An AllTrials project

Non-reporting of clinical trial results is an ongoing global public health problem.

The best currently available evidence shows that the results of clinical trials are routinely withheld from doctors, researchers, and patients. This is a global public health scandal.

How do we contact you?

Email us! [email protected]. We're always keen to hear feedback, user stories, insights you’ve made, errors, omissions, and feature requests.

What’s this all about?

The FDA Amendments Act 2007 (“FDAAA”) is a law which says that certain clinical trials have to report their results, within 12 months of completion. This is really important. Doctors and patients use the results of clinical trials to make informed choices about which medicine works best. But the best currently available evidence shows that trial results are routinely withheld. Previously, laws and guidelines aiming to fix this issue have been widely ignored, despite energetic campaigns. So we built this website, to increase public accountability, and to help researchers make sure they report all their trials. You can read more about the problem in general, and this website, in our paper on the project.

Who made the FDAAA TrialsTracker?

We are the Evidence Based Medicine DataLab at the University of Oxford, a team of academics, software engineers, and clinicians working together to make live interactive tools as well as academic papers. Seb Bacon was the software engineer; Nick DeVito the researcher; Ben Goldacre devised the project and was PI; Ben O’Brien was the illustrator; everyone in the DataLab team contributed. You can read more about our all work - on research integrity, and health informatics more broadly - at ebmdatalab.net

Are you publishing a paper on FDAAA compliance?

We certainly are, yes. We have a forthcoming paper on reporting rates in compliance with FDAAA; and on the response to this TrialsTracker. You can read about our team's other research outputs on research transparency and clinical informatics at ebmdatalab.net.

I’m a nerd! How is this website built, and what assumptions do you make about the data?

Our paper is the best place to read the full technical details of how this site was made, how the data is processed, how the law operates, and how we translate that into code.

There is an API. It's somewhat self-explanatory; however, if you would like to use it and need some help get in touch.

If you want to read the code for the website, it’s on Github here.

How can I help the FDAAA TrialsTracker?

In many, many ways!

What can I do to help make sure sponsors share their results?

Trial reporting is a huge, global, ongoing problem. But there are more people working towards a solution than ever before.

  • The AllTrials.net campaign calls for all trials to be reported, and now has the support of over 700 professional bodies, patient groups, government organisations, industry bodies, research funders, and more. Join them, support their work, and encourage your organisations to sign up to their simple pledge.
  • You can advocate for better trial reporting in your own work community.
  • You can use our data to encourage trialists to report their trials.
  • You can call your elected representatives and tell them to pressure the FDA to levy fines and publicly censure researchers and sponsors who fail to report their trial results.

I think some trials on clinicaltrials.gov are missing from your site! Doesn’t the law cover all trials?

Sadly, FDAAA 2007 only covers a narrow subgroup of trials. We think all trials should report results: you can read more about the AllTrials.net campaign online. On this website we track only compliance with the law. We have a range of “TrialsTrackers” which monitor reporting for other kinds of clinical trial at TrialsTracker.net

What is the difference between an ACT and a pACT?

ACT stands for “Applicable Clinical Trial”. These are the specific trials required to report results under FDAAA. ACT, in the context of this tracker, is specifically referring to trials that began after the effective date of the 2016 “Final Rule” (January 18, 2017). While the Final Rule created much clearer criteria for identifying an ACT, trials that began before, and end after, the effective date may be required to report, but data on the criteria for judging whether they are is less available. These trials are called “probable ACTs”, or “pACTS” (by everybody, not just us!) and are identified by a different methodology. We have added a tag to every trial so you know which methodology was used to identify that trial. You can read the details on ACTs and pACTs in our paper.

If you think we have wrongly identified a trial, please get in touch!

I found a trial I think is due to report results, under FDAAA, but it’s not listed as due?

There are a number of reasons why a trial that completed 12 months ago may not yet be due to report results under FDAAA.

