Are you a BC doctor using AI for patient care? Here's what you need to know

The College of Physicians and Surgeons of British Columbia has released interim guidance for physicians using publicly available artificial intelligence tools such as ChatGPT and Bard AI. The new “ethical principles,” released on April 11, are meant to fill regulatory gaps pertaining to the use of AI in medicine.“AI has demonstrated the capability to assist health care providers with multiple elements of care such as diagnosis, creating treatment plans and writing patient communications,” the college noted in its guidance. “However, AI must be used with caution to ensure patient safety and wellbeing.”Health Canada rolled out a national regulatory framework in 2022 governing the application of AI in medical devices and systems. In 2023, the federal agency published pre-market guidance for manufacturers of machine learning-enabled medical devices, along with requirements for physicians to use only AI-enabled medical devices approved by Health Canada. However, there remains little to no guidance for physicians’ use of AI tools such as large language models, which include ChatGPT and Bard AI.Read: AI chatbots are here to help with your mental health, despite limited evidence they workSaskatchewan’s guidance outlines principles in six areas: patient privacy and consent, accuracy and reliability, transparency, interpretability, bias, and monitoring and oversight. Many AI tools—such as ChatGPT—do not currently comply with privacy and security regulations, according to the college. Because of this, physicians using such tools must ensure not to include personal patient data or any data that identifies the patient or allows the AI to reconstruct the patient’s identity.“If using personal patient data, the data must be securely stored, accessed, and transmitted and the AI tool must comply with applicable privacy and security laws and regulations,” wrote the college.Physicians must also ensure the accuracy and reliability of AI-supplied information relevant to patient care. While AI tools have become increasingly proficient in generating responses that appear to be accurate and reliable, “they can be partially or completely wrong, leading to erroneous decision-making if relied upon without critical thinking.”Physicians must also be transparent about their use of AI tools in clinical decision-making and explain to patients how these tools work and what their limitations are. Doctors must also be able to interpret the clinical appropriateness of results generated by AI and be mindful of the inherent bias of these results, particularly from the perspective of equity, diversity and inclusion.Read: Patient perspective: Is my doctor using AI?The guidance also emphasized the need for physicians to monitor the use of AI in patient care to ensure appropriate and effective application of the technology.“This includes critically reviewing and assessing whether the AI tool is suited for its intended use and the nature of practice it is being applied to,” wrote the college. “Registrants must also ensure that there is appropriate oversight of the AI tool that they or their employees are using, including regular updates and maintenance.”Health Canada’s legislation focuses on directing the development and commercialization of AI devices … it does not regulate the application of publicly available AI-enabled tools, including large language models like ChatGPT and Bard AI, by individual health care providers.“Read: AI and the primary care crisis: Not the solution but part of the solutionA previous version of this article attributed the interim guidance discussed herein to the College of Physicians and Surgeons of Saskatchewan, when it was, in fact released by the College of Physicians and Surgeons of British Columbia. We regret the error.

Are you a BC doctor using AI for patient care? Here's what you need to know

Register for free access or login.

Related Topics