Can you believe another year has almost come to a close? It’s outstanding what the PHUSE projects have achieved over the past 12 months. We’ve launched 6 new projects, and produced 19 deliverables ranging from white papers, abstracts and workshops to new software tools like the R Shiny App for simplified TS Domain.

We continue to build upon our collaborations, with the extension of our MOU with CDISC, the launch of a new cross-collaboration project focussed on creating an R Shiny Forrest Plot Application with the American Statistical Association (ASA) and the FDA, and we welcomed new industry members and regulatory liaisons from CDER, CBER and OBS to our Steering Committee.

Our published deliverables have undergone a makeover and have been given a new professional look. All our deliverables can be found on the PHUSE website. Under Useful Information, you will find PHUSE templates such as the Working Groups PowerPoint slide deck and white papers, how to submit a New Project Request and processes, and guidance for our Working Groups members.

Early in the new year, our focus will be on the planning of the Computational Science Symposium (CSS) event. This will take place in Washington 7–9th June and will be the 9th co-sponsored event with the FDA. The projects look to address the challenges we face as an industry and formalise best practices, meet like-minded folks like yourselves, generate new ideas and seek opportunities where we can cross-collaborate. We host a well-attended Poster session together with very interesting and successful workshops. Registration and Call for Posters will open in the new year.

This quarter’s Working Groups news will highlight our successes and what’s on the horizon for 2020. If there is a project you would like to get involved with, email workinggroups@phuse.eu.

Data Transparency:

Data Transparency Workshop at the CSS with regulators

In June 2019, the Working Group met in Silver Spring and had the opportunity to discuss data transparency with representatives from the FDA and Health Canada, who both gave an update from their respective agencies. The workshop also touched upon the Best Practices Guide, GDPR and Data Utility projects, with 20 participants. A representative from the EMA Policy 0070 team also presented a status from the EMA and participated in a workshop at the EU Connect 2019 in Amsterdam.

Data Anonymisation and Risk Assessment Automation Project

The project team, led by Lukasz Kniola, has now finalised the first draft of its deliverable, which provides a practical and step-by-step approach to data anonymization, building up from the PHUSE Data De-identification Standard, and includes concepts from key publications in the field. The team plans to send this out for public review in Q1 2019.

Answer to call for review of regulatory documents

Two sub-teams from the Working Group joined forces to review a Draft Order from the Ukrainian health authorities around the sharing of “synopses-like” documents in May 2019 and answers to call for comments on the FDA Clinical Data Summary Pilot Program in August 2019. The comments from the PHUSE sub-team were also featured in a Pink Sheet article.

Data Utility Project

A sub-team was formed around this topic and repurposed the project to map the different types of research requests against the different data sharing initiatives and policies to further inform and enable secondary-purpose research.

• Asset transfer(two white papers) from TransCelerate to PHUSE in June: Protection of Personal Data in Clinical Documents - A Model Approach:Version 1.0, 10-Jun-2019 De-identification and Anonymization of Individual Patient Data in Clinical Studies - A Model Approach: Version 1.0, 10-Jun-2019.

• Data Transparency – A Global View of the Clinical Transparency Landscape: Best Practices Guide: This is a sister document to the similarly named white paper that was published in Dec 2018. This Best Practices Guide brings together subject-matter experts who share their recommendations with the aim of fostering a holistic approach to clinical data transparency and disclosure that provides accurate and complete information across multiple disclosure platforms. This document is currently undergoing finalisation after both a recent Steering Committee and public review. In 2020, we plan to explore additional methods to publish the information as the guide is very extensive.

• Recommendation for GDPR Compliance: A reflection paper was developed offering practical recommendations for applying GDPR rules. This paper was the result of two online surveys where a broad range of opinions was sought and is currently being finalised after also receiving comments from PHUSE, the Steering Committee and from a public review cycle.

Educating for the Future:

• Data Engineering: What a great year for the Data Engineering project! With 10+ new members in 2019, we divided into four sub-groups to focus our efforts – Data Sources and Use Cases, Automation, Data Engineering Techniques and the Squarespace web portal – starting the year with an overview of what data engineering means to our industry at both the US Connect and the CSS. Then, taking some time to curate information regarding Data Sources and Use Cases, Automation, and Data Engineering Techniques, our team presented two papers at the EU Connect (“Processing Big Data from Multiple Sources” and “Educating for the Future: Data Engineering Techniques”). In 2020, we will begin our Big Blog Bonanza to draw more traffic to our web portal while also building up the content. Look for us at the US Connect by attending our presentation on processing data from multiple sources and by stopping by our poster presentation for Data Engineering Techniques. We are also looking forward to sharing our learnings on Automation in a white paper, while also looking for new future ways to educate our industry, perhaps through webinars or video blogs. Stay tuned!

