ISPOR 2025 Conference Recap

Lionbridge’s COA team was proud to attend and present at the ISPOR Europe 2025 Conference in Glasgow, Scotland. This year’s conference focused heavily on improving the patient experience, particularly the role of AI in reaching this goal. Participants explored many angles of this topic via presentations, networking events, and posters. Our team presented two posters:

• Balancing Innovation and Propriety: The Question of Copyright Holders’ Opinions on GenAI Usage While Localizing Their Intellectual Property
• Evaluating GenAI Vs. Human Screenshot Review Outputs in eCOA Localization: Does GenAI Hold the Key to Improved Feedback?

Read on to learn more about our posters covering AI-powered COA solutions.

Poster #1: AI-Powered Solutions and Copyright Holders

Lionbridge created this poster to summarize the results of a survey sent to 100 copyright holders. It assessed their perspectives on leveraging AI on their content. These were some key takeaways from our research:

• Copyright holders are more open to proposed AI solutions for eCOA migration than for linguistic validation.

• Copyright holders would feel more secure when exploring AI solutions if they could have certain documentation in place, such as a data security certification, data privacy certification, data handling trust certification, and AI architecture diagrams.

• Most copyright holders were not proactively providing AI usage policies in publicly available forums.

• The pharma industry is currently exploring how to best use AI, and copyright holders are keen for industry-wide guidance on AI usage to be published to bring clarity to this space.

Poster #2: AI-Powered COA Solutions for eCOA Translation and Migration

This poster covered one of the most exciting AI-powered features of our COA services: AI-powered screenshot review for eCOA translation and migration. This tool harnesses the power of AI to compare source master screen reports and target screen reports. It can also compare target screen reports to target legacy content. The first version of the feature, which compares only source and target screen reports, is available to customers now. We’re still completing the final steps in rolling out simultaneous AI review of source versus target screen reports and target screen report versus target legacy content.

The poster covered the tool’s ability to successfully identify 11 types of errors:

• Missing Content
• Untranslated Content
• Incorrect Translation
• Missing eCOA Edits
• Version Number Mismatches
• Formatting Issues
• Morphology Issues
• Tag Issues
• Line Break Issues
• Capitalization Issues
• Scale Anchor Issues

We tested the tool with 16 target languages that spanned a wide variety of alphabets and language families (Latin, Cyrillic, Chinese, Korean, Thai, and more).

Crucially, the poster shared that the tool uses a “human-in-the-loop” process. We use AI-powered screenshot review as a supplemental QA check, rather than fully replacing human linguists.

Get in touch.

Didn’t make it to ISPOR or our poster presentations? Let’s discuss how we can use our AI-powered clinical trial translation tools and services to help your team enhance and improve patient outcomes. We have decades of experience providing life sciences translation solutions to top pharmaceutical and medical device companies. Let’s get in touch to discuss.