Collaboard Export

Version 1

Development and consensus-based validation of (i) a reporting guideline and (ii) a bias and risk assessment tool for studies developing, updating, validating or evaluating AI- or ML- based systems in elderly care settings.

This investigation brings together studies on what AI in the care of older people is, and how it differs from AI in other fields of health care. The aim is to identify the characteristics that matter for reporting and appraising such studies, in order to set the scope of the reporting guideline and the bias and risk assessment tool developed in this project.

This modified nominal group technique study, conducted in October 2026, compared AI in the care of older people with clinical prediction models as covered by TRIPOD+AI and PROBAST+AI. Seven experts answered open questions on six contrastive case pairs in a pre-questionnaire, discussed the emerging characteristics in a 60-minute online meeting, and rated 18 resulting statements on whether they are true and whether they matter for reporting and appraisal.

Some files on the workshop.