News & Events

Dr. John Frew, the first Pioneer Award winner, from Rockefeller University, sat down for an interview with Vivli to share more about his research, the impact his findings are having on future clinical trial design and patient care, as well as his advice to researchers who were going to embark on a research project using individual-patient level data from completed clinical trials.

Congratulations on being named the first Vivli Pioneer Award winner. Can you tell us more about your research?

Thank you very much. Firstly, I’d like to acknowledge that this work is definitely a team effort. It would not be possible without our excellent statistical team: Neha Singh, Caroline Jiang and Prof Roger Vaughan; as well as our clinical research team Kristina Navrazhina, David Grand and Prof James Krueger, and, of course, Vivli and AbbVie, the data provider. Our research focused on Hidradenitis Suppurativa, which is a chronic inflammatory skin disease. Currently, there are very few approved therapies to treat it. There are also very few large-scale clinical trials that examine this condition. By going back and reviewing these large-scale clinical trials we can retrospectively look at the population with the condition we are researching.

The first aim of our research project focused on the utility of the outcome measures used in the completed clinical trials. We wanted to have a deeper analysis of the existing data sets on Hidradenitis Suppurativa prior to moving forward with expensive, long and large clinical trials and see if we need to reconsider the outcome measures used or develop new, or novel, outcome measures. We want to avoid the issue that have plagued clinical trials in this area that have failed due to their design.
The second aim of our research project was to look at the individual patient-level data to look at patient characteristics and see if they would inform response rates to therapy.

What were the results of your research?

In our recent publication in the Journal of the American Academy of Dermatology, our research showed that there were multiple outcomes measures for this disease and depending on which outcome measures were used, changed the efficacy of the treatment. We showed that additional confounding effects weren’t being taken into account in those outcome measures.
This research was able to provide an evidence base for the anecdotal comments we had heard from dermatologists and immunologists about the what they were observing in patients with Hidradenitis Suppurativa.

What difference do you hope your research will make?

Actually, I was at a conference in Europe last month and a half a dozen speakers were citing my research as the basis for re-thinking trial designs and the outcome measures used and whether multiple measures need to used or a new one developed to take these findings into account.

In addition, I had a European counterpart come up to me and say how important it was to not only be asking the right questions but to have a large, reliable data set to interrogate with these questions. Having all this data available via the Vivli platform is a key tenet to evaluating the epidemiological aspects of this condition. This is beyond valuable.

What’s next in your research?

These initial findings have validated our approaches to on-going analysis and confirmed that this work is more than “just re-analysis”. This work is highly useful with high utility and direct clinical applicability in the design of clinical trials and patient care.
Developing an evidence base for the treatment of Hidradenitis Suppurativa is really in its infancy and so being able to stimulate a discussion about the best way to design clinical trials using data at the individual patient-level is invaluable.

It has also opened up new lines of enquiry for our work, particularly in terms of mechanistic biochemical studies that we are doing in the lab with tissues and considering new variables.

That’s so exciting. Can you tell us more about the process of requesting data using Vivli?

I had never come across a concept such as the model that Vivli employs providing access in collaboration with companies who produced the data. It is a unique and fantastic opportunity. I found the user interface easy and logical to use. I cannot commend and recommend Vivli enough for the degrees of support and transparency throughout the whole process. It has been an absolute breeze. The data itself was of such a high quality. The clinical trials were developed to answer a very specific question but so much more can be done with the data beyond answering the original question it set out to address.

How did you find working in the research environment?

I think for me personally it was a bit of a learning curve. While I had a statistician that I worked with, it was largely me doing the coding from scratch. I was worried at first about the speed of access working in the secure research environment, but this was never an issue and I really appreciated the flexibility of being able to access the research environment from home and work.

I really loved the Chat function. It was so smooth to be able to ask the data providers and Vivli a question directly and to be able to get a response in a few hours was excellent.

What advice would you give to other researchers who want to request and analyze data via the Vivli platform?

Ask for help early and don’t try to bang your head against a brick wall. Make sure you work closely with your statistician every step of the way. I originally extracted some data out of the original data sets myself and had to go back and re-do it after discussions with the statistician.

I can see now why Vivli has taken the need for a statistician into account at the very beginning and have thought that out. You just have to trust their system and follow it through.

