To the Editor:

I have read with great interest the manuscript titled “Comparing ChatGPT and Bing, in response to the Home Blood Pressure Monitoring (HBPM) knowledge checklist” by Nico et al. [1], which provides a rigorous evaluation of the accuracy and completeness of AI chatbots in addressing questions related to hypertension. The findings provide valuable guidance on how these technologies can effectively support patients.

In addition to evaluating the manuscript, we would like to offer the following insights and suggestions for further consideration: First, employing a standardized knowledge checklist and utilizing repeated questioning can enhance the reliability of the evaluation by reducing variability compared to open-ended queries. This systematic approach could serve as a model for evaluating other health-related subjects. Second, if feasible, comparing chatbot responses to those of clinicians would offer valuable insights into their capabilities compared to human experts. Collaborating with physicians could facilitate this process. Third, conducting tests to assess the real-world impact on patient-centered outcomes, such as understanding, adherence, and clinical measures, could demonstrate the utility of these tools beyond mere accuracy. Furthermore, the lower-rated ChatGPT responses provide an opportunity to identify knowledge gaps and prioritize targeted improvements to the model. Collaborating with the company could facilitate ongoing learning.

Overall, this study lays an important foundation for further advancements. With refinement and broader testing, AI chatbots hold promise for enhancing hypertension care and improving outcomes on a larger scale.