Approval of a novel drug by the FDA garners worldwide respect. It signals that the safety and efficacy of a medication has been demonstrated in carefully monitored research studies. According to written standards, effectiveness must be demonstrated by substantial evidence. Until relatively recently, "substantial" was defined as at least two well-controlled clinical investigations. However, in 2017 the implementation of the relaxed the to make it easier to gain FDA approval on the basis of a single trial.
Over the last several years, I have been working on an independent accounting of evidence used for each new FDA approval in collaboration with colleagues Veronica Irvin, PhD, MPH, at Oregon State University, and Amanda Koong, MS, a McGovern/University of Texas medical student. Earlier this month, we published showing that in 2022, a single study was used to justify 65% of novel drugs approvals. In 2016, the year prior to the implementation of the 21st Century Cures Act, a single study was used to justify only 20% of new approvals. In 2016, 55% of the products were approved on the basis of three or more studies. By 2022, three or more studies were used to justify approval of only 11% of novel drugs.
This week, we published that focused on the 46 FDA approvals in 2017. We choose 2017 because it was the first year the Cures Act was in effect. Using ClinicalTrials.gov, we were able to document how many studies were initiated, completed, and reported before and during the 5 years after each product was approved. We found that in 2017, 41% of the products were approved on the basis of a single trial. When we probed further, it was apparent that other studies had often been initiated before the approval. In many cases these studies had finished data collection many years before. Among studies completed prior to approval, only about one in four had made their results public. Even after 5 years, less than half of the studies had disclosed their results. We also found that companies commonly wait until a drug is approved before releasing results. For 72% of medicines approved in 2017, new information was made public shortly after approval.
We are not the first to report that many completed studies fail to report results. And, we join others who have questioned the level of reporting. Along with these colleagues, we worry that limited availability of trial results deprive consumers and healthcare providers of the full picture necessary to make good decisions.
Although these issues are hiding in plain sight, attempts to address them have fallen flat. The (FDAAA) includes specific provisions to remedy non-publication of clinical trial results. In particular, the FDAAA allows a fine of up to $10,000 per day (with a ceiling of $1 million) for failure to disclose trial results within a year of study completion. The policy was in January 2017. Yet, a revealed that the FDA had never issued a non-compliance notice up to that point. Further, they had never imposed a single fine. Simply enforcing the authority to collect these penalties would have yielded $19 billion in legitimate fines. To put the $19 billion in context, it is about triple the 2022 FDA budget allocation. Further, the 2017 deadline for reporting compliance seems to have had little effect. Our analysis of 2017 approvals showed a greater gap between completed trials and those reporting results than analyses completed .
It might be argued that companies should be allowed to report only their most convincing evidence. In science practice, it is common for labs to present only their most pristine data. In their evaluation, FDA could consider proof of principle rather than considering averages across many studies of varying quality. But focusing on single trials raises important questions. We expect about one trial in 20 to provide a statistically significant result by chance alone. Our concern is that the single study used to establish the safety and efficacy of a particular product could be a false positive result. Putting trial results within the context of all studies that have been completed offers patients and their providers a better basis for making crucial treatment decisions.
We concluded our recent publication with the suggestion that consumers deserve access to all evidence regarding the drugs they might take. It is unfair to limit public disclosure to a subset of selected studies that may show a product in the best light. And, achieving better transparency may be as easy as enforcing existing federal policy.
is a faculty member at Stanford University's Clinical Excellence Research Center, a former associate director of the National Institutes of Health, and a former chief science officer for the Agency for Healthcare Research and Quality.