From data to narrative: Contextualizing patient reported outcomes in narratives

Introduction and aim: As part of their routine practice, some hospitals ask patients to fill out questionnaires on a range of patients reported outcome measures (PROMS) for health-related quality of life topics. These PROMS can be used to inform clinicians, but they can also be communicated back to patients to give them an indication of how they are doing during their disease trajectory. However, understanding PROMS-scores might be difficult for many patients since many have low health literacy or numeracy skills. In this study, we therefore want to explore if giving more context to these numerical PROMS by providing patients with testimonials has added value. More specifically, we aim to explore if personalising such testimonials (based on patient' characteristics) is something patients would value.

Methods: Testimonials are personalised on gender, age and PROMS-outcome (relatively good/bad score). In April-May 2022, patients (N=30-60) will take part in an experiment and interview study. Patients will view two versions of a dashboard (dashboard only/dashboard + narrative). Experimental outcomes include risk understanding, perceived relevance/usefulness, intentions to discuss and format preference. We control for numeracy and health literacy. In the subsequent interview, patients will be asked about their experiences with the dashboards. Interviews will be analysed using thematic analysis.

Results: We hypothesise that personalised testimonials will help patients understand their PROMS-score better, will increase perceived relevance/usefulness of PROMS-scores and will lead to higher intentions to discuss their health outcomes. We also expect that we can identify key themes in topics the testimonials should cover to help improve understanding of PROMS-scores.

Conclusion and discussion: We expect that personalised testimonials could be a valuable addition to routine clinical oncological practice. Since testimonials can be personalised automatically using data-to-text techniques, more health contexts could implement patient testimonials.