Meeting Minutes - October 30, 2014
Shelley Hurwitz, Sally Thurston, Frank Harrell, Tracey Weissgerber, Bob Oster, Leah Welty, Mat Soukup, Liz Heitman, Susan Perkins, Susan Duke, Simona Carini, Brenda Crowe, Dale Plummer, Andy Cucchiera
Introduction: CTSpedia - BERD, Research Ethics, and CTSA projects
Frank Harrell: We had hoped to have funding for CTSpedia from NCATS, but that funding did not come through. Frank was wondering if we can get some BERD conversations re-started.
Leah Welty: A group is restarting with Susan Perkins and the ACTS. There is no formal invitation right now. The group will be a special interest group (SIG) within ACTS. They will have conference calls and will be broadening the invitation to all of the original BERD members.
Liz Heitman: The Research Ethics group has one PI in Communications, who is interested in continuing the CTSA
work. They have quarterly meetings. They are not part of ACTS.
CTSpedia: HOW IS IT BEING USED AND WHAT IS THE LEVEL OF SUPPORT
Safety Graphics: Susan Duke, Mat Soukup, Brenda Crowe, Mary Banach
Susan Duke reported on the FDA/Industry/Academia Safety Graphics work, which involved statisticians from the FDA, industry, and academia. The goal of working group was to answer questions about common safety design issues in clinical trials. It was an industry-wide collaboration. They needed a common website to post their work and CTSpedia has turned out to be the perfect venue. Just recently 2 publications have featured links and work from CTSpedia Safety Graphics. The first is a book chapter on safety graphics for pharmaceutical clinical trials and the second is a paper on Tuffte-type graphs in medical research.
Mat Soukup reported on the history of Safety Graphics Working Group and CTSpedia. They were searching for a website in the public domain that was curated by statisticians in clinical research. The site needed to be more than a place that just had publications but had a structure where users and working group members could interact. They felt that CTSpedia and the Foswiki, administered by Vanderbilt, was the most advanced wiki that they had used and provided the venue for the working group to display their suggestions, including code and sample graphics, as well as best practices.
Safety Graphics: Safety Graphics Presentations at national and international meetings, Code, Example Graphics, Best Practices, and an area for contributions and questions from CTSpedia users.
CTSpedia Statistical Tools: Sally Thurston
Sally Thurston at Rochester has been using CTSpedia since its inception. She uses the references, important papers, and BERD presentations in her work. The Reproducible Research links and papers have been an invaluable resource for her. She continues to do BERD education projects for clinical research and uses the papers, tools, education materials for her projects. One example that Sally has found very useful is Peter Bacchetti’s post on p-values. CTSpedia is very relevant resource to her to this day. Sally was just speaking with someone at Rochester who does a great deal of consultation. She said that she often uses the SAS macros that others from Rochester uploaded to CTSpedia.
Educational Tools Template Testing: Tracey Weissgerber
Tracey Weissgerber is a relatively new contributor to CTSpedia. Felicity Enders at Mayo Clinic showed Tracey the educational tools area of CTSpedia. Tracey has presented a brief overview of her work.
The Problem: The inability to reproduce key scientific results is a growing concern among scientists, funding agencies, journals and the public. The problem of irreproducibility was recently highlighted in a series of articles from the NIH and Nature series journals. We propose that the types of figures that scientists use to present continuous data may contribute to irreproducibility. Bar graphs are designed for categorical variables; yet they are commonly used to present continuous data in laboratory research, animal studies and in human studies with small sample sizes. This is problematic, as many different data distributions can lead to the same bar graph. The full data may suggest different conclusions from the summary statistics. We systematically reviewed all full-length, original research articles published in the top 25% of physiology journals between January 1 and March 31, 2014 (n = 703) to assess the types of figures that were used to present continuous outcomes. Most continuous data were presented in bar or line graphs. Figures that showed the data distribution were uncommon (univariate scatterplots, boxplots and histograms). This was particularly concerning given that most studies had small samples.
Using CTSpedia to Promote Better Data Presentation Practices: Our results suggest that as scientists, we urgently need to change our approach to presenting and analyzing continuous data in small studies. To facilitate this, we created free Excel templates that allow investigators to quickly make univariate scatterplots for small studies). CTSpedia provided an excellent opportunity to make these free templates available to researchers around the world. I presented posters detailing our findings and handed out cards listing the website at two meetings (International Society for the Study of Hypertension in Pregnancy, BIRCWH Scholars Annual Meeting).The manuscript reporting our findings is currently in review. The CTSpedia link for the templates will appear in the paper and will also be available through websites for the CCaTs at the Mayo Clinic and the University of Belgrade Medical School. We are exploring opportunities in create an online Continuous Professional Development seminar through CCaTs at the Mayo Clinic and to incorporate this into the FunCATS curriculum. Our statistical collaborator at the University of Belgrade will also incorporate the templates into their biostatistics curriculum (3,000 students/year).
