Reflections on DFUG 2018

Date: October 9, 2018
Contributed by: Eric Bosch, Director of Software Development at DF/Net

Hi, I’m Eric Bosch, the Director of Software Development at DF/Net Software. This year I took part in my 18th DFUG meeting. Every year I leave the meeting feeling energized about ideas that our users bring and the future of our software. DFUG 2018 was no exception. Here are my thoughts about our 24th Annual DFUG Meeting.

DFUG 2018 took place September 27 and 28 at the Hyatt Regency Lake Washington, located at the southern tip of Lake Washington and directly adjacent to the Boeing factory in Renton. The DF/Net team went to DFUG with a wide range of new software tools that we wanted to share with clients.

Day One

The DF/Net Software team agreed on the idea of telling a story, using our tools to simulate clinical trials data collection in imagined settings.

Our role-playing counterparts, Anna, Connor, Simon and Evan (me), told this story during the morning. Study definition was amended by Simon using DFsetup, Evan performed remote data collection online and offline using DFcollect, Connor entered lab values using DFexplore, and Anna used DFweb in an "application restricted" setting to review and query Evan and Connor’s data. These tools and people collaborated to get clean data into the study database efficiently and accurately, from disparate sources, using tools that interoperate by design.

Between presentations, we kept attendees sharp by asking Jeopardy-style questions about our software, our host city of Seattle, and the home country of the development team, Canada. Through these questions, we were reminded that DF_SSschema does sound like it could be the name of a boat, residents of Seattle buy more sunglasses per capita than residents of any other US city, and Canada produces 80% of the world’s supply of maple syrup. Attendees were very creative in their question-style answers, generating lots of laughter and eventually, the right answer - er, question.

Our attendees also had stories to share.

Lisa and Darryl opened the meeting with the lighthearted reasoning behind the theme of “Take Flight”. In keeping with this theme, Darryl provided glimpses into future areas of focus for our software. Beyond DFcollect and DFweb, we will continue to explore new uses of our API to connect people with their data, provide seamless CDISC integration, and create new data views optimized to meet the needs of monitors and managers. The audience that needs data to make decisions is expanding; we need to make sure DFdiscover is the “go-to” tool for this growing group. Lisa shared that it’s not just about our software, it’s what our users do with our software that is so motivating. Lisa provided many examples of the visible and positive impact our software has not only on global research, but also on projects benefiting local communities. This was the perfect introduction to the next talk by Lara.

Lara from University of Washington engaged us with her story about offline data collection in townships of Kenya and South Africa.  Lara identified the needs of the study that eventually led to the genesis of DFcollect and the inspiration for our latest version of DFcollect. Lara had lots and lots of questions - attendees are clearly keen for the DFcollect solution.

Gavin shared with us his extensive experience intertwining DataFax and clinical trials data collection. His story of a “rolex” that could not stop data collection on a paper CRF, gave everyone a great chuckle and reminded all of us that even in the face of adversity, there is durability to a paper CRF that is not easy to surpass.

Mark, from the University of Calgary, presented a follow-up to his DFUG 2014 presentation and updated attendees with progress on the Study Manager project. Five new deployments have occurred since then, each adding new features and functionality. Mark and the ‘B153 team’ are actively soliciting input for a product development feedback pool.

Samer further explored the landscape of possibilities now that the DFws API is available. Lots of great examples showed that there are many possible programming uses for the API, limited mostly by our imaginations. There was a lot of DFws API discussion after the presentation and throughout the remainder of the meeting.

Sadeesh, from our newest client, the University of Manitoba, summarized his journey from DataFax experience at McMaster University to a brand new DFdiscover installation. He reminded all of us to focus on tasks, not technology, and that a flexible solution will open the door to possibilities that may otherwise be blocked by inflexible thinking. Welcome aboard, Sadeesh and team.

At the end of Day One presentations, everyone participated in the Marshmallow Spaghetti Tower challenge. Twenty teams had 15 minutes to build the tallest structure that they could, using only spaghetti and marshmallows. It was reassuring to know that we are all involved in clinical trials and data management, because we certainly are not engineers. After removing the goo from our hands, we boarded the Lady Mary for an evening dinner cruise on Lake Washington. Everyone enjoyed food, drinks and stimulating conversation.

Day Two

The start of a new DFUG day. On Day Two:

I was fortunate to speak first to a meeting room full of refreshed minds. DF_QCupdate is a huge advantage in DFdiscover (and of course DataFax). Our software has always known how to deal with visit scheduling, data that is missed and visits that are overdue. Rather than resting on our laurels, we are continuing to improve the usability and feature set of this important monitoring tool.

Beth from ShiftLabs spoke to us about DripAssist. Beth and her team are focused on solving the problem of getting accurate gravity infusion data. Can their novel device, together with the DFws API, put the entire picture of gravity infusion within easy reach of the study team?

Jaskiran, at the Office of Cyberinfrastructure and Computational Biology, NIAID, NIH, shared their workflow and structure for a clinical data warehouse. This warehouse structure greatly simplifies the task of acquiring data from CRFs created with CDASH in mind, pushing it through DataFax/Dfdiscover's extensive data cleaning / validation features, linking to other research data, and pushing out SDTM data sets.

We learned from John at iRespond that over one billion global citizens do not have access to a form of identification that allows them to get a job, a license, healthcare, or travel across borders. With the iRespond biometric identity solution, individuals are uniquely identified and verified so that they are who they are, and no one else. Individuals now have a secure digital solution as an alternative to paper-based forms. John also spoke about using iRespond’s technology to prevent co-enrollment.

Fax images with poor quality barcodes can delay data entry. Nick shared with us the investigative, and now standardized, approach that his team at NETE/OCICB-NIH has developed to ensure that every page of every fax document lands in the correct, final study database. As a DataFax client with 100+ active studies and over 3 million CRFs received and processed in the first 9 months of 2018, this is a vital solution.

John works at the Population Health Research Institute, a client that has been using DataFax since the early 1990s. John shared his work with DataFax and Excel to create a toolkit that presents a dashboard of key performance indicators in an active, and very busy, DataFax installation. Staying on top of workflow statistics, disk space consumption, etc. is greatly simplified with always available metrics and visualizations.

DF/Net's Sandi has been exploring CDASH domains and DataFax modules, tying together the pieces to simplify SDTM output. Sandi shared several great insights and suggestions for using DFexport to get a long way toward a workflow that delivers SDTM output.

I had one more chance to lead discussion with attendees and present the development team’s thoughts about meaningful simplification of our code. We are motivated to continue to improve the software’s offerings. And sometimes that means that we need to simplify too. Thanks to everyone for the great discussion and feedback.

Darryl and Lisa closed Day Two, and the meeting, with a review of our journey together. They spoke of new challenges ahead and how DF/Net is uniquely positioned to address these challenges. Everyone is excited for the coming year.

Feedback

Back at the office on Monday, we put together a slideshow of photos from the meeting and we posted PDFs of all the presentations. A quick notification email to everyone that the materials were posted and we were ready … to start thinking about DFUG 2019.

Since DFUG 2018 we’ve received lots of email feedback from attendees including:

  • You hit a Grand Slam.
  • I really thought the ‘role-playing’ presentation ... got across the idea of what all the products do, how they integrate with each other, & also ‘when’ to use each product.
  • Great job on DFUG 2018 !!