Reinventing real-time patient-reported outcome measures (PROMs) in clinical practice: Realizing the potential of artificial intelligence (AI) to improve communication between patients and clinicians.

Patient-reported outcome measures (PROMs) are questionnaires in which patients self-report symptom severity. While PROMs can improve clinical communication and health outcomes, there are common barriers to PROMs completion, such as difficulties interpreting and answering questions and cultural differences in symptom expression, which may disproportionately affect anglophone and allophone patients, in addition to the overall challenges these groups face in navigating Québec's predominantly francophone healthcare system. Artificial intelligence (AI) has the potential to revolutionize PROMs collection for these underserved populations. Our team aims to develop an AI-driven process by which anglophone and allophone patients can report PROMs-like data verbally, in an organic, narrative format using their preferred language, with information transmitted to clinicians in real-time as a concise structured summary. As a first step, we will develop and validate a process for extracting PROMs-like information from narrative clinical notes in the electronic health record, using a large language model (LLM). We will use data from the electronic Implementation of Patient-reported Outcomes Across cancer care in Québec (e-IMPAQc) project, focusing on anglophone and allophone patients. Using an LLM on the McGill University Health Centre’s internal server, we will iteratively refine a process to extract PROMs-like data from narrative text. This project will help develop AI-powered solutions to reduce language barriers, improve clinical communication, and enhance healthcare access for Québec's linguistic minority communities.

---

Outputs:

• Daskalo, C., Moore, A., Boone, E., Durieux, B., Loban, K., Kreutzer, T., Lambert, S., & Poenaru, D. (2026, July). Reimagining patient-reported outcomes with generative AI: From clinical narratives to patient voice [Abstract submitted for conference presentation]. 24th International Conference on Artificial Intelligence in Medicine (AIME 2026), Ottawa, ON, Canada. https://aime26.aimedicine.info/

Speaking of Care: Comparing Indicators of Language Identity and Preference for Equitable Access to Health Services in Quebec English-Speaking Populations

Many English-speaking residents in Quebec face challenges accessing health care in their preferred language. In advancing research and changing practices to better meet their needs, a key challenge lies in how we identify individuals who are English-speaking. Shifts in Canada’s demographic composition and increases in immigration has broadened the concept of English-speakers to include not only native English-speakers, but also persons with another mother tongue who speak English most often at home, as well as those who can conduct a conversation in English but not in French. Existing studies use varied and inconsistent approaches to defining linguistic identity such as mother tongue, first official language spoken, or language most often spoken at home.

This project will explore how traditional variables about language translate to the application of healthcare encounters. We will begin by forming a Patient Partner Council to collaborate closely with our research team throughout the project. Using existing large surveys of the Quebec population, we will identify different ways used to measure language and examine patterns and associations between these measures of language and health care access. Finally, we will present these results to English-speaking patients and health care providers in Quebec to obtain their perspectives of the most appropriate measures for capturing language identity and preferences as they relate to healthcare access. Ensuring language is measured in a way that reflects people’s healthcare preferences is a critical step toward building more inclusive, equitable health systems that meet the needs of the diverse English-speaking minority communities in Quebec.

Outputs:

Coming soon…