About

What is this?

This is the COVID-19 TrialsTracker, part of our larger TrialsTracker.net project. Here we will bring together structured, cleaned data from clinical trial registries on studies of COVID-19 and track the availability of their results. Currently, we get our trial data from the ICTRP which holds structured data from 18 individual trial registries including ClinicalTrials.gov, updated weekly or monthly depending on the registry. We may collect additional information directly from certain registries over time.

Initial data on COVID-19 trials was published as part of the Oxford COVID-19 Evidence Service

This service is a work in progress. We hope to iterate, expand, and improve the Tracker in the coming days, weeks, and months as time and resources allow.

How does this all work?

Each week we download the ICTRP COVID-19 dataset and run it through some code to clean up and standardise the data. We then add additional information on the interventions involved in each study, assess for cross-registration, add information on any results we know about, make some basic data visualisations, and then post it all online as a free resource for anyone to view or use!

A few notes about the data:

We've limited the dataset to all newly registered studies since the start of 2020, so some older trials that have added a focus on COVID-19 may be excluded.
In addition we've removed any trials that have identified as "Withdrawn" - that is the trial never happened and this is noted in the registry entry.
If we know about a trial that the ICTRP dataset has missed, we add it manually.
Cross-registrations are collapsed into a single parent entry with the known cross registrations noted in the last column of the dataset (for instance, all registered versions of the WHO SOLIDARITY trial are collapsed to the ISRCTN entry). We usually default to the ClinicalTrials.gov registration where possible.
Most data on the tracker is automatically taken or derived from the ICTRP dataset. We try to catch the most blantant nonsense automatically but we can't manually verify every datafield so mistakes or conflicting information in registration data given to the ICTRP are inherited by our dataset.
Normalisation of sponsor names, extracting information about the interventions in each study, accounting for cross-registration, and adding information on completion dates and results is done manually. If you spot mistakes, typos, or have any suggestions to improve this, please get in touch.

Can I use your data for...?

Yes! No need to even ask. Our dataset, and everything else about this project, is fully open and available for anyone to grab and use for their own purposes. Attribution is of course appreciated. If you do find our data useful and do something amazing with it, please share it back! We are excited to see what people can do!

Who made this?

The TrialsTracker project is run by the Bennett Institute for Applied Data Science at the University of Oxford. The COVID-19 TrialsTracker is maintained by a team of researchers, coders, and clinicians across the Bennett Institute and The Centre for Evidence-Based Medicine (CEBM) with additional input from external collaborators.

Contributors to Date

Project Lead

Nicholas DeVito (Bennett Institute for Applied Data Science/CEBM, University of Oxford)

Analysis, Interpretation, and Visualisation

Jeffrey K Aronson (CEBM, University of Oxford)
Henry Drysdale (Bennett Institute, University of Oxford)
Michael Liu (Bennett Institute, University of Oxford)(
Helen Curtis (Bennett Institute, University of Oxford)
Carl Heneghan (CEBM, University of Oxford)
Robin Ferner (University of Birmingham)

Website Development Help

Peter Inglesby (Bennett Institute, University of Oxford)
Seb Bacon (Bennett Institute, University of Oxford)
Ben Goldacre (Bennett Institute, University of OXford)

Data Extraction Volunteers

Florence Rogers (UAEM, Imperial College London)
Sarai Keestra (UAEM, University of Oxford)
Maymunah Malik (UAEM, University of Manchester)
Holly Melvin (UAEM, University of Manchester)
Tricia Tay (UAEM, University of Manchester)
Frances Kenworthy (UAEM, University of Manchester)
Till Bruckner (TranspariMED)

In addition, many thanks to the websites listed below in the Other Resources section many of which we use to cross-check our data!

How can I help?

We are open to improving the Tracker, either in content or appearance, through collaboration with committed individuals with specific expertise. If you have time, knowledge, and an idea to improve a specific aspect of the project, please get in touch at [email protected]. See below for other specific opportunities to contribute.

Ways to contribute

If you come across mistakes, omissions, or information you know about that we don't account for in our data you can submit them here.
Let us know if you think you have come across information on unregistered trials.
Suggestions for improvements to our categorisation and normalisation schema. Especially for Chinese sponsors.
Download our data and use it to make amazing data visualisations! If you would like this to be considered for inclusion on the Tracker (with full credit given) please get in touch.
Please see our Github below if you are interested in contributing to our code base.

The Github page for the project is located here. This includes our data cleaning notebook, normalisation and manual data collection files, and the website html files. Everything is available under open MIT Licensing.