• Design Thinking: Design thinking methods foster a way of thinking that reframes the problems and solutions we assume we have the answer to. With that in mind, this project facilitated a session at the PHUSE US Connect in Baltimore earlier this year and ran a two-day workshop at the CSS. The project is currently planning for the Design Thinking Workshop that will form part of the Leadership session at the US Connect in March.

• Data Science: This project was formed early January and will look to help support people to familiarise themselves with data science and all its various branches. It will provide a training repository, where consolidated materials around topics like Machine Learning, Artificial Intelligence, Deep Learning and Quantitative and Advanced Analytics will be made publicly available. Currently the team are working on their paper for the US Connect next March, which will highlight the differences between data science and a data scientist.

Emerging Trends & Technologies:

• The group has continued to be very active in all project areas, and with the addition of Real World Evidence and Open Source Technologies the group now has 9 active projects.

• The Cloud Adoption in the Life Sciences Industry project released Framework Version 4 in May, a comprehensive guide to using the cloud for clinical research, and is now focussing on developing a framework to address the challenges the cloud brings.

• The Going Translational with Linked Data project held the ever-popular ‘Let’s Make a Knowledge Graph!’ Workshop at both the CSS and EU Connect this year. The project has recently re-launched its efforts, with a focus on data validation and potential new collaborations. Watch this space!

• The KPI metrics initiative (PHUSE Key Performance Indicators & Metrics Project) revised and reissued the common data reporting metrics in response to the Year 1 survey. The Year 2 survey is now underway and will be reported in the new year.

• The FHIR, Blockchain and Machine Learning projects continue to progress solidly, having all contributed throughout the year to PHUSE conferences and SDE events worldwide.

• The Data Visualization Working Group initiated two technical proof of concept projects to explore how to share interactive tools with the FDA in 2019. The goal is to assess the feasibility and benefits of leveraging interactive tools as a reviewing aid to complement the standard eCTD submission. We are happy to report that we have successfully completed the first pilot, which included an interactive data visualisation (a volcano plot written in JavaScript) as part of the submission, and are currently conducting the second pilot to explore the containerisation approach to submit an open-source R Shiny interactive application to the FDA. Several FDA staff will participate in this pilot and share their experiences afterwards. Assuming all goes well, Genentech will be interested in conducting a real submission for the R Shiny application, using the docker containerisation process in 2020.

Nonclinical Topics:

Collaboration was the word of the year for the Nonclinical Topics Working Group. Nearly every project fostered interactions outside the Working Group and across our PHUSE landscape. The nSDRG team established a working relationship with the Japanese CDISC User Group around exchanges of questions and issues related to effective fulfilment of the nSDRG. All participants are motivated and enthusiastic about this new relationship, which started at the CSS this year.

The Modeling Endpoints project achieved a white paper, expected shortly, through extensive collaboration with the CDISC workstream evolving the Immunogenicity Specimens Domain. By working together with the CDISC team and ADA experts from nonclinical laboratories, the project developed an interim recommendation for modeling nonclinical ADA data in SEND 3.0 and 3.1 that supports e-data submission of this important data in anticipation and consideration of forthcoming CDISC SEND team efforts.

The Nonclinical Scripts project team made a contribution, in collaboration with the Standard Analyses & Code Sharing Working Group, to further develop automated creation of simplified ts.xpt for clinical and nonclinical studies. It has provided a tangible tool for anyone in industry to take advantage of an efficient tool which helps sponsors successfully sail through one of the Technical Rejection Criteria at the time of submission. And there is more to come for 2020!

