Is perfection necessary for AI in healthcare?

AI can be optimally applied in healthcare environments without crossing human boundaries.

“When it comes to AI, I still apply the principles of the Hippocratic Oath, but restated: AI will do no harm, we will uphold equality and confidentiality, and above all, a human must be kept informed,” said Dr. Shankar Sridharan. , the Chief Clinical Information Officer at Great Ormond Street Hospital, United Kingdom. He spoke in the plenary session ‘Deploying Generative AI in Clinical Environments’ at HIMSS24 APAC.

“Perfection (in AI) is not necessarily great; it takes away from the user’s responsibility. AI is already faster, but it can be kinder and safer – kinder to paying attention to the patients we (doctors) monitor, and safer when a human is always kept informed to verify the AI.”

Dr. Shankar gave a live demonstration of using Ambient AI to automatically generate clinical notes. He spoke to a fake patient about his symptoms and medical history.

“The transcriptions come immediately,” he noted as he showed the output of the AI ​​solution. “The platform automatically segmented notes into clinical notes, including the patient’s profile, allergies, summary of symptoms, observations, and plans for follow-ups with the physician.”

Despite the benefits, Dr. Shankar emphasized that hospitals need to focus more on governance and digital infrastructure for optimal use of AI.

“AI protects the cognitive load, but it has to be done (within a governance framework). We need to look at the rates of word errors and hallucinations. The responsibility lies with hospitals to have systems in place. This could include AI contracts that test an evolving technology. Such systems should not just be for Ambient AI, but for any AI technology.”

Dr. Shankar expressed interest in applying AI to documentation for nurses and other healthcare workers, in addition to physicians.