Insights from PSI 2024: Advancements in digital health technology

The Statisticians in the Pharmaceutical Industry (PSI) held their annual conference in Amsterdam in June 2024. Four ICON biostatisticians attended and in this article we share their insights. They also showcased their expertise in an individual presentation and during a panel session.

During the three-day conference there were talks on a multitude of relevant topics of interest. The conference provided an ideal opportunity to expand our knowledge of current methodological considerations and trends within the clinical trials industry. ICON participants attended sessions on:

digital biomarkers
advanced methods in vaccine trials
patient focused drug development
Bayesian borrowing
adaptive designs
estimands for non-inferiority trials
use of R programming
decision making
European Union (EU) Health Technology Assessment (HTA) in 2025

What the COVID-19 pandemic taught statisticians

The opening keynote set the tone for the conference by putting statistical theories into a real world context. Geert Molenberghs, Director of I-Biostat, Professor at Universiteit Hasselt KU Leuven explored the lessons learned from the COVID-19 pandemic. Under the title “Biostatistics in a post-pandemic era”, he looked at ways of presenting and interpreting COVID-19 incident rates. He illustrated his talk with examples from Belgium and covered the implications for research, public health, drug and vaccine development, policy advice and communication.

Index-based ranking, R and Machine Learning

Samdhan Ghubade, Senior Biostatistician II, ICON, presented a paper in the decision making stream. He spoke about the detriment index-based ranking technique for painkillers for noncommunicable diseases (NCDs).

The ICON attendees found the use of R programming stream particularly useful. Guiyan Lei’s paper shared how Roche handled the use of R for primary analysis and submission. Both Lei’s paper and one from Janina Linnik of Novartis touched on using open source code.

Another emerging trend is the use of machine learning (ML) for data driven biomarker identification. Bayer’s Karl Kochert, Nils Ternes of Sanofi and Mairzia Scelsi of Roche spoke about ML use for digital endpoint validation and analysis methods.

Digital health technology and challenging conventional thinking

Dr Mia Tackney, MRC Biostatistics Unit at Cambridge University, joined Rosemary Abbott, Associate Director, Biostatistics, ICON and Susanne Schaefer, Senior Director Biostatics, ICON, for their one and a half hour session. Karen Smith, Director, Biostatistics, ICON, chaired the session on “Advancements in digital health technology (DHT): Opportunities, challenges and solutions”. Their presentations covered decentralised clinical trials, digital endpoints, and statistical challenges.

The presentations took an industry-wide perspective, outlining considerations and challenges for researchers contemplating adopting DHTs in clinical trials. As well as the broader picture, the session covered more particular challenges such as handling missing data. It was emphasized that missing data was most likely not to be missing at random. Approximately 200 people attended the presentation and the link between academia and industry was considered noteworthy. The audience had many questions which contributed to an interesting discussion. Some commented that the session had challenged their usual thinking, especially in relation to missing data.

Another talk that challenged conventional thinking was the day two keynote speech from Theo Smith, founder of Neurodiversity World and Neuromine. His thought-provoking speech “Neurodiversity at work: Unleashing potential in every mind”, initiated a lively panel discussion.

As well as learning through the formal sessions, we enjoyed the informal learning opportunities that arise during networking sessions and breaks. We left the conference with important insights on the latest trends. We were happy to share ICON’s experience of using DHT in clinical trials with a wider audience of our industry colleagues. We look forward to seeing them again at next year’s conference and seeing how the trends develop in the meantime.

What the COVID-19 pandemic taught statisticians

Index-based ranking, R and Machine Learning

Digital health technology and challenging conventional thinking

Discover how our Biostatistics team can support your clinical operations.

In this section

Connect with us