Research & Insights

Artificial intelligence and healthcare

Evaluating technology in resource-limited healthcare

WHO publications and DECIDE-AI address different parts of healthcare technology evaluation.

Health-system conditions

WHO’s 2019 digital health guideline assesses interventions through benefits, harms, acceptability, feasibility, resource use and equity. These dimensions make staff capacity, infrastructure and implementation requirements relevant to the evaluation of a healthcare tool.

AI performance and live evaluation

WHO’s 2023 publication sets out high-level considerations for evaluating and monitoring AI systems. DECIDE-AI addresses reporting of early, small-scale live evaluations of AI decision support. It concerns use in clinical care, a different stage from testing on previously collected records.

Scope of these sources

The WHO publication is an overview of regulatory considerations, not a regulatory approval. DECIDE-AI is a reporting guideline; adherence does not certify the effectiveness of a particular tool.

Sources and further reading

  1. WHO. Recommendations on digital interventions for health system strengthening (2019)
  2. WHO. Regulatory considerations on artificial intelligence for health (2023)
  3. Vasey et al. Reporting guideline for the early stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI. BMJ 2022;377:e070904

Educational summary of the cited sources. No original study findings are reported.

Back to Research & Insights