Sahlgrenska University Hospital coordinates a Swedish–American PROM-AI project with support from Vinnova

In a partnership between Sahlgrenska University Hospital, Boston Children’s Hospital, and Karolinska University Hospital, an advanced AI tool will be developed to provide researchers, healthcare, industry and policymakers an effective process that delivers an accurate understanding of patient health experiences.

The hospital has received close to one million SEK from Vinnova, Sweden’s innovation agency, to support this new, groundbreaking project that combines AI, international collaboration, and patient-reported outcome measures (PROM). A PROM refers to a valid and reliable questionnaire that gather information of the patients’ health, treatment and quality of life experiences reported directly by the patient, without interpretation by anyone else.

The project is coordinated by Michaela Dellenmark Blom, Associate Professor and Director of the Center for Patient-Reported Outcomes and Experiences in Pediatric healthcare at Queen Silvia Children’s Hospital, Sahlgrenska University Hospital. She has long-standing collaboration with all partners involved, making this joint project a natural step.

-The field of PROM and AI is developing rapidly, yet many questions remain unanswered, questions that when answered ultimately could help strengthen patients’ health, and quality of life. International partnerships deepen our understanding of complex challenges. By international partnership, we can bring together data, expertise, and experience that no single actor possesses alone, and solve problems collaboratively. This is how we move from individual insight to global knowledge,” says Michaela Dellenmark-Blom.

The project aims to develop a collaborative AI agent, a support system capable of analyzing interviews and focus groups that contains patients’ narratives about their health experiences, and additionally generating items to a questionnaire, a PROM. Such a questionnaire can measure how patients perceive their symptoms as well as the perceived impact of disease and treatment in daily life.

The collaborative AI-agent will be tested and developed in English and Swedish through a proof-of-concept study of children born with esophageal atresia, a congenital condition where the esophagus does not fully connect from the mouth to the stomach. Within the framework of the partnership, there is a principal investigator in Sweden, Michaela Dellenmark-Blom, and one in the United States Benjamin Zendejas.

-We want to explore whether AI can serve as a co-researcher in qualitative research in two different languages, for example, in the analysis of focus groups and explore whether a collaborative AI agent can achieve the same level of quality as a human researcher, but with significantly shorter time and lower costs, says Benjamin Zendejas, pediatric surgeon, and Surgical Director of the Esophageal and Airway Treatment Center, Boston Children’s Hospital.

- Healthcare is generating more patient-centered data than ever, yet our ability to analyze it meaningfully hasn’t kept pace. If we can use collaborative AI to accelerate qualitative analysis without

compromising quality, we can bring patient-reported insights into clinical care and research much faster, which is especially important for children with rare diseases, he says.

Sahlgrenska University Hospital’s Center for AI is also a driving partner, and Director Magnus Kjellberg highlights the project’s significance:

-This is an exciting example of how we can build national and international research infrastructure around AI and PROM with a strong focus on patient value, and how two centers of excellence at Sahlgrenska University Hospital can come together to achieve this,” says Magnus Kjellberg.

Karolinska University Hospital’s representative is Tomas Wester, Professor of Pediatric Surgery. With extensive experience in international collaborations, he sees clear benefits in the project.

-In the long run, this may facilitate data structuring, clinical risk assessment, and individualized follow-up, which is highly important for children with rare conditions, says Tomas Wester.

The long-term goal is for the collaborative AI-agent to be used across more diagnoses, in different countries, and in languages where PROMs are needed. This project establishes a strong foundation for this continued work.

-We see clear potential for this Collaborative AI-agent to support qualitative analysis and item generation for more patient groups. This is about making healthcare and research more patient-centered and integrated, at Sahlgrenska University Hospital, in Sweden, and globally. I am grateful for the trust shown by Vinnova and proud of our Swedish–American partnership team,” says Michaela Dellenmark Blom.