EMBRACING CHANGE
leverage machine learning for applications , ranging from diagnosis to promoting healthy behaviors . IDx-DR is the first FDA-approved fully autonomous AI . 2 It is used to screen for diabetic retinopathy in primary care settings to identify patients who should be referred to an ophthalmologist .
One of AI ’ s greatest early promises to clinicians is its potential to offload repetitive administrative and documentation tasks , in hopes that physicians can focus more time on patient care and less time in the electronic health record ( EHR ). The American Medical Association ( AMA ) recently published an article discussing the implementation of AI scribes to assist with patient encounter documentation . 3 As described in the article , The Permanente Medical Group in Northern California , the largest in the U . S ., introduced a scribe technology tool that emphasizes the importance of ease of use , accuracy of transcriptions and protection of patient data . The scribe tool uses a smartphone ’ s microphone to transcribe the visit and then uses ML and natural-language processing to summarize the clinical content and produce a note . Physicians who used the tool saved an estimated one hour per day . Notably , time saved was not seen as a productivity tool to encourage squeezing in more appointments . With improved efficiency , clinicians have more time to devote to tasks that require human interaction and physician expertise .
Moreover , AI tools have the potential to augment clinician diagnostic , prognostic and treatment decisions ; and potentially improve quality , affordability and accessibility of health care . Integration of artificial intelligence provides the medical field with many avenues for advancement , but it remains critical to utilize this technology responsibly to enhance patient care .
ARTIFICIAL INTELLIGENCE : NOTABLE CONSIDERATIONS
Despite enormous promise , the integration of artificial intelligence in health care is not without significant ethical and societal implications . Several concerns exist surrounding algorithmic transparency , patient privacy , interpretability and accountability , and the potential transfer of human bias . The four tenets of medicine that we first learn in medical school — patient autonomy , beneficence , nonmaleficence and justice — should extend to AI development and application as they govern our clinical practice . As humans possess inherent bias , bias is inherent in AI development . AI trained on biased data may produce similarly skewed results , potentially exacerbating health disparities . 4 Therefore , AI models should be assessed across diverse populations to mitigate further bias during development and ensure that AI tools augment physicians ’ abilities to care for everyone rather than detract from it .
Notably , generative AI applications , such as ChatGPT , can “ hallucinate ” or fill in gaps with knowledge and do so with confidence . In the previously mentioned AI scribe tool roll-out , a small percentage of notes contained “ hallucinations ,” which is the term used when AI provides false information from the recorded voices or from other input . For example , one physician mentioned scheduling a patient ’ s prostate exam , but the AI scribe recorded that the exam had been performed . In another example , the physician discussed issues with the patient ’ s hands , feet and mouth , which the AI-generated summary recalled as “ the patient being diagnosed with hand , foot and mouth disease .” 3 As we can see , AI-powered technology in medical documentation has yet to be fine-tuned and always requires human oversight . The precision required in patient care documentation must be reviewed by its human originator .
As we continue to discover AI ’ s potential in health care , it becomes increasingly critical to view AI as a way to augment the role of a physician rather than to stand in its place . The success of AI as a supportive tool in health care is predicated not only on the intellectual curiosity we demonstrate as physicians but also upon our constant need for optimization in every facet of our chosen discipline . The intent should remain that physicians and medical trainees are central to the delivery of patient care , using AI to enhance , not replace , their clinical judgment . National efforts are required to ensure that AI systems are transparent , ethically designed and subject to regulatory oversight . In direct response to the growing call for regulatory standards and guidelines in AI development and application in health care , the National Academy of Medicine ( NAM ) launched a three-year project in 2023 to develop an Artificial Intelligence Code of Conduct ( AICC ). 5 Hopefully , this code will serve as the national framework needed to support equitable and responsible use of AI in medicine and health research .
ARTIFICIAL INTELLIGENCE : THE FUTURE IN MEDICINE
Beginning at the undergraduate medical education level , trainees must be equipped to understand and adapt to the evolving health care landscape and contribute to shaping a system that upholds high standards of care . Patients are modern consumers of the health care system , and their questions will continue to evolve to ask how artificial intelligence affects them and their care . To be good stewards of our patients , we must choose to educate ourselves on AI ’ s both realized and untapped potential . AI holds the capacity to reconcile medications , decrease recurrent patient questionnaires , resolve scheduling conflicts and cancellations , enhance weather and disaster planning , and accelerate biomedical research in drug development and clinical trial eligibility : all of these directly impact our patients . Over one hundred years later , the Flexner model and its foundational principles maintain our responsibility to embrace scientific advancements and create progress in medicine to deliver the highest quality patient care .
As selected members of the AAMC Organization of Student Representatives ( OSR ) for the University of Louisville School of Medicine , we are uniquely positioned to learn about systemic changes in academic medicine and participate in national discourse on medical
( continued on page 18 ) March 2025 17