Finalist
Collaboration With Patient Partners and Data Scientists to Develop a Lexicon for AI-Enhanced Medical Communication
Summary of work
Non-specialist audiences increasingly require access to healthcare research to empower shared decision-making and foster understanding and trust. However, the language used in patient-facing materials often remains technical, inconsistent, and disconnected from patients' own words. Content development is often constrained by time and cost. Amica Scientific, working with the healthcare business of Merck, Darmstadt, Germany and four patient partners living with myasthenia gravis (MG), co-created a patient lexicon and purpose-built an artificial intelligence (AI) application to address these gaps.
The lexicon captures 118 scientific terms and patient-preferred alternatives relevant to MG. A large language model was used to build a proof-of-concept AI app that identified technical terms in existing content and recommended patient-friendly alternatives. This has evolved into Patient Voice, a generative AI tool which creates plain language text and infographics from scientific source materials.
This project is distinctive because it grounds AI in something that can't be algorithmically inferred: the authentic voices of patients themselves. Human oversight was embedded throughout, from co-creation to medical writer validation. The work won the 2026 International Society for Medical Publication Professionals Best Original Research Award, advancing industry thinking on the responsible integration of patient-led AI tools in medical communications.
Judges’ comments
This entry from Amica Scientific and Merck KGaA was well structured with a strong approach to the methodology. The judges could see a direct line through the development and it had end to end clarity. The entrants used tech to bring to life the meaningful difference to patients.
