When we launched our FDAAA TrialsTracker we wanted to produce a tool that would improve clinical trial reporting, rather than another repetitive academic journal paper that simply documents the extent of the problem. This reflects our ethos in the Bennett Institute: clinicians, academics and software engineers, working together to produce tools, as well as papers.

Two weeks after launch we have had extensive media coverage, and a lot of great user feedback. Here are two emails we’ve had from users who are employed to improve clinical trial reporting in major US institutions.

Firstly, Anthony Keyes of the Johns Hopkins School of Medicine Institute for Clinical and Translational Research wrote, explaining how our FDAAA TrialsTracker helped him find a trial that was about to go overdue, which they would otherwise have missed:

“I first logged on to TrialsTracker on 02/22/18 and discovered NCT02018562 which had a primary completion date of 02/28/17. This record was unfortunately not on our radar so I urgently reached out to the PI and we worked together into the evening to submit results that same day.

In our research to prevent records from falling through the cracks again we discovered that the date format for Planning Report columns “Primary Completion Date” and “Results Expected” is not standardized. When we attempted to filter the dates in order not all records were visible.

I am also grateful that your tool is designed to recognize when a record has been released and is currently in quality review. The majority of our records go through multiple submission cycles and it is not uncommon to spend a total of 90 days or more in review (currently writing a paper to quantify). The unpredictable delay is the main reason why we begin reaching out to our PIs and study team 4 months in advance of the results expected date.

Please extend my thanks and appreciation to your colleagues for a well-designed, pertinent tool that has already proven itself to assist Academia in compliance. We plan to use TrialsTracker as a supplement to our other efforts and look forward to future expansion of this tool.”

Another university staffer who helps their institution report trials on time (and preferred to be unnamed) shared the following:

“This tracker seems to be very well designed, with methodology that is sound and fair. My institution’s data is correct. I have bookmarked this site, and I intend to use it…to cross reference my data and to keep tabs on [my institution’s] public perception.

Unlike other trackers, this doesn’t tag records with late results as non compliant in perpetuity. It’s clear that the goal is to encourage results reporting, not embarrass responsible parties (despite the headlines). The choice to use 13 months (including the 30 day grace period to resolve noncompliance, as would be allowed by the FDA) and the focus on post-final rule reporting are examples of the reasonableness of their approach.

Similarly, I appreciate how the authors acknowledged the ambiguity around the “submitted” vs “posted” date. For example, [my institution] has a record with results that were due on [redacted]. We released the results on [redacted], and they are under PRS review and therefore late. However, this tracker indicates that the results were reported, not that they are overdue.”

We are pleased to hear that our tracker is already having real-world impact, and helping compliance staff improve trial reporting at their universities. In addition, it’s great to see that staff like these - who work with the FDA Amendments Act all day long - approve of how we have implemented this complex legislation. We have set out to be constructive with our features; and conservative in how we implement the law in code, to ensure that we only identify genuine breaches.

Our methods paper is online already. We will be publishing our journal paper on the FDAAA tracker shortly, describing: the process of launching such a tool; the community reaction to it; and the data on compliance so far. We already have some exciting ideas on how to expand the functionality of the tracker. Constructive and specific feedback is always helpful. If you’re finding our TrialsTracker useful, or have specific feedback about our methods and the website, we always want hear from you. Similarly, if you’d like to discuss funding our work, to help us deliver more, and for longer, then again, do please get in touch!