The Society for Translational Oncology (STO) has created a Clinical Conversation Guide resource to help clinicians prepare to respond to patients who use AI to research their diagnosis and treatment.  

In a blog post about the new resource, the society explains that patients often turn to AI chatbots to feel more informed and involved in their cancer care. Rather than dismissing patients’ use of AI or rushing to correct its output, the organization recommends that clinicians respond with curiosity, leaving the door open for meaningful conversation and connection. The resource suggests that when a patient mentions asking an AI tool questions about their diagnosis or treatment, clinicians respond by asking what concerns led the patient to AI or what the chatbot told them.  

Next, the guide suggests that clinicians validate their patients’ efforts to find more information as a way to build trust and foster a collaborative environment for shared decision-making. Patients may then be more willing to hear guidance about what AI can and cannot help them with, allowing clinicians to set realistic expectations for AI use without discouraging patients from seeking further answers to their questions.  

The resource provides suggested responses to a handful of common statements. For example, if a patient tells a clinician, “ChatGPT says this isn’t the best treatment,” the clinician can respond, “That recommendation may apply in some situations. Let’s look at how it compares with your diagnosis and treatment goals.” 

Similarly, if a patient says, “AI suggested a different option,” the clinician can respond, “AI can describe treatment options, but it doesn’t know the details that make your situation unique.” 

If a patient expresses doubts about AI, the clinician can respond, “AI can be a useful starting point, but we’ll make decisions based on your medical history, test results, and what’s right for you.” 

Overall, the guide notes that AI’s role in how patients prepare for appointments and engage in their own cancer care is growing. As such, clinicians should acknowledge patients’ desire for more information and help place AI-generated information in a broader clinical context. “AI is not replacing the clinician. But it is changing the room before the clinician even enters it. The providers who adapt to that reality will be the ones who use it to strengthen trust, not strain it,” the society explains. 

“The most effective response may be the simplest: ‘Tell me what AI recommended, and let’s review it together,’” the resource guide concludes.

ASCO AI in Oncology is published by Conexiant under a license arrangement with the American Society of Clinical Oncology, Inc. (ASCO^®). The ideas and opinions expressed in ASCO AI in Oncology do not necessarily reflect those of Conexiant or ASCO. For more information, see Policies.