Presentation (CTI)

OBJECTIVES: Canada’s Drug Agency (CDA-AMC) conducted an evidence review of RapidAI, an artificial intelligence (AI)-enabled health technology. The review supported deliberations for developing evidence-informed recommendations on implementation for stroke detection in Canada.
METHODS: The assessment included a review evaluating the effectiveness, accuracy, and cost-effectiveness of RapidAI for stroke detection. Ethics and equity considerations were integrated throughout, and were informed by literature and patient, clinician, and other expert input. The Health Technology Expert Review Panel (HTERP), an advisory body to CDA-AMC, reviewed the evidence and developed recommendations on the appropriate use of RapidAI for stroke detection. HTERP used a new deliberative framework to guide their recommendations, considering the following domains: unmet clinical need, clinical value, economic considerations, impacts to health systems, and distinct social and ethical considerations.
RESULTS: Given the uncertainty and gaps in the evidence regarding clinical, economic, and equity value of RapidAI, HTERP did not provide recommendations for or against its new implementation for stroke detection. The clinical evidence suggests that using AI functionalities of RapidAI to assist clinicians making stroke diagnoses may result in clinically important reductions in radiology report turnaround time. The effects of RapidAI on other clinical outcomes were very uncertain. Ethical and equity considerations related to patient autonomy, privacy, transparency, access, and algorithmic bias have implications across the technology lifecycle when using RapidAI for stroke detection. We found no relevant economic evaluations assessing its cost-effectiveness. In sites where RapidAI has already been implemented for stroke detection, HTERP recommends the generation of evidence to evaluate its value.
CONCLUSIONS: While RapidAI has potential to improve time to diagnosis, its impact on many outcomes and its cost-effectiveness are uncertain. Our appraisal and deliberative processes identified evidence limitations that may be common across many AI-enabled health technologies. Lessons learned will support future assessments of AI-enabled health technologies by CDA-AMC.