Vivli launched its first data analysis challenge shortly after our launch in an effort to generate interest and move data sharing forward. The Vivli Pioneer Award was conferred to the first approved 10 data requests that initiate analysis. The lead investigators of these data requests received $1,000 awards. The Vivli Global Data Sharing Innovator Award will be conferred to the first 10 data requests that complete and publish their analysis in a peer-reviewed journal. The lead investigator of the data request team (or designated team member) will receive a travel award to enable presenting these results at a future global Vivli Data Sharing meeting.

The Vivli Pioneer Award recipients are:

John Frew, Rockefeller University
Vojtech Huser, National Library of Medicine/NIH
Akira Kimata, University of Tsukuba
Frederikus Klok, Leiden University Medical Center
Vivek Rudrapatna, UCSF
Changyu Shen, Beth Israel Deaconess Medical Center
Mirjana Stanic Benic, University Hospital Centre Rijeka
Sharon Straus, St. Michael’s Hospital
Diane van der Woude, Leiden University Medical Center
Michael Ward, National Institutes of Health

Vivli launched its first data analysis challenge shortly after our launch in an effort to generate interest and move data sharing forward. We are delighted to announce the first Vivli Global Data Sharing Innovator and our Vivli Pioneer Winners.

John Frew, Rockefeller University, is the first Global Data Sharing Innovator for his publication in the area of Dermatology. Dr. Frew will receive a travel award to present at the Vivli Annual Meeting on Nov. 18 in Amsterdam.

“Our congratulations to Dr. Frew,” said Rebecca Li, Vivli Executive Director. “We look forward to being able to name more Vivli Global Data Sharing Innovators in the near future.”

The Vivli Pioneer awards, granted to the first ten lead investigators whose data requests are successfully approved and have initiated their analysis, have also been named. They are eligible to receive $1,000. We are grateful to the Doris Duke Charitable Foundation for their initial grant in supporting these awards.

These winners are:

John Frew, Rockefeller University
Vojtech Huser, National Library of Medicine/NIH
Akira Kimata, University of Tsukuba
Frederikus Klok, Leiden University Medical Center
Vivek Rudrapatna, UCSF
Changyu Shen, Beth Israel Deaconess Medical Center
Mirjana Stanic Benic, University Hospital Centre Rijeka
Sharon Straus, St. Michael’s Hospital
Diane van der Woude, Leiden University Medical Center
Michael Ward, National Institutes of Health

View of summary of the research requests to learn more about these research proposals.

Few things are more frustrating than wasted time. The Vivli team wants to help you make the most of your valuable time, and to help make sure that your data request has the best possible chance for success. While access to data is a decision made by our data contributors (see our Members Page for more details), we do have some general tips and tricks for drafting your data request.
1. Use our resources. Vivli has several resources to help data requestors, including a Quick Start for Requesting Studies and a Data Request Form Worksheet that you can fill out offline, and share with other members of your research team. Both resources are full of specific information about the fields of the data request form that may require greater attention.
2. Ask for help. Vivli has dedicated staff ready and waiting to answer any questions you may have about the Vivli platform. You can contact our team via email to support@Vivli.org.
3. Watch a webinar. Review our webinar on Keys to Submitting a Quality Research Proposal to a Data Sharing Platform, with representatives from our platform as well as the YODA Project; the Wellcome Trust IRP; and AbbVie. Vivli co-founder Dr. Ida Sim has also created a webinar to explain how to share and request data on Vivli; and Dr. Sarah Nevitt of the University of Liverpool provided Vivli users with an in-depth presentation on how to perform IPD meta-analysis. Check our library for additional webinar content.
4. Tell us about you & your team. When filling out the Lead Researcher and Statistician Researcher fields on the Data Request Form, make sure you list the specific training, qualifications, education or experience that qualifies you and your team to do the analysis described in your proposal.
5. Watch for abbreviations. We frequently see data requests full of abbreviations and acronyms. These time-savers can be helpful, but make sure you spell out the entire phrase the first time it appears in your data request so that it can be more easily understood when it is published on our website.
6. Be specific. Data contributors generally want to know how your project relates to the data you’ve requested. Make sure you tie the studies you’re requesting to a specific scientific question or hypothesis.
7. Check your dates. On the data request form, you must enter an anticipated start and completion date for your project. This refers to the timeline of your work on the Vivli platform only, so make sure the dates you enter reflect that.
8. Give it a title. After your data request is approved and you get access to the data, certain parts of your data request will be published on the Vivli website. This includes the title, so make sure you give your project a title that briefly describes your project.
9. Check it twice. While your data request will not be evaluated for its spelling or grammar, it is important to be sure that anyone who may be reviewing it can clearly understand the information. You can use our downloadable Data Request Form to draft your proposal and then copy the information into the platform when filling out the data request.
10.  Revisions requested? Don’t sweat it. Even if you follow all the tips above, your data request may still be sent back for revisions. These may come from the Vivli team, or from Data Contributors or other reviewers asking for more information or clarifications. Revising your request is an easy, straightforward process and the Vivli team is available to help every step of the way.

