though all communications are based on FDA-approved labeling, they are often at odds with each other. The format, layout, and even the takeaways from these materials are likely to be quite different—and confusing for patients. As is often the case, consumers get the story on benefit, and then the story on risk, but no real way to piece these bits of information together.
For physicians who have dealt with package inserts and product promotion for years, such a mixture of messages is not only tolerated, but also expected. Doctors know that drug company attorneys insist that companies include every possibility. Doctors also know to take marketing messages with a grain (or, more likely, a pound) of salt and to rely on independent sources of information, such as academic investigators and medical journals. They are accustomed to scanning these resources, speaking with reps, and forming their own opinions.