VAAT: Validating the quality and the performance of AI-assisted triage and diagnostic tools

Digital and online symptom checkers are an increasingly adopted class of health technologies that enable patients to input their symptoms and biodata to produce a set of likely diagnoses and associated triage advice. In order to make sensible use of this healthcare technology, its potential value and impact must be examined carefully.

As of now, the lack of a standardised process of clinical evaluation has resulted in significant variation in approaches to validation of AI-based diagnostics and triage. Furthermore, a few of these approaches focus on the safety, clinical efficacy, cost-effectiveness, and in some cases, ethical and social implications of these technologies.

The aim of this project is to design and implement an objective, standardised validation process to systematically evaluate the effectiveness, accuracy, and benefits of AI-assisted triage and diagnostics tools in the healthcare system. More specifically, our goal is to develop new methodologies and metrics to quantify the clinical, economic, and societal values of these technologies in the healthcare system in Sweden by using real-world health records of patients.