Long-term cancer survivor and patient advocate Jane Perlmutter tells us why data sharing is so important for people with rare diseases.

Q: Why should we share data from clinical trials? 

Jane Perlmutter: I spent half my life as a cancer survivor. At first, I thought I wouldn’t survive it–and I’ve had three more cancer diagnoses since. I’m still here. And I’m here because people have contributed their own experience by being in clinical trials.But we need to learn faster. I like to say patients don’t have the luxury of patience. By allowing researchers to get access to data from multiple clinical trials, we can get to better solutions more rapidly and help more patients.

Q: What does data sharing mean for people with rare disease?

JP: When patients are diagnosed with a disease, especially a rare disease, it’s often very traumatic, and they want to find some way to get meaning out of that traumatic experience. One way is to participate in a clinical trial. They hope that their experience helps answer some questions and lead to some solutions so that their children or future generations won’t have to suffer the same things as they did. Sharing data from these trials increases the chances that their experience will be meaningful. There are fewer trials for very rare diseases, and that’s one reason why it’s even more important that we share those data and make the most of them.

Q: Aren’t you concerned about privacy with data sharing? 

JP: To make people feel secure in sharing their data, we need to ensure there is some privacy so that their personal information doesn’t get out. When it comes to rare diseases, there are fewer of those patients, so protecting their data becomes even more of a challenge. The good news is, there are many ways to protect patients’ identities. And, while privacy is important, what’s most important for the patients–and they will tell you this–is that many researchers can use the data. A recent study showed that 93% of patients who are in clinical trials are happy to have their data shared with more and more researchers.

Q: How can platforms like Vivli help?

P: It can be difficult for a researcher that runs a clinical trial to figure out how to share data, but Vivli makes it easy. To learn as rapidly as possible, and get to better treatments and cures faster, researchers need as much data as they can get their hands on. That’s what Vivli is all about.

By Rebecca Li, Vivli Executive Director

I started my career at a small biotech company and understand the intensity of the start-up experience as well as the experience of the more mature mid-size biotech and pharma company. Data sharing can be an after-thought to the challenges of bringing treatments to the clinic. We at Vivli often receive inquiries along the lines of: “When should we start implementing a ‘real’ data sharing program?”  The inquiries run the gamut from companies just beginning their data sharing journey to those that have done this proficiently.  We know many who have started to outgrow homegrown sharing systems and processes that functioned well initially, and now the rate of incoming enquiries may be more than those systems can handle. Companies beginning the data sharing journey may be clinically early stage or have a marketed drug.

What’s it take to get serious about data sharing?  Some questions and answers may help.

Why should we share?

There are numerous ethical reasons to share clinical trial data responsibly.  Various industry organizations have set public commitments for their members as principles. For example, BIO a trade association that represents thousands of small and mid-size biotechnology members, committed in 2014 to make transparent and facilitate the process for qualified data requests of individual participant level data for approved medicines [1]. Likewise PhRMA/EFPIA[2] and IFPMA[3], influential trade organizations in the US and internationally for the industry, have committed to data sharing for the pharmaceutical industry

When should we begin the program?

If a company has one or more programs in late-stage clinical development, then 18 months prior to either marketing authorization or regulatory decision is a realistic timeframe to seriously considering putting a data sharing program in place. Like most things in life, starting the planning earlier allows for a high-quality program to be implemented with greater degrees of freedom. Therefore, if you anticipate that there will be trial data to share within the next 18 months or less it is important to start the planning process for a data sharing program now.

How do we start to build a data sharing program?

