How is AI Changing Your Doctor Visit?
From intake forms to ambient scribes, artificial intelligence is transforming your medical visits. A Stanford expert explains the questions every patient should ask.
AI is changing how we understand disease and discover new drugs. But it’s also changing the patient experience. Intake forms are moving from clipboards to iPads, appointment scheduling takes place via chatbots, and telehealth platforms use AI to match individuals to the best doctors for their needs.
How should we think about AI in our medical care?
So far, the patient care experience hasn’t changed all that much, says Nigam Shah, professor of medicine at Stanford University, chief data scientist for Stanford Health Care, and an affiliate of the Stanford Institute for Human-Centered AI. “Nationally, we seem to be using AI to do things we already know how to do, like automated billing.”
But he envisions a future where AI will advance our understanding of disease, enhance the practice of our healthcare professionals, and improve the efficiency of our clinics and hospitals.
In a recent interview, Shah explained what the healthcare AI landscape looks like from the patient’s point of view, and where the hidden risks may lie.
A Convenience or a Privacy Erosion?
You may have already noticed tablets have replaced many of the paper-and-pen patient intake forms you receive during an appointment. These AI-powered patient forms not only build your electronic health record, but the data can also potentially be de-identified and sent to other companies for research, marketing, or training AI models.
Once you get to the exam room, your doctor may use an “ambient scribe” product that leverages generative AI to listen to and transcribe your conversations, noting relevant details such as medical histories, symptoms, diagnoses, and recommended treatments. This information flows into your electronic health record, streamlining the notetaking process while capturing more detail from each visit.
These technologies come with pros and cons for patients, says Shah. Providers say ambient scribes reduce cognitive burden and may also help get better reimbursement from insurance, but the scribes also enable creating more codes that providers can bill for, potentially increasing patient costs in the long term.
While more patient data may train better, more useful AI or help develop life-saving drugs, patients who agree to have their de-identified information used for other purposes may not realize that “de-identification is not anonymization,” Shah explains. “And once your data is stripped of personally identifying information, it’s no longer protected under HIPAA, and it can be bought and sold in a thriving data industry.”
People should focus on the scope of acceptable use, he adds. For example, can a company use your information to price insurance for you? To offer you advertisements for drugs? And would you want to know who is profiting from your information?
Shah recommends patients ask questions of their providers to better understand:
Where will you be storing my intake data?
Which third parties will have access to my personal information?
Can I control whether my de-identified data is used for marketing, research, or training AI models?
How long do you keep my data in this system?
Is this visit being recorded or transcribed in real time?
Where will this conversation be stored and for how long?
Who has access to this conversation and the notes generated from it?
Will a human review the AI-generated notes for accuracy?
Where can I view the transcript that’s generated from this visit, and how do I submit corrections if I see any mistakes?
Can I decline the use of an ambient scribe during my visit?
Enhancing Treatment
Besides AI tools that do mundane tasks, patients might soon see AI that assists doctors in planning your treatment. For example, Atropos Health, which Shah co-founded at Stanford, pulls together hundreds of millions of de-identified healthcare records into a system that doctors can query to find what happened to similar patients in any given health scenario.
“Healthcare AI has reached the point where doctors can ask a question in a chat window and the system creates a study within minutes to generate real-world evidence based on the past care of similar patients,” Shah explains. As a result, the patient may get more effective care, in record time.
Outside the Doctor’s Office
Consumers are also increasingly using AI apps and devices outside their doctor visits.
Shah recently contributed to an article in the Journal of the American Medical Association on AI in medicine that noted there are more than 350,000 mobile health apps available today, with AI frequently embedded in the software. From diagnosing skin cancer to detecting falls to providing mental health support, these popular tools hold promise, but the authors conclude that we need a clearer understanding of their effectiveness and better regulatory oversight if they are to be integrated into the healthcare system.
Looking Forward
In the bigger picture, AI will change not only how we interact with our doctors but also how we manage disease from cancer to diabetes to Alzheimer’s.
But, Shah cautions, consumers need to have realistic expectations for what’s possible and when. “Good progress will happen, but researchers should not overpromise. It’s going to take time.”
Related News
A Stanford HAI workshop brought together experts to develop new evaluation methods that assess AI's hidden capabilities, not just its test-taking performance.
A Stanford HAI workshop brought together experts to develop new evaluation methods that assess AI's hidden capabilities, not just its test-taking performance.
Stanford researchers have developed a deep learning model that transforms overwhelming brain data into clear trajectories, opening new possibilities for understanding thought, emotion, and neurological disease.
Stanford researchers have developed a deep learning model that transforms overwhelming brain data into clear trajectories, opening new possibilities for understanding thought, emotion, and neurological disease.
AI Leaders Discuss How To Foster Responsible Innovation At TIME100 Roundtable In Davos
HAI Senior Fellow Yejin Choi discussed responsible AI model training at Davos, asking, “What if there could be an alternative form of intelligence that really learns … morals, human values from the get-go, as opposed to just training LLMs on the entirety of the internet, which actually includes the worst part of humanity, and then we then try to patch things up by doing ‘alignment’?”
HAI Senior Fellow Yejin Choi discussed responsible AI model training at Davos, asking, “What if there could be an alternative form of intelligence that really learns … morals, human values from the get-go, as opposed to just training LLMs on the entirety of the internet, which actually includes the worst part of humanity, and then we then try to patch things up by doing ‘alignment’?”