Unveiling the Implications of AI in Health Care through Clinical Notes
Health Care and AI: An Imminent Connection
All of us have the experience of visiting a doctor. Before any medical procedures, the doctor might rely on AI to handle the fundamental task of writing clinical notes. There are over 90 different companies marketing their own AI medical scribes or software applications that claim to free healthcare workers from note-taking and allow them more time to address patients' concerns. These applications play a vital role in quality care, as Altschuler and her co-authors argue in a recent issue of The Lancet.Skills trained in the humanities, such as narrative and editorial skills, history, ethics, and observation, are essential for dealing with the radical changes brought about by AI. However, the use of these applications has already shown some issues like omissions, inaccuracies, and the loss of implied information. Given these risks, clinicians need proper training to navigate the new technology.
Storytelling is significant in medicine. Patients share their stories with doctors, and doctors do the same among themselves. Notes serve as an important place where care occurs and are deeply narrative objects. Note-taking is a part of clinicians' training process, where they process patient information and consider possible diagnoses. But what happens if clinicians don't write the notes themselves? There is a risk that care might deteriorate.
The Importance of Notes in Medicine
Notes ensure the continuity of care and facilitate communication between doctors. They also provide legal protection and enable insurance companies to reimburse for care costs. With the advent of patient advocacy movements, patients now have access to their charts via electronic patient portals. Scrutinizing notes can make patients feel heard or ignored, depending on how their accounts are documented. A well-written note can instill confidence in patients, while a poor one can create problems.Some research suggests that some doctors are omitting differential diagnosis from open notes or creating ghost charts to avoid issues. This shows the complexity and challenges associated with the use of AI in health care.
Possible Pitfalls of AI Scribes
AI scribes produce transcriptions and summaries of doctor-patient conversations. But they don't create narratives from scratch; they generate notes using various data sources. There are several problems identified in the article. For instance, one co-author, Ian Huntington, has experienced mistakes like misgendering and generating prejudiced notes. AI can also miss implied information and have automation bias. It is less effective at recording and transcribing the speech of people with heavy accents, especially immigrants. Privacy is another concern as companies promise to anonymize information but there are ways to de-anonymize it.These issues highlight the need for careful consideration of the shift in clinicians' roles from note authors to note editors when using AI scribes. Training in editorial and narrative skills is crucial to maximize patient care.
Finding Solutions
Altschuler suggests that healthcare organizations should assemble an advisory group of various experts to make recommendations on when and how to use AI scribes. It's important to think about whether residents and students should use AI scribes or learn to write notes themselves. Introducing AI scribes too early in the training process might diminish essential medical training.Overall, the use of AI in health care through clinical notes presents both opportunities and challenges. It requires careful consideration and proper training to ensure the best outcomes for patients.