facebook twitter google plus youtube flickr
President and Co-Founder, CMPI
Read Peter Pitts' bio
The debate around clinical trial transparency is certainly nothing new, with some arguing that the Clinicaltrials.gov registration requirement in the Food and Drug Administration Amendments Act (FDAAA) isn't being properly implemented and that more transparency (and oversight) is needed to satisfy full public disclosure and address the issue.
Unfortunately, it’s just not that simple.
As the former senior government official in charge of clinicaltrials.gov, I think it’s important to look at the facts - and the numbers.
In 2000, the National Institutes of Health (NIH) launched ClinicalTrials.gov to provide public access to information on clinical studies. Although it initially contained information primarily on NIH-funded research, it has been expanded to include both publicly and privately supported clinical research. Since the launch of the site, it has been enhanced to significantly increase data sharing, and the database now includes information on nearly 140,000 clinical trials in all 50 states and 182 countries.
And it’s being utilized. The NIH reported last year that ClinicalTrials.gov “receives more than 95 million page views per month and 60,000 unique visitors daily.”
As we progress through the many factors involved in data transparency, here are some issues we need to consider:
According to Dr. Richard Moscicki, CDER’s new Deputy Director for Science Operations, there’s a “transparency dichotomy” of “the promise versus fear and loathing.” As to “why” transparency is promising, he offers reproducibility, re-analysis of the potential to identify new information (placebo effects, biomarkers, endpoints, trial designs). According to comments Moscicki made at a recent conference on this issue, one group that is silently against transparency is academics – because they don’t want to be cornered into making studies public if it impacts their ability to publish.
The FDA’s impediment to data sharing, per Moscicki, is legal (data ownership, HIPAA/privacy, proprietary information), technical/practical (format, data standards, CDISC, redaction), a question of resources and the agency’s need to focus on its key mission.
On that last point, he shared that the FDA does not view (at least as of right now) the issue of clinical trial data transparency as a key agency agenda item – unless there was a move to move it to the head of the regulatory queue via user fees. But transparency is important and, per Moscicki, “inevitable” – and to that end he discussed the agency’s recent Federal Register notice (Masked and De-identified Non-Summary Safety and Efficacy Data).
Moscicki stressed that FDA’s approach has been under development for several years and the agency is not contemplating routine preparation and release of de-identified and masked clinical and non-clinical study data. The agency is encouraging independently organized efforts to create, curate and share clinical trial datasets from all sources.
Maybe the FDA’s incremental and collaborative approach is best. Slow and steady isn’t always sexy, but with respect to public sharing of clinical trials data, it is in the best interest of the public health.
Executive Director, MRCT
Read Rebecca Li's bio
Clinical trial data represent important scientific resources, and the potential public health and health care delivery benefits are far-reaching. These significant benefits can be leveraged to a greater degree if patient level data are made more freely available to researchers, accompanied by appropriate safeguards. Increased access to patient level data will, however, inevitably lead to a greater risk of breaching patient privacy. The optimal balance between privacy, research access, and incentives for innovation will need to be established prior to the adoption of any regulatory imperatives to release patient-level data.
The Multi-Regional Clinical Trials Center at Harvard’s multi-stakeholder workgroup on Models of Clinical Trial Data Sharing and Transparency has examined these issues in depth over the past year. Through our work in this area, several imperative principles have emerged on how best to balance patient privacy, innovation and public health interests:
1) Provide access to data in cases in which access to the patient-level data is necessary to answer a relevant scientific question.
2) Provide the minimum amount of data that is required to answer a specific scientific question.
3) Recommend a model that incorporates the principles of a ‘learned intermediary’ or ‘controlled access to data’. This model would not necessarily preclude data generators from providing clinical trial data to researchers directly, based on these principles.
4) Ensure that safeguards are in place to minimize the possibility of re-identification of patients, adequately protects patient privacy and also adheres to the terms of the informed consent. A robust data-use agreement could serve this purpose.
5) Assure accountability of both data generators and data requesters for responsible use of data and analyses.
6) Treat equally all sponsors of data, including industry, academia and government.
7) Make practical any proposed system of data sharing, minimizing the cost and burden to data generators and requestors, and assuring transnational adoption of uniform rules, to lower transaction costs.
If the above practices are implemented, this will serve to benefit all parties and ensure that the full value of clinical trial data is leveraged, while also balancing the rights of patients and commercial interests.
Lastly, policymakers must respect that participants who volunteer for clinical trials may have a different perception of public health benefit than scientists. In the dialogue that is currently ongoing, the participant voice is an important one. It is clear that they are key stakeholders and have much to lose if innovation is stifled or privacy is breached, but also much to gain if increased data sharing leads to faster drug development cycle times or if safety signals across drug classes are ascertained earlier.
