FDA "Breakthrough Drugs": A Trial Testing the Effect of Alternative Language on Public Perceptions

An online sample of 600 American participants was recruited from Amazon's Mechanical Turk. Respondents were randomly assigned to 5 conditions in which they read an excerpt from a news article describing a newly FDA-approved fictionalized breakthrough therapy, Zykanta, for treating metastatic lung cancer. Information on this fictionalized drug was drawn from a real-life breakthrough designated drug so as to be able to provide as accurate an example as possible with respect to inherent uncertainty in risk and benefit. Each condition communicated information about Zykanta with a greater or lesser degree of transparency regarding the contingent nature of a breakthrough approved drug. All conditions contained the same factual information about the risks and benefits of the drug. In the first condition - a Purely Factual condition - participants read a clear and concise technical definition of breakthrough therapy designation about a drug that was approved 4 months ahead of schedule, without seeing the term 'breakthrough' attached to the drug. This condition provides the basic risk-benefit information made available by the FDA without loaded terminology. In the second condition, the factual information was enhanced by describing the drug as "Promising" - the language the FDA uses in press releases about breakthrough drugs. This condition provides a more tempered label for this designation, but a label nonetheless. In the third condition, the factual information was enhanced by describing the drug as "Breakthrough." In the fourth condition, factual information about the drug was enhanced by describing the drug as "Breakthrough", but additionally contained a Tentative Disclaimer about the possible uncertainty surrounding the drug. This condition reflects the information found in the small print of the labeling of some breakthrough drugs. In the fifth condition, the factual drug information was enhanced by describing the drug as "Breakthrough", but contained a more Definitive Disclaimer about the uncertainty surrounding the drug and its need for further testing.

Measures To determine whether the term 'breakthrough' resulted in mistaken inference about its safety and effectiveness, all participants were asked to make a choice between a 'breakthrough drug' or a drug that met the definition of breakthrough without the explicit use of the word "breakthrough" in the description ("If you had a potentially deadly medical condition and could choose between 2 drugs recently approved by FDA, which would you choose? (1) 'Axabex, a "breakthrough" drug newly approved by the FDA' or (2) 'Hypapax, a drug that has shown some early promise in drug trials but which has not been shown to improve survival or disease-related symptoms'). All participants were asked whether they had heard the term "breakthrough drug" prior to being asked the question. All participants also provided a written definition of a breakthrough drug.

After reading the news excerpt on the fictional breakthrough drug in each of the 5 conditions, participants were surveyed on the safety and efficacy of the drug described in the news excerpt ("How safe is Zykanta compared to other drugs approved to treat metastatic lung cancer?", "How risky is Zykanta compared to other drugs approved to treat metastatic lung cancer?", "Has Zykanta been proven to save the lives of patients with metastatic lung cancer," "How effective do you think Zykanta is at treating metastatic lung cancer?" "If a patient with metastatic lung cancer started to take Zykanta, what do you think would happen to their tumor over the next 6 months?" and "How effective is Zykanta compared to other drugs approved to treat metastatic lung cancer?") and the strength of the scientific evidence that the drug helps metastatic lung cancer ("How strong is the scientific evidence that Zykanta helps patients with metastatic lung cancer?" and "If you were deciding to take Zykanta, would you feel confident making that decision given your current knowledge of the drug?"). We also asked about the likelihood of a symptom that was explicitly stated as a common symptom in the drug description ("If a patient with metastatic lung cancer started to take Zykanta, how likely is it that they would experience abdominal pain within 6 months?") and about their own anticipated behavior with respect to the drug ("If you were diagnosed with metastatic lung cancer, would you want to take Zykanta?" and "If a close friend or family member was diagnosed with metastatic lung cancer, would you want them to take Zykanta?"). Each response was measured on a 4 or 5-point Likert scale.