The Data Sharing Landscape in Clinical Trials
By Maxine Bookbinder
August 6, 2014 | In 2003, The National Institutes of Health (NIH) said in its Final NIH Statement on Sharing Research Data that “NIH reaffirms its support for the concept of data sharing. We believe that data sharing is essential for expedited translation of research results into knowledge, products, and procedures to improve human health.”
But for many companies, sharing information outside company walls was likened to comforting the enemy. Particularly left out of the loop were clinical trial patients who rarely saw their own files, and test redundancies and product failures were part of the process.
But the rules are changing. Competition is now collaboration, transparency is golden, and discovering cures trumps swelling market shares. Craig Lipset, head of Clinical Innovation at Pfizer, describes data sharing as “a collaborative effort” to pull data together for the benefit of other studies, to move faster on current research, and to develop new discoveries. Data sharing collaborations, says Lipset, “want to ensure that data is as open and accessible as possible to scientists and researchers.”
Not everyone, however, has adapted. According to a study published in PLOS One*, 30% of 400 randomly selected 2008 trials were never published, either in a journal or on ClinicalTrials.gov, by 2012. Smaller, Phase II trials funded only by pharma were even less likely to be published. Singling out Phase III studies for publication may mean that only successful studies—not all studies—are published. Article author Christopher Gill told the Wall Street Journal that “the majority of scientific discovery from industry is not being published.”
Many experts argue the scientific community has a responsibility to trial participants and to the public to be transparent, to publish all data, to find new discoveries and to save lives as quickly, efficiently – and honestly – as possible. Secrecy can lead to short-cuts and scares, evidenced by drugs removed from market due to safety issues and health emergencies.
The Investigator Registry is proof of the industry’s evolution. Many companies have been pooling and normalizing their data, as demonstrated through a precursor project known as the Investigator DataBank. Johnson & Johnson, Eli Lilly, Novartis, Merck and Pfizer began pooling their investigator data about two years ago. Today, investigators can update and share information. As Lipset notes, “It is ultimately good for all—research sponsors remove overhead and use larger amounts of data to find appropriate sites, investigators are able to share the experience and capabilities with a large number of research sponsors.”
In May 2013, GlaxoSmithKline initiated its Clinical Study Requests website, listing all studies, as of that date, completed or terminated, giving external researchers access to the company’s trial results. “It provides transparency, adds knowledge, and is the right thing to do scientifically and for society,” says Perry Nisen, M.D., Ph.D, senior vice-president of Science and Innovation for GlaxoSmithKline.
The website is updated at least twice yearly and studies are added within six months after inclusion approval. GSK is in the process of adding global studies from 2000 to the present onto the website; it already posts summary information about every trial, regardless of outcome at https://clinicalstudydata.gsk.com, lists almost 5,000 clinical trial summaries, and averages nearly 11,000 visitors monthly.
To ensure confidentiality, GSK removes all personally identifiable information from the data set, recodes identifiers, removes free-text verbatim terms, uses age, not birth date, and revises all dates relative to a random anchor. To meet certain European requirements, the key code between the provided data set and the original data set is destroyed. Investigators must agree to not try to identify subjects.
Non-profits have embraced data sharing, as well. TransCelerate, a non-profit collaboration of 18 pharmas created almost two years ago, is a convergence of companies collaborating to share pre-competitive data to decrease trial times and expenditures and to increase completion success rates.
At the Summit for Clinical Ops Executives in Miami in February, Jackie Kent of Eli Lilly said that TransCelerate has so far focused on data standards, but has made available several valuable tools (see, Collaboration and Competition in Clinical Trials). The consortium is developing an Investigator Registry as a shared resource regarding investigators worldwide. It is also exploring a shared standard-of-care dataset, bringing together the control arm data from multiple sponsors.
The Coalition of Major Diseases (CAMD), a non-profit alliance of government, industry, and non-profit organizations, researches neurodegenerative diseases, such as Alzheimer’s and Parkinson’s, in order to better understand these illnesses, improve assessment tools, and possibly develop new therapies. A note on its website states, “Resources required for the success of these efforts exceed the ability of single organizations, and will require the collaborative efforts of multiple partners from industry, academia, patient advocacy groups and government.”
Even as data sharing efforts grow, the clinical trial patient is still often left out. Pfizer’s 2013 Blue Button project was a first of its kind program that allows clinical trial patients to access their own health data. It’s a model that has a great return on investment, Lipset says.
“Sharing data with patients improves engagement with trial participants,” he says. “It allows them to feel more as active participants and that they are getting something back.” Patient contentment improves retention, increasing trial completion and success rates.
Sharing Patient Data in the Clinic
At the point of care, direct patient care data sharing, or electronic health information exchange (HIE), also saves cost, time, and lives. HIEs are currently used in conjunction with some electronic health record (EHR) systems in a subset of the 6,000 U.S. hospitals through various vendors as well as international facilities. They are most frequently used to share information regionally between primary care physicians and specialists, but their applications could extend far beyond that.
“It’s time to take blinders off,” says Christopher Longhurst, MD, pediatrician, Chief Medical Information Officer, and Vice-president of Data Analytics and Informatics at Stanford Children’s Health. “This (HIE) offers more timely care and a continuum of care. It’s a strategic disadvantage not to exchange information. This helps us offer integrated care.”
Dr. Longhurst recently saw a family in the hospital who brought a 3-month-old baby born with a hole in his heart. “They were visiting from North Carolina and didn’t know many specifics. I made an electronic match to the baby’s records in Chapel Hill in about 22 seconds. I saw the doctor’s notes, the genetic and echocardiogram test results. If this was six months ago, I would have had to call Chapel Hill and ask to have the records faxed.”
Using HIEs prevented additional and redundant tests and helped diagnose a mild bleeding disorder. When the family returned home, the baby’s Chapel Hill pediatric hematologist had full access to the Stanford files.
Dr. Longhurst says that Stanford launched its HIE in early May and physicians have already used it hundreds of times to access thousands of patient care records. “Of course, we need to follow the rules of the road on how data is exchanged,” notes the doctor. Patients must give consent and records can only be accessed with other HIE-participating hospitals, which must agree to upload information in order to also download it.
For all of the advantages to data sharing, the issues of data privacy risks and unethical usage remain. Dr. Nisen of GSK questions whether small companies and academia will be able to successfully follow Big Pharma’s lead. “We have invested millions to do this. It will be a challenge to accommodate diverse expectations across the scientific and medical community. Is there a means and mechanisms by which everyone can do this?”
Updating the status quo still brings promise of improvements. “The community,” says Lipset, “is still trying to figure it out with many different approaches that will help lead toward best practices.”
* Saito H, Gill CJ 2014, How Frequently Do the results from Completed US Clinical Trials Enter the Public Domain? – A Statistical Analysis of the ClinicalTrials.gov Database. PloS ONE 9(7): e101826.doi:10.1371/journal.pone.0101826
Text Only 2000 character limit
Page 1 of 1