The next Industry SEND survey will be out early in Q1, to prepare for results presentation at the next CSS. In past years, the “SEND Survey” has provided important insights into the status of implementation across industry data providers. The newest project, Best Practices for QC of SEND Datasets, is already running fast. This project has co-leaders from industry, the FDA and a motivated team from around the world working towards establishing best practices to assure quality and integrity of e-data submitted to authorities. The team has already started surveying the industry for current practice and we look forward to their initial findings in early 2020. We thank the members of the Nonclinical Topics Working Group, who worked diligently throughout 2019 to collaborate, collaborate, collaborate. They bring to life many ideas for harnessing data challenges faced by us all.

IMAGE

Optimizing the Use of Data Standards:

This Working Group has had another successful year. Thank you to the teams for your contributions and volunteer hours, which has resulted in global impact. We said goodbye to Lisa Brooks as Co-Lead and welcomed Janet Low to co-lead with Jane Lozano. We'd like to thank Lisa for her contributions to the team. Here are some highlights from 2019:

Clinical Integrated Study Data & Analysis Data Reviewer's Guide is a NEW project that recently launched. The cSDRG and ADRG apply to a single study. Sponsors have used these templates when creating a reviewer’s guide for integrated study data and a reviewer’s guide for integrated analysis data. The template has to be changed in order to document the information for multiple studies. There is an opportunity to create a template for integrated study data and integrated analyses data. The project team will determine if this can be accomplished in one reviewer’s guide or if there should be two separate reviewer’s guides. Deliverables include a template, completion guidelines and example documents.

Publication of Regulatory Referenced Deliverables:

• The Legacy Data Conversion Plan & Report was incorporated into the cSDRG and ADRG, along with other updates, to align with the Technical Conformance Guide.
• Study Data Standardization Plan Template, completion guidelines and example documents.

Publication of White Papers & References:

• Best Practices for Documenting Dataset Metadata: Define-XML vs Reviewer's Guide
• Best Practices for Submissions of Event Adjudication Data
• Define-XML 2.0 Stylesheet Recommendations & Completion Guidelines
• DRG-XML Prototype updated package Version 2.0 – next steps: collaborate with the FDA to pilot the cSDRG in XML.

We continue to publish responses to our SDTM Implementation FAQs and these can be found on our Wiki page.

Q1 2020: We are looking to finalise and publish the following White Papers:

• Two papers from the Industry Experiences Submitting Standardized Study Data to Regulatory Authorities
• Integration Strategies in Support of ISS/ISE.

Look for us at the PHUSE US Connect 2020

• The Best Practices for Data Collection Instructions project will be doing a poster along with the CDISC Implementation Primer team.
• Industry Experiences Submitting Standardized Study Data to Regulatory Authorities will give a presentation and a poster.

Standard Analyses & Code Sharing:

This Working Group has had a very productive year, with six active projects continuing to publish deliverables for our industry members. Here is a highlight of their accomplishments:

• A successful Workshop on Safety Analytics took place at our CSS back in June. The Workshop recordings together with their slide decks can be found here.

• Publication of an R Shiny app to create simplified TS Domain Simplified TS Files provide a Study Start Date for NDA/BLA/ANDA studies that began on or before December 17, 2016 and Commercial IND studies that began on or before December 17, 2017. This allows the FDA to determine that study data is not required to be in a CDISC-standardised format. Sponsors should submit a dataset named ‘ts.xpt’ with four variables (STUDYID, TSPARMCD, TSVAL, and TSVALNF) and one row of information. Additional details can be found in the Study Data Technical Conformance Guide.

• The Working Group created a landing page on Squarespace, which improves the Script Repository Front-End.

• The Script Metadata for Sharing White Paper was published and can be found here.

• The R PHUSE package version 0.2.0 has passed tests and loaded to the CRAN; this package sets the web framework for accessing the scripts stored in the GitHub Repository.

• In January 2020, we'll be looking to publish a White Paper on Best Practices for Quality Control & Validation and start our planning for the CSS.

We have many exciting things to look forward to in 2020 such as an upgrade to our Wiki platform, new and exciting projects on the horizon and, of course, the Computational Science Symposium.

We want to thank each and every one of our project members, whose time, enthusiasm and dedication to collaborating with industry members makes it all worthwhile. We are all working towards one common goal: a better, brighter healthcare...Cheers to that!

Both Wendy and Lauren have thoroughly enjoyed supporting you all. The camaraderie, team spirit and friendliness we experience every day is very warming. We hope you have an amazing festive holiday break with your loved ones and we look forward to welcoming in 2020 with you all.

Merry Christmas and a Happy New Year.