- Full participation across all stakeholders — clinical trial sponsors, sites, investigators, patients and their healthcare providers.
- Information is fully used to improve the overall quality, design and development process. We protect patients through shared, proactive safety science.
- Improving the execution of research and development through greater harmonization with the potential to conduct collaborative trials.
Newest Resources and Company Milestones
- BioCelerate’s SEND Harmonization Initiative published its second manuscript, co-authored in collaboration with the FDA as part of its Public-Private Partnership, in Chemical Research in Toxicology’s Computational Toxicology Special Edition. The article, “Leveraging the Value of CDISC SEND Datasets for Cross-Study Analysis: Incidence of Microscopic Findings in Control Animals,” highlights the potential of SEND datasets for cross-study analysis and proposes solutions to address the challenges of doing so currently.
- SCOPE 2021 finished on March 4 after three days filled with informative sessions related to the R&D and pharma industry. TransCelerate had the opportunity to present 10 sessions over the course of the conference covering many topics related to our solutions and initiatives. To read the in-depth recap of our sessions, click here.
- We welcomed Andrew Hudson (Roche), who is the new Chair for the TransCelerate Pharmacovigilance (PV) Steering Committee. He will be taking over for Paul Chang (Johnson & Johnson), who we would like to thank for his dedication as Chair over the last two years. The innovative and synergetic energy he has brought to the committee during his time within TransCelerate’s leadership has been appreciated.
Pulse On Progress Updates
Improve the Patient & Site Experience
Improve the patient and site experience by decreasing burden, enabling a better-informed patient and improving clinical research awareness, participation & engagement.
- Our Patient Experience Initiative has been diligently working to increase awareness and usability of critical deliverables such as the Patient Protocol Engagement Toolkit (P-PET) and Study Participant Feedback Questionnaire (SPFQ). In participating with volunteers across the Country Network, the team is pleased to announce that there are now Japanese translations of the P-PET and more than 15 different translations of the SPFQ available for download on the Translated Solutions Webpage. Languages include: Spanish, Russian, Czech, Danish, French, Hungarian, Flemish, Polish, Slovakian, Korean, Japanese, German, Italian, Dutch, and English.
- Our Site Qualification & Training Initiative had some exciting new changes to its Good Clinical Practice (GCP) Mutual Recognition website. As one of the most popular pages on our website, TransCelerate solicited input from the Training Provider community and determined there was a need for an update to address data quality issues, operational simplification, and outdated technology. This update will allow for automation of existing processes, improved user experience, and enhanced execution and oversight of the GCP Mutual Recognition Program.
Changes in this update include:
- An improved user interface to make it more intuitive to submit a new course
- New search function for course certificate templates
- Optical Character Recognition (OCR) technology to reduce manual errors
- There is a new URL
- All bookmarks will need to be changed to this new link
- The page will still be accessible from TransCelerate’s Site Qualification and Training Solutions page
- Integration with the Shared Investigator Platform (SIP)
Enhance Sponsor Efficiencies
Facilitate the advancement of innovative healthcare and clinical research through improved technologies, advanced data collection systems and simplified processes.
- Our Protocol Deviations Initiative published a manuscript, highlighting the contents of the Protocol Deviation toolkit entitled, “Protocol Deviations: A Holistic Approach from Defining to Reporting.” In addition to the paper, the team made updates to its Protocol Deviation Decision Tree and Protocol Deviation Process Guide.
Enhance Drug Safety
Facilitate solutions that address a common set of challenges in drug safety surveillance such as engaging the public, collaborating with regulators, incorporating data and analytics, developing model approaches, and assessing the impact of efforts.
- The Intelligent Automation Opportunities (IAO) in Pharmacovigilance Initiative published an article through Drug Safety entitled, “Validating Intelligent Automation Systems in Pharmacovigilance: Insights from Good Manufacturing Practices.” This position paper provides considerations to the industry on validation strategies that can be utilized for emerging AI-based automation technologies, while also outlining a risk-based approach to AI-static systems. It is intended to facilitate dialogue and collaboration with Health Authorities and industry on this topic and assist pharmacovigilance professionals by providing the knowledge to lead technology implementations within their organizations with considerations given to the building, implementation, validation, and maintenance of assistive technology systems.
- The Value in Safety Information Data Sources (VSIDS) Initiative released a new infographic entitled, “Identifying the Value of Information Data Sources to Improve Patient Safety.” This infographic overviews study findings from two papers on the value of information data sources to improve patient safety and can be widely adopted for patient safety improvement.
