An exploratory study found that the very popular online artificial intelligence (AI) model called ChatGPT provided mostly appropriate responses to simple heart-disease prevention questions, as evaluated by preventive cardiology clinicians. The findings suggest the potential of interactive AI to assist clinical workflows by augmenting patient education and patient-clinician communication around common questions, according to the authors. The findings may also point to using the model for other types of patient-education information.

The researchers created 25 questions addressing heart disease prevention. Examples of the questions for which appropriate responses were received include: What is the best diet for high blood pressure and high cholesterol? I have muscle pain with a statin. What should I do? What is lipoprotein(a)?

Each question was posed to the ChatGPT three times and the responses were recorded. Each set of three responses was graded by an experienced preventive cardiology clinician for two hypothetical contexts: as responses on a patient-facing information platform (similar to hospital-based informational websites) and as AI-generated draft responses to electronic message questions sent by patients for clinician review. Responses to 21 of 25 questions (84%) were graded as appropriate in both contexts.

The study had a number of limitations, including use of an older version of the model; lack of standard criteria for appropriateness; and no references to evidence to support any of the responses. Nevertheless, it points to the potential of the model and directions for future research.

To read the research letter, published in JAMA Network Open, click here

To learn more about ChatGPT and to try it out, click here