The Public Accounts Committee (PAC) report released last week looked back at access to clinical trial information and stockpiling of Tamiflu. While it raised some interesting and welcome points, I feel the recommendations and conclusions relating to clinical trial transparency don’t fully reflect the current position.
I am certainly not claiming that there is no problem. I strongly believe that clinical trial transparency is a global issue that we need to continue to work on. It also does not apply only to the pharmaceutical industry; all those involved in clinical research recognise the need to work towards greater transparency of trial information. However, statements from the PAC such as “research suggests that the probability of completed trials being published is roughly 50 per cent" seems to imply that this is still the case today. I do not believe that to be so for clinical trials sponsored by industry.
The global industry first signed up to a voluntary Industry Joint Position in 2005 that agreed clinical trials should be registered in a public registry and the results of those trials should be made available publically within 12 months of marketing approval of the medicine or trial completion (whichever was the later date). Since 2007 the industry has also followed the requirements of the FDA Amendments Act, which makes it mandatory to register certain trials and post their results on clinicaltrials.gov. In 2008 the ABPI included a requirement to register and report clinical trial results in our Code of Practice.
In order to assess the impact of the above changes on public reporting of trial results as well as benchmark the current environment, the ABPI commissioned research in late 2012. Published in November 2013, this peer reviewed work examined the rates of disclosure of clinical trial results for all the clinical trials that companies conducted in patients in support of all of the new medicines that got approved by the European Medicines Agency (EMA) during 2009 – 2011. This included trials conducted over the preceding 10 or more years – i.e. during the entire development program for that medicine. The research shows a positive trajectory towards greater transparency, with almost nine out of ten of all industry sponsored clinical trials having at least some results placed in the public domain by 31 January 2013 on those medicines launched in Europe between 2009 – 2011. If a 12 month cut off from the medicines marketing approval or trial completion is applied, then 76% of those trials had some results disclosed within that time limit. No doubt there is room for improvement, but these figures clearly demonstrate that industry is “on the case” with respect to public disclosure of trial results.
In December 2013 the ABPI, in partnership with the Statisticians in the Pharmaceutical Industry (PSI), held a very productive second workshop on clinical trial transparency. This brought together speakers from industry, academia, research funders, the Cochrane Collaboration and AllTrials and a cross-section of delegates from many key stakeholders involved in discussions on clinical research transparency. A very lively and stimulating discussion point of the workshop was how to disclose historical trial data and the challenges this presents, some of which have not been publically recognised by all stakeholders.
Understandably pharmaceutical companies are conscious that if they say they are going to disclose all their trial data, they need to be absolutely confident that this is achievable and doesn’t risk patient confidentiality. Retention of clinical trial data has been covered by EU legislation since 2001. We do of course expect companies to abide by this and the many other related pieces of legislation, both national and international, that cover archiving of trial information. However difficulties can arise when you look for historic data from the pre-legislation and Good Clinical Practice era and this needs to be recognised by all parties so that we can work productively and within the limits of what is possible.
A number of companies have already implemented specific processes for seeking access to historical trial data and where a company-specific process does not already exist, the ABPI recommends that requests for access to datasets on past trials should be made directly to trial sponsor companies. On reviewing these requests companies may take into account factors such as the purpose and scientific validity of the request and any logistical/legal/technical challenges in identifying and sharing older datasets and, in some cases, the information to address the scientific question may be available from alternative sources.
To facilitate clinical trial transparency, we have already launched a clinical trial disclosure toolkit to assist companies in registering and reporting results and will continue to engage with key stakeholders on this issue, which is the best way to make progress. Meanwhile our European and American counterparts are driving things forward internationally through their ‘Principles for Responsible Data Sharing.’ I feel very positive about the direction of travel on this issue for our industry and look forward to building on this in 2014 and beyond.
Dr Bina Rawal
ABPI Research, Medical and Innovation Director