A data sharing program consists of three primary components – a Policy, a Mechanism and the Resources to manage and oversee the program – we describe these components in brief below. A Policy for data sharing typically includes a written public statement for governance (covering areas such as which trials are available for sharing, when those data are available, who is qualified to make a request and how, and whether there are exceptions to these policies). Additionally, a Data Use Agreement – the legally binding agreement which governs data access to outside researchers – should be developed to protect all parties involved including the research participants. A Mechanism for data sharing should be described – this is the IT infrastructure, portal or platform management system that has been developed to manage the data sharing program. Lastly, sufficient Resources (human and capital) should be tapped to oversee and implement the program and scalable to meet demand once Policy and Mechanism are set in place.

If you are interested in learning more you can watch the webinar.

How can we manage a data sharing program?

On your own internally or through a trusted external partner. Similar to other partnering decisions such as the tactical use of a CRO, externally managed mechanisms for data sharing exist.  If there are available resources internally that are capable of initiating this effort – they should also be available to oversee and sustain this effort over time. Internal efforts, may be typically comprised of a multi-functional team drawn from biostatistics, clinical data disclosure, biometrics and at times members from other functions such as medical writing, program management, legal, IT, and clinical operations are called upon for their expertise.

What can partners like Vivli do for us?

VIVLI was established several years ago and offers our members a comprehensive data sharing program including Policies, a Mechanism for data sharing and Resources. This includes assistance with setting up policies and access to harmonized legal agreements that are already vetted with a broad range of stakeholders. Additionally, the Vivli team can advise your leadership if necessary on public-facing data sharing policies for public or private organizations and will discuss a solution tailored to your current situation. The Vivli platform is user-friendly and flexible, and provides a team – an extension of yours – to ensure seamless data sharing.

If you are interested in learning more you can watch the webinar.

[1] https://www.bio.org/articles/bio-principles-clinical-trial-data-sharing

[2] https://www.phrma.org/press-release/joint-efpia-phrma-principles-for-responsible-clinical-trial-data-sharing-become-effective-today

[3] https://www.ifpma.org/wp-content/uploads/2010/11/IFPMA-Principles_Data-Sharing-FINAL-w-QA-vF.pdf

 In June of 2017, the International Committee of Medical Journal Editors, or ICMJE, released new requirements for data sharing statements, for submissions to their publications. The ICMJE represents the most well-respected and influential publications in the biosciences, including the New England Journal of Medicine; the Lancet; Annals of Internal Medicine; the BMJ; and PLOS Medicine: These publications, and the many others which follow their lead, now require the inclusion of a data sharing statement in all submissions for publication. ICMJE publication editors may take these statements into consideration, when making editorial decisions. Researchers should also know that as of 1 January 2019, ICMJE requires inclusion of your data sharing statement at the time of trial registration.  

In their guidance to researchers, the ICMJE notes that when it comes to data sharing, “undecided is not an acceptable answer.”   When pre-registering interventional trials on clinicaltrials.gov, if “Undecided” is selected for the “Plan to Share IPD” data element, a note now appears indicating that the ICMJE data sharing policy requires a “Yes” or “No” answer to this question. The “Plan to Share IPD” data element is optional on CT.gov, but is required by the ICMJE as part of registration information for interventional studies  Rather than remaining “undecided,” see the chart below to learn how to use Vivli in completing your ICMJE data sharing statement: 

ICMJE Question	How to respond, if using Vivli to share your data:
Will IPD (and data dictionaries) be made available?	Yes
What data in particular will be shared?	Final cleaned individual participant-level data, de-identified*
What other documents will be made available?	Final protocol, statistical analysis plan, and the data dictionary. (Note: additional documents such as CRFs and analytic code may also be included.)
When will data be made available?	X months /years after study completion
With whom?	Anyone with the relevant skillsets to conduct the analysis,  who submitted an approved proposal on Vivli. Proposals are submitted on Vivli.
For what types of analysis?	To achieve the aims and objectives in the scientific proposal as approved via Vivli.
By what mechanism will data be made available?	Following an approved request, a data use agreement must be signed.   Data are made available via a secure research environment or download.

 

For more information or assistance in using Vivli to meet data sharing requirements, email suppport@vivli.org.  

* Vivli partners with Privacy Analytics, to provide de-identification of data sets at a discounted rate for researchers who store and share their data on Vivli.