Leaders at the American Society of Clinical Oncology published a formal document June 25 that outlines what the group deems as the most appropriate use of artificial intelligence when it comes to clinical oncology care.
Like many sectors, AI has propelled cancer care into "a new era of discovery," the document states. Because this emerging technology is finding its way into hospitals and health systems, providing clinicians with guidelines for its application in oncology is already necessary.
When considering AI technology for oncological care, the American Society of Clinical Oncology says clinicians should keep the following principles in mind to guide their use:
- Transparency – Algorithms, tools and applications of AI should be "transparent throughout their lifecycle."
- Inform Stakeholders – Ensure that patients and clinicians are both informed thoroughly if/when AI is used in clinical decision-making and patient care.
- Equity and Fairness – Ensure any AI tool used has been evaluated for training biases and protect against any unfair model designs that are not equitable or accessible to all patients.
- Accountability – Adhere to legal, regulatory, and ethical requirements regarding both the use of data as it is related to the tool and the AI system overall.
- Oversight and Privacy – Ensure leaders are aware of institutional politics regarding AI and take steps to build-in "protections that guard clinician and patient autonomy in clinical decision-making and privacy of personal health information."
- Human-Centered Application – Do not eliminate human interactions from healthcare when using AI for care delivery.
"In the coming years, we expect to learn a great deal about how AI will change our healthcare system in both negative and positive ways," the document states.