Kenneth A. Getz, MBA
Director of Sponsored Research Programs, Tufts University
Read Kenneth Getz' bio
Perhaps the question, as posed, reflects a narrow and outdated perspective? The current question suggests that clinical research professionals have a choice in addressing — or choosing not to address -- elements of patient privacy, clinical trial results disclosure and proprietary innovation knowledge. That choice no longer exists. It is not a matter of balance, but rather an acknowledgement that success hinges on new models and strategies that defy conventional drug development notions of privacy, disclosure and competitive sensitivity.
It is widely accepted that the R&D paradigm that served our industry well for more than 60-years is no longer viable. Our current operating environment is fraught with large and rising costs, high levels of complexity, inefficiency and risk. My colleagues and I at the Tufts Center for the Study of Drug Development have been looking at one very promising new R&D approach -- collaborative and open innovation -- characterized by partnerships that share innovation and operating risk, better leverage expertise and capabilities, and drive higher levels of efficiency and performance. For open innovation to work effectively and successfully, collaborative partners must respect one another and must share intellectual property and proprietary knowledge.
Many open innovation approaches include patients as engaged team members or collaborative partners helping to set the research agenda and study design, providing funding support, assisting in recruiting and retaining research participants, and disseminating study findings. Honoring the patient at the center of the research team requires adhering to the highest standards of patient privacy and transparency.
The disclosure of clinical trial results is not only required by regulatory mandate, it also rests at the heart of building trust among the public and patients and ultimately engaging them as partners in the clinical research process. Sadly, Very few companies are meeting the obligation at this time to disclose clinical trial results to study volunteers but there is hope: Two years ago, my non-profit organization CISCRP (Center for Information and Study on Clinical Research Participation) began assisting sponsor companies in routinely providing non-promotional, non-technical clinical trial results summaries to study volunteers through investigative sites. Nearly two dozen companies are now providing clinical trial results to study volunteers and they are seeing improvements in patient retention rates, in volunteer comprehension and satisfaction, and in patient engagement and support in the clinical research process. The number of companies supporting this important initiative is growing.
The goal is not balance – it is acknowledging that sustainable success requires embracing and respecting open innovation models and the profound value created by collaborative teams made up of professionals and patients.
Douglas Peddicord, Ph.D.
Executive Director, ACRO
Read Douglas Peddicord's bio
As clinical research organizations (CROs), ACRO member companies are involved in the conduct of the majority of clinical trials that lead to new treatments and therapies. Working with confidential and proprietary information to facilitate rapid advances in the development of safe and effective diagnostics and medicines is at the core of our business. While we are not the “owners” of clinical trial results, maintaining the privacy of clinical trial participants is central to the mission of producing reliable and unbiased data.
ACRO endorses the “Principles for Responsible Clinical Trial Data Sharing” announced recently by EFPIA and PhRMA. Sharing clinical and pre-clinical data within a framework that protects the privacy of research participants and the confidentiality of commercial information, under a governance structure that ensures responsible and fair use of the data, has great potential for enhancing the public health. Voluntarily giving patient-level and study-level trial data, along with study protocols, to qualified scientific and medical researchers could foster genuine advances in development by allowing independent reanalysis of rich data sets and significant learning from previous successes and failures.
Notwithstanding our support for this kind of scientific collaboration, we are concerned that some so-called transparency proposals from regulators and others could have a chilling effect on clinical trial participation. It is important to note that clinical trial participation is already inadequate, and this hurts the development of new treatments and therapies.
According to a recent Research!America poll, 53.4% of Americans believe privacy and confidentiality are “very important” to their participation in clinical trials. Also, 31.3 percent of those polled – and 40.6 percent of African-Americans – believe people have been included in clinical trials without their knowledge.
When discussing the transparency issue, we must keep in mind this “trust factor” as well as privacy concerns. Surely no one advocating making all clinical trial results public would want to discourage participation in clinical trials. Participation in such trials is critical to the development of new treatments and cures.
It is important to note that in that same Research!America poll, 72.5% said they would be willing to share their personal health information to advance medical research. Perhaps regulations and informed consent documents could be revised to allow for greater research use of this valuable clinical trial data as long as patient anonymity is protected.
We strongly believe that any subject level data made available for research purposes should be de-identified consistent with the requirements of HIPAA. The HIPAA de-identification model is the most widely accepted and followed standard for de-identifying health data consistent with the appropriate protection of individual privacy. The use of any different standard will create confusion by allowing distribution of “de-identified” data utilizing differing safeguards, and potentially place research subject privacy at risk. Research participants must have confidence that their privacy is protected. In order to fully protect patient privacy, we believe this HIPAA de-identified and masked subject level data sets should be provided only to qualified researchers undertaking IRB-approved projects.