  • The trial may have become due within the last month: After the trial becomes due, even when the results are submitted to ClinicalTrials.gov, there is a 30 day administrative window to post results. If a trial has submitted results, even if they haven’t been fully posted yet, this should be noted in the study record, and our “webscraping” tool will automatically find that information. However, to be safe, and to account for any updating delays by ClinicalTrials.gov, we wait this entire 30 day period before calling a trial “overdue”. If results show up in the meantime, we will show them as reported. You can read more detail on this in our paper.
  • The sponsor may have applied for an exemption: Sponsors can apply for a certificates that can delay the posting of results for certain trials, for up to 2 additional years. You can see if this is the case by clicking on the “Results’ tab on ClinicalTrials.gov. We might start flagging this on our website too, if it’s useful and interesting for you.
  • The trial may not be required to report results under FDAAA: Unfortunately, only a subset of all trials registered on ClinicalTrials.gov are required to report results: essentially the law covers trials on drugs and devices: so a trial on a surgical technique, a public health intervention to reduce smoking, or a talking treatment for mental health problems, would likely be exempt. The trials covered by FDAAA are known as “Applicable Clinical Trials” or ACTs. Unhelpfully, although ClinicalTrials.gov know which trials are ACTs, they do not make this information public. We are, however, able to infer whether a trial is an ACT by using other publicly available data. You can read in detail about how we do this, in our paper.
  • We may not have updated our data yet: If your trial became due very recently, you may need to wait for a data refresh. We’re planning to refresh the data on this site daily, at first, resources permitting. You can check in the footer of the website to see when we last updated our data, and when the next update is planned.

If after reading this you think we are missing a due trial, please get in touch!

I found a journal article/conference abstract/other document containing results from this study. Why are you saying it has unreported results?

The FDAAA 2007 is a law that requires all applicable trials to report their results on ClinicalTrials.gov within 12 months of completion. This is to ensure that results are shared in a timely manner without editorial delays; to ensure that results can be easily located; and to ensure that results are completely and correctly reported. FDAAA sets out a standard format for reporting clinical trials. Previous research has shown that overall journal publications are significantly less complete than structured reports on ClinicalTrials.gov: for example, journals often permit misreporting of prespecified outcomes, and are less complete on safety data. Journal reporting is often interesting and useful but the law requires results to be reported on ClinicalTrials.gov.

This trial’s primary completion date has passed, but it has not yet reached its completion date. Why are you saying it is required to report?

Under FDAAA trials are required to report within 12 months of their “Primary Completion Date”. The “Completion Date” is a different data field. These data fields have very specific meanings and definitions on ClinicalTrials.gov. The Primary Completion Date is the last visit of the last subject for measurement of the primary outcomes. Some trials may additionally continue to follow-up with patients for weeks, months, or even years after they finish collecting data on their primary outcomes. The responsible party is required to report the results for the primary outcomes 12 months after the Primary Completion Date. We use the “Completion Date” field to measure when a trial is due to report if the “Primary Completion Date” field is missing. ClinicalTrials.gov allows trial results records to be updated and amended, as needed, to accommodate additional results if and when they become available.

How do you calculate what fines a sponsor could receive?

The FDA can fine (per 333(f)(3)(B) in table here) non-reporting sponsors up to $11,569 a day following a 30 day notice period. We calculate the maximum possible fine assuming the sponsor was notified the day after they failed to meet their submission deadline. So each day after 1 year plus 30 days since trial completion without submitted results is another $11,569. You can read more on this in our paper.

How can I cite this work?

Please cite the website as:

FDAAA TrialsTracker. EBM DataLab. 2018. Available from: https://fdaaa.trialstracker.net/

For citations regarding our methods, please cite our preprint paper:

DeVito NJ, Bacon S, Goldacre B. FDAAA TrialsTracker: A live informatics tool to monitor compliance with FDA requirements to report clinical trial results. bioRxiv. 2018; https://doi.org/10.1101/266452.

When was the data last updated?

The current data was taken from ClinicalTrials.gov on 20 August 2018. Our next update of the data is planned for 22 August 2018.