Clinical trial reporting by European universities – best and worst performers
New data compiled by TranspariMED shows that major universities across Europe have uploaded a record number of clinical trial results over the past year.
Many universities are now racing to make hidden clinical trial results public – but some are still not meeting their obligation to fully share medical research findings.
Just over a year ago, the 29 largest medical universities had only made 162 of their clinical trial results public on the European trial registry. Since then, they have uploaded an additional 297 trial results, bringing the total number of trials with results available to 459.
This is excellent news for medical researchers and patients worldwide. Posting results onto trial registries makes more, and more reliable, information on the benefits and harms of drugs publicly available.
The chart below shows the performance of the top ten universities in terms of uploading missing clinical trial results.
UK universities were the first to tackle the problem, and therefore dominate the list. Imperial College alone has uploaded 69 additional results since March 2019, and has by now achieved an outstanding reporting rate of over 96%.
Efforts by leading universities in Austria, Germany, Belgium and Denmark to fix the problem started later, but are rapidly gaining speed.
The table below shows how many clinical trial results each of the 29 universities has uploaded over the past year.
13 universities have uploaded three or more trial results. These universities are presumably tackling the problem systematically, and on track to fix it.
9 universities have uploaded only one or two trial results each. Many of these institutions are probably committed to fixing the problem, but it is too early to tell. Future TranspariMED updates will track their progress.
7 universities have not uploaded any clinical trial results to date.
If your university is not listed above, you can type its name into the EU Trials Tracker search window to check its performance.
Major Italian and French institutions are rapidly falling behind their European counterparts.
None of the three Italian universities have made any of their clinical trial results public yet.
The performance of the two French institutions is equally dismal.
Neither AP-HP nor Hospices Civils de Lyon have yet uploaded a single clinical trial result. (Contrary to what the table further above suggests, AP-HP has not itself uploaded any trial results; instead, a different sponsor posted the result of a multi-country trial in which AP-HP had participated.)
TranspariMED and allied groups will continue to encourage universities, funders and regulators to ensure that all clinical trial results are made public.
TranspariMED’s collection of transparency tools provides guidance for universities that wish to improve their performance.
NOTE ON METHODOLOGY
The data above provide an update on a report published in April 2019 by TranspariMED, BUKO Pharma-Kampagne, Test Aankoop / Test Achats and Health Action International.
The cohort comprises the 29 non-commercial trial sponsors in Europe that had sponsored the largest number of clinical trials listed on the European trial registry, the EU Clinical Trial Register, as of end March 2019. Data was extracted using the EU Trials Tracker and reflects data publicly available on the EU registry at the end of August 2020.
The use of a single metric – due trial results uploaded – is deliberate.
In keeping with TranspariMED’s standard methodology, which reflects that used by the EU Trials Tracker, only results uploaded for trials that are verifiably identifiable as ‘due to report results’ from the trial protocol are included.
In many countries, universities typically have dozens of clinical trials in their portfolios that are falsely marked as ‘ongoing’, making it impossible to determine the true number of trials that are due to report results according to EU transparency rules. In this context, it does not make sense to publish cross-country data on the percentage of due trials with results, or on the number of trials still missing results.
In addition, when a university first addresses the problem, it may request its national regulator to correct data on false ‘ongoing’ trials and mark them as completed. This increases the number of trials shown as overdue by the EU Trials Tracker. Paradoxically, this means that Tracker data might show a decline in the percentage of due trials with results for that university during the first few months of its efforts.
TranspariMED’s choice of metric to track progress thus avoids creating perverse incentives.
On the downside, TranspariMED’s metric does not place universities’ efforts in the context of the overall size of their trial portfolios, or the size of their backlogs of overdue results. It also fails to fully reflect the efforts of universities (such as Charite) that are making rapid progress, but only started the process recently. It also undercounts the efforts of universities (such as King’s College London) that started uploading trial results before March 2019.
In future, TranspariMED will provide regular updates on universities’ performance to ensure that all institutions, including late movers, get full credit for their efforts. This will include updates on national performance (example here), so smaller universities' efforts will also become visible.
Notes on individual sponsors:
Heidelberg. In keeping with the methodology used in the 2019 European report, the data above only includes trials sponsored by Heidelberg University itself. Heidelberg's performance is far more impressive when trials sponsored by Heidelberg University Hospital are also taken into account.
Danish sponsors. Data provided in the 2019 report for one sponsor, “Copenhagen University and Hospital,” aggregated the performance data of several Danish trial sponsors. This sponsor is therefore no longer listed above. The EU Trials Tracker now lists the various Danish sponsors separately. Overall, Danish sponsors appear to be making rapid progress.