CTSpedia Links and Resources: Susan Perkins
Susan Perkins created an area on CTSpedia to report on collaboration and consultation good practices. They have three sections currently: 1) consultation, 2) links for walk-in clinics, and 3) clearly explained guidelines. A new paper on the paper on the consultation paper has just been accepted. The CTSpedia site is set-up for others in the community to provide additional links and resources. Susan has found CTSpedia to be exactly the solution that was needed to communicate their work.
Educational Materials: Bob Oster and Andy Cucchiera
The Educational Tools group earlier started going through the materials on CTSpedia and updating them. Their goal is to get more peer-review of the materials and to post new materials as they receive them. Bob’s group is anxious to get re-started with CTSpedia activities.
Andy has been working with REDCap. He would like to establish a working group between CTSpedia and REDCap to design good clinical trial forms.
Ethics in Biostatistics: Shelley Hurwitz
Jon Gelfond has taken over as the lead in this group. Shelley reported that their session at JSM was packed. CTSpedia has afforded this group an opportunity to post meeting abstracts, slides, and updates as well as share educational tools and controversies for biostatistician. Shelley has posted the latest information on Ethics in Biostatistics presentations.
Research Ethics: Liz Heitman
Research Ethics area on CTSpedia was started years ago by a non-educator at UC Davis. There are a number of people who are interested in both curating the materials that are currently on the site and expanding it. This group would not really be candidates for ACTS.
Study Design Typology: Simona Carini and Ida Sim
Simona was on the call but she was driving and could not talk to us. She and Ida were early contributors to CTSpedia for the Study Design Typology work. They went through numerous iterations and papers are now published on their CTSA
work. They could use CTSpedia for a forum to interact with clinical researchers who would like to use the Study Design Typology tool. Simona has provided us the following description of their work:
The CTSA Human Studies Database Project (hsdbwiki.org) developed a Study Design Typology that classifies quantitative research on individual humans into one of eight high-level study design types, further defined by a set of Additional Factors http://hsdbwiki.org/index.php/Study_design_typology
An early version of the typology is already in CTSpedia for use as both a teaching and a classification tool for clinical researchers and statisticians.
We would like to update the Study Design Typology to its current version and also create a way to collect usage data and feedback from users as evaluation of our work. Renewed funding for CTSpedia would allow us to expand the user base of a unique tool already judged as extremely helpful by BERD members.
Ida was the PI for the project.
Google Analytics: Dale Plummer
CTSpedia continues to be used. There were approximately 11,000 hits last month. There was a slight up tic usage last month. It continues to be used by the community. Our biggest problem in using the wiki is helping users to add materials to the site. With funding Dale and Mary can look at better, more efficient ways to increase dialogue and add more materials to the site.
Entire CTSpedia: Peter Bacchetti
Peter Bacchetti could not make it to our call, but we talked before the call in order to capture how Peter has been using CTSpedia and his suggestions for the future of CTSpedia. CTSpedia owes its existence and continuing use within the CTSA
and biostatistics communities to Peter, Sally, Laurel, and Frank.
Peter created these materials
for the biostatistics consulting service at UCSF. As you can see he gave links as well as explanations and descriptions of the materials. He believes that consultants need a site that provides: 1) information on what is useful for consultants, and 2) concepts that need to explained for collaborations in clinical research.
The more CTSpedia can address glaring needs in our community, the more helpful it will be. Peter has always felt very strongly about curating and producing high quality content on CTSpedia. Peter reiterated the need to really curate a revitalized CTSpedia.
Peter pointed out one of our most successful collaborations on CTSpedia came after Frank presented a critique on ROC curves at one of our BERD face-to-face meetings. A very fruitful discussion continued on CTSpedia with a real examination of the pros and cons of ROC curves. CTSpedia offers an important resource to continue these discussions and this work.
Do we really need CTSpedia? Can ACTS fill the void?
Frank wanted to know if we really needed CTSpedia. Susan Perkins pointed out that she can post files and we can have discussions on the ACTS website but it is limited. Leah pointed out that the ACTS site is closer to a very simple Rocket tool for internal communications and does not have the exposure provided by CTSpedia. Susan and Leah are talking with Brad and Chris. They will be discussing how best to revitalize the BERD conversations and will be in touch with us
Archive CTSpedia Materials
Return to CTSpedia Web Home
ONLINE JOURNAL CLUB
Success Stories/ BERD Watch
Collaborations: Adaptive Clinical Trial Design
CTSpedia Mission Statement
Return to CTSpedia Web Home
This topic: CTSpedia > WebHome >
Topic revision: 14 Sep 2015, MaryBanach
Copyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback