A new culture is needed to promote the equitable and open sharing of data, to advance science, whilst protecting the interests of all stakeholders, says the US Institute of Medicine

 

IOM_Sharing Data_hi res cover-smallThe sharing of clinical trials data is clearly in the public interest, but huge effort is required to develop the policies and put in place the technical infrastructure and create a culture of openness, according to a new report from the US Institute of Medicine (IOM).

 

“We know this is not an easy topic,” said Victor Dzau, President of the IOM. “There are many disincentives to data sharing. Competition, loss of control, loss of credit, giving away trade secrets, so on, and so forth.”

 

But said Dzau, the “unprecedented opportunities to advance the world’s health” mean it is essential to respond to these challenges. “We have access to an incredible wealth of data, and we have hardly begun to tap into the knowledge that it can elucidate,” Dzau said at the launch of the report on Wednesday (14 January 2015).

 

At the same time, methodological improvements in how to collect, manage, analyse and share data, have vastly improved the ability to evaluate new therapies. “We must be prepared to responsibly respond to these changes, to ensure that we are using data ethically and to its full potential,” said Dzau.

 

Sharing clinical data will allow investigators to carry out additional analyses, strengthen the evidence base for regulatory and clinical decisions, increase scientific knowledge and stimulate new ideas for research. It will also maximise the contribution to clinical research made by patients taking part in trials and avoid studies being duplicated.

 

The IOM’s report, supported by 23 sponsors, including the US Food and Drug Administration, 17 pharmaceutical companies, and clinical research funders including the US National Institutes of Health, the UK Medical Research Council and the Wellcome Trust, sets out expectations, guidelines and standards of how and when clinical trial data should be shared.

 

This matters to everyone

Everyone has an interest in responsible sharing of clinical trial data, said Bernard Lo, chair of the committee that drew up the report, outlining the main recommendations. “We are all patients or future patients, or relatives of patients. And we all really want to have the widest and most rigorous information about what therapies are effective and what the side effects are.”

 

But Lo noted, currently much of the data collected in clinical trials is not analysed and not published in a timely manner, with a third of clinical studies not made public four years after completion. “So, all of that effort, the contributions of the participants, the effort of the clinical trial staff, and also the financial funding for the trial go to naught if the results aren’t published,” Lo said.

 

“We think that the question today is not whether you share clinical trial data, but instead, what types of data do you share, when do share it, and how do you share it?”

 

Defining data sharing

 

Whilst there are already requirements in the US and Europe to share a summary of results of trials, the IOM report takes a wider sweep, defining data sharing as making data from clinical trials available for secondary research, and calling for far broader disclosure.

 

“Clinical study reports from companies presenting material to the FDA can run into thousands of pages of tables, charts, summary data,” Lo noted. Beyond that, the committee also looked at sharing (anonymised) data on the level of individual participants. And, Lo added, “We really want to emphasise that it is not just the data, it’s metadata which is trying to describe what the data are, how it is supposed to run, plus the study protocol, all versions, plus the final versions, the statistical analysis plan the investigators used, and the actual analytic code, the computing code used to calculate the published results.”

 

“We think all of that really needs to be shared to maximise the benefits.”

 

However, the IOM committee is not suggesting a free-for-all. Data sharing need not mean posting data on the web for anyone to download. Instead, it may be shared only after a request and might be shared only with certain conditions of use. “In fact we make specific recommendations on those elements of the conditions of use and other protections,” said Lo.

Fair shares

It is critical that data sharing is fair and takes account of the interests of different stakeholders. “First and foremost, we want to protect participants,” Lo said. Beyond that objective, scientists who spend long years working on conducting a trial should be given time to analyse, publish and get credit for their work.

 

However, it is also important that independent investigators get access to carry out additional analyses and at the same time, that pharma companies can protect their intellectual property and commercially confidential information.

 

In the face of countervailing interests, sharing data should be underpinned by the vision of advancing the science as the foundation of medical care. In future, the norm should be that data is shared, and to support this there should be incentives and protections.

 

At a technical level, there need to be multiple, inter-operating platforms and work to identify best practice in data sharing. “People are doing it in a lot of different ways. Let’s find out what works and how we can improve,” Lo said. “We want the way that we share data to be continuously modified, improved and that process of change to be driven by data, by evidence.”

 data-sharing-cycleclick for full scale image 

To make the committee’s recommendations feasible, there needs to be a funding model for sharing clinical trial data that is sustainable and equitable.

“Big tasks, big challenges, we could not address them all, but we could at least point to a path forward,” said Lo.


 

 

IOM_Sharing Data_hi res cover-smallLink to summary of the report : Sharing Clinical Trial Data: Maximizing Benefits, Minimizing Risk

Link to full report : Sharing Clinical Trial Data: Maximizing Benefits, Minimizing Risk

 


Report Sponsors