Check Out Our Blog
Learn about what we’ve been up to this past quarter:
- February 11, 2021: Despite the Pandemic, 2020 Brought Immense Innovation Across the Biopharmaceutical Research Ecosystem
- February 25, 2021: TransCelerate BioPharma In the News
- March 5, 2021: Nimbleness and Innovation Are Top Themes at SCOPE 2021
- March 11, 2021: How Shionogi and Bristol-Myers Squibb leverage data sharing to make data-driven decisions on compounds
- March 25, 2021: Validating Intelligent Automation Systems in Pharmacovigilance: Insights from Good Manufacturing Practice
Intelligent Automation Opportunities in Pharmacovigilance Initiative
Health care organizations and pharma companies are required to detect, assess, and report adverse events to ensure regulatory compliance and patient safety. Established pharmacovigilance processes allow this to be done, but can often be highly manual, tedious, and resource intensive.
Enter intelligent automation technologies, which have become increasingly popular due to their ability to automate manual work. Some of these automations can streamline the steps of a workflow while others can transform unstructured or semi-structured adverse event information into structured data, further streamlining the pharmacovigilance process. However, there are obstacles to its implementation. This challenge led TransCelerate to create the Intelligent Automation Opportunities in Pharmacovigilance Solutions Initiative, which focuses on identifying how intelligent automation technologies can be better used to support execution of pharmacovigilance activities.
We interviewed Kristof Huysentruyt (UCB) about TransCelerate’s recently published paper “Validating Intelligent Automation Systems in Pharmacovigilance: Insights from Good Manufacturing Practices” and discussed how machine learning, natural language processing, and artificial intelligence (AI) are becoming more prevalent across the biopharmaceutical industry as well as the challenges the industry is facing in validating this technology.
Q: How is automated intelligence and other machine learning algorithms impacting research and development, especially in pharmacovigilance?
A: Automated technologies are impacting all departments within research and development, not just pharmacovigilance. There are three big buckets that companies are working on already:
- Optimization of existing processes: By applying these technologies to rote tasks with large and ever-increasing volume, such as adverse event reports, we can better manage the growth in this volume. This allows human capital to be used on value-added processes rather than having a department full of data entry professionals.
- Insights from data: These algorithms can be very useful to sift through enormous amounts of data collected during the pharmacovigilance process. Both regulators and sponsors are interested in ensuring that we use this type of technology to quickly identify the most important cases, which ultimately can improve safety for patients.
- Unlocking new data sources: Real World Data is an area many companies and health authorities are looking at artificial intelligence to help solve. These algorithms can help mine this data and extract unstructured data from electronic health records. This will complement data that is already captured in a structured way using existing processes. One great example of what could be abstracted is a patient’s smoking status, which is often not captured in a structured way in electronic health records. Overall, these technologies promise to supplement the value that you can get out of Real World Evidence.
Q: Why did your paper focus on validation of emerging technologies?
A: Demonstrating that something is fit for purpose is a requirement and that does not change because it is a new type of system. What changes is how we demonstrate a newer technology is fit for purpose, and the present guidance is not yet sufficient to cover all the questions that arise from the usage of AI-based systems. Companies are starting to implement these systems and it can be a challenge for health authorities and sponsors to be confronted with systems they’ve never seen before. In many cases, they are not sure how to assess these newer systems or don’t know what questions to ask, so part of the goal in writing this paper was to spark a dialogue around approaches to validation.
Q: What type of systems are going to be the most challenging to validate?
A: Dynamic AI-based systems are going to be the most challenging to validate, because they evolve while in production. This is very unlike the traditional principle of validation, where something is validated before being put into production and when there is a change, it happens via a change control process. This would be the first time systems in production could potentially evolve by themselves independent of human quality checking. We can cover static systems with an extension of the existing framework, but dynamic systems change the framework entirely.
Q: Why do you find it so important to provide insight into why decisions are made within AI-based systems behind the scenes?
A: There is a clear expectation from regulators that a system cannot be a black box in terms of how it makes decisions. There is an expectation that you can at least demonstrate the device is fit for purpose. More complex AI algorithms are so sophisticated and built in such a way that explaining them is not possible in a meaningful way to a human. That may make it harder to implement some of the more complex algorithms.
Q: How will automation impact the pharmacovigilance profession?
A: Automation will not eliminate the role of a medical reviewer but will instead allow human labor to focus on more value-added tasks and analysis of safety signals. Filtering and recognizing patterns are tasks these algorithms do well and jobs that are more apt to the capabilities of this technology. However, a medical reviewer will still be required to interpret the results and decide whether further work needs to be done in the safety signal detection process before moving forward.
Implementation of AI technologies does change the expectations of a medical reviewer, as they cannot be “allergic” to technology anymore. PV professionals need to be at least comfortable working with and understanding the inputs and outputs for these types of algorithms. This is even true for all professionals across drug development. As these technologies continue to be deployed, all of us will need to become more familiar with these technologies, their limitations, and advantages, at least on a high-level basis.
Mark Your Calendars
TransCelerate leaders and Member Companies are invited to participate and present at many biopharmaceutical research industry conferences and meetings across the globe to provide a perspective on industry challenges and clinical trial issues. Take a peek at a few places we will be presenting early next year. For detailed information on speaker presentations, visit the events page on the TransCelerate website.