Quality, safety and artificial intelligence (17 May 2024)

Summary

This seminal study by Cabral et al delves into the transformative potential of artificial intelligence (AI) in oncology, highlighting its pivotal role in enhancing healthcare quality and safety. The study aligns with the broader discourse on AI’s capacity to revolutionise healthcare outcomes, drawing from insights previously proposed on the synergy between human expertise and AI across various medical disciplines.

Content

Key points

• Global perspectives from 881 artificial intelligence (AI) and cancer researchers into the future impact of AI on cancer care highlight AI’s potential to improve cancer grading, classification, diagnostic accuracy and follow-up, while also identifying significant barriers to its integration into clinical practice and the need for standardisation in cancer health data.
• Additional information (from human peers or AI) can have a strong influence on prescribing decisions made by intensive care doctors, particularly for AI recommendations, but the presence of an additional simple explanation did not significantly further increase adherence to AI suggestions.
• A methodical approach to assessing and improving the safety of AI systems in a clinical setting is important, as demonstrated through a case study with the AI Clinician for sepsis, providing a concrete example of AI’s potential impact and the complexities involved in safely implementing AI in healthcare.