Known as ambient listening, this new AI technology is increasingly being used by providers, with the idea being that it can help them avoid the burden of paperwork, allowing them to spend more time in true patient diagnosis and treatment. It uses microphones in exam rooms to listen to conversations between doctors and patients. It then creates a medical report based on the particulars of such conversations. According to a 2024 article produced by the Association of Health Care Journalists (AHCJ), “The technology is already being implemented at multiple medical centers around the country, including Yale New Haven Health in Connecticut, Emory Healthcare in Georgia, the University of Michigan Health-West, the University of Pittsburgh Medical Center and the University of Kansas Health System.”
The Benefits
It has been estimated that physicians can save an average of 10 minutes on notes per day by incorporating this technology. The ambient listing tool produces a progress note or visit report. The doctor reviews it, perhaps adding or revising it, and then approves it. At that point, the AI generated report becomes part of the official medical record.
The AHCJ additionally points out that, as an added bonus, “patients reported their visits were much more engaging, and some said they enjoyed seeing their recorded words posted to patient portals because they felt the doctor understood what they were saying.” And the doctors are equally pleased, as well, with one physician describing the accuracy of the system as “amazing” and adding that he can pay more attention to the patient conversation. He went on to note that he felt less mentally fatigued at the end of the day, per the AHCJ article
The Limitations
Adding a new technology to your practice usually involves an extra expense. And that is certainly the case when it comes to the ambient listening AI tools. Indeed, according to one source, AI listening devices can range from $200 to $600 per month, per physician.
Then, of course, there is the small hitch of patient privacy rights. Before implementing the session using AI listening technology, you would first need to inform the patient that you are doing so and get their permission. It’s a minor inconvenience as it may take a grand total of seven seconds to communicate and receive the patient’s response. Nevertheless, it must be done. And, of course, there is always the chance that the patient may not like the idea of being recorded and thus may not grant you permission to use the ambient listening device.
A final limitation of the system that providers must take into serious consideration is that, while these AI listening tools are often accurate in recording the conversation, assessing them, and producing a nice-sounding report, inaccuracies do sometimes occur; and this can lead to serious consequences if the doctor or practitioner is not careful to review and correct the AI-generated record. Here’s what one doctor stated concerning this issue in an article found in Physicians Practice:
Ambient listening tools are well-positioned to help tackle physician burnout and reduce documentation burdens, allowing providers to spend more time face-to-face with patients. But when the technology adds the occasional symptom or diagnosis that does not align with the patient presentation, or switches a patient’s sex from male to female in the middle of the note, it’s critical that we call out the potential limitations of these technologies. In the same way that busy and overwhelmed clinicians sometimes sign off on transcription without reviewing the notes in detail, it might be tempting to do the same with documentation created via ambient listening. However, to minimize risks to patient safety and to avoid downstream problems with care coordination, billing, and insurance, clinicians must take a more measured approach when embracing these tools.
So, the uptake is: proceed with caution. If you determine it’s worth spending the money on these systems, then it is incumbent on you to use them legally and carefully. Always get the patient’s permission and always double-check the tool’s final readout to ensure accuracy.