← All research storiesHealth

Peer-reviewedHealth

AI note-taking saved doctors about a minute per visit, but didn’t help them finish paperwork sooner

Across 1.6 million patient visits, an AI scribe cut time spent on notes, especially for clinicians who used it often, but other measures of workload didn’t improve.

How far alongPeer-reviewedReviewed by independent experts for a journal before publication. Peer review is a quality check, not a grade of evidence strength: confidence still depends on study design, limitations, replication and the wider evidence.
How soon could it reach you?Could reach you soonCould show up in products, services or advice soon. Our rough editorial guess is about 1 to 3 years, not a forecast from the research.
Where it comes fromAnnals of Internal MedicineBeth Israel Lahey Health; Beth Israel Deaconess Medical Center · Oct 27, 2026Read the original →

What the scientists did

Doctors spend a large part of their day writing notes about each visit, often late into the evening. “AI scribes” listen to the conversation between doctor and patient (with permission) and draft the visit note automatically, and many hospitals are rolling them out.

Researchers at Beth Israel Lahey Health in Massachusetts looked at more than 1.6 million patient visits handled by 902 clinicians from July 2024 to January 2026, after the health system offered an AI scribe to anyone who wanted it. They compared how many minutes each clinician had notes open in the electronic health record per visit before and after signing up, and checked whether heavy users benefited more.

Words to know

AI scribe

A tool that listens to a medical visit and uses AI to write a draft of the doctor’s notes.

  • Turning a conversation about a cough into a written visit summary
  • Drafting the “plan” section of a note for the doctor to edit

Electronic health record (EHR)

The digital file where a patient’s medical history, visit notes and test results are kept.

  • Visit notes
  • Lab results
  • Medication lists

Quality improvement study

A study of a change made in real-world care, checking whether it helped, rather than a controlled experiment.

  • Measuring wait times after a clinic changes its check-in system

What they found

After signing up, clinicians spent 1.2 fewer minutes per visit with notes open, on average. Clinicians who used the scribe the most saved 2.5 minutes per visit, while those who rarely used it saw little change.

However, the scribe didn’t increase how many notes were finished the same day, didn’t reduce documentation done after hours, and didn’t change clinical productivity, such as how many patients were seen.

1.6 million+patient visits analyzed
1.2 minless note time per visit on average
2.5 minsaved per visit by the most frequent users

Why it matters to you

At your next checkup, your doctor may ask whether an AI tool can listen in to help write notes. You can ask how it works and who sees the recording. This study suggests these tools can save doctors some typing, but they aren’t a magic fix for doctor burnout yet. It is not medical advice; questions about your own care belong with your doctor.

Where this could lead

Smarter rollouts of AI scribes

Hospitals could focus on training and supporting clinicians who will use the tool often, since that’s where savings showed up.

1–3 years

Studies of note quality

The next step is checking whether AI-drafted notes are accurate and complete.

1–3 years

These are possibilities the research points toward, not promises. Most early findings take years of testing before they reach everyday life, and some never do.

Why scientists care

AI scribes are one of the fastest-spreading uses of AI in medicine. Large real-world studies like this one show what they actually change, and what they don’t, beyond what companies promise.

Keep in mind

This was one health system, and it wasn’t a randomized trial. The study measured time, not whether the notes were accurate or complete. We read the press release, not the full paper.

How far along is the science?

  • Peer-reviewedReviewed by independent experts for a journal before publication. Peer review is a quality check, not a grade of evidence strength: confidence still depends on study design, limitations, replication and the wider evidence.
  • PreprintA manuscript shared publicly, often before or alongside journal review. Check its separate journal and review status (each story says what we know). Results may change.
  • Conference talkPresented to other scientists at a meeting. Usually short and early; a full paper may come later.
  • Agency reportPublished by a science agency such as NASA, NOAA or NIH. Reviewed internally, but not by an outside journal.
  • Press releaseAn announcement from a university or company. Useful, but always check the research it describes.

How soon could it reach you?

  • Could reach you soonCould show up in products, services or advice soon. Our rough editorial guess is about 1 to 3 years, not a forecast from the research.
  • A few years outPromising, but needs more testing or engineering first. Our rough editorial guess is several years, with a lot of uncertainty.
  • Long-term scienceBasic research that builds knowledge. Real-world uses, if any, are far off.
  • Small stepA small update to earlier work. Useful to researchers, but not a change you would notice yet.

Every story links to its original source: the paper itself, or the conference program when no paper is out yet. Explanations are written with AI, usually from the paper’s abstract (each story says what it was written from), then checked against the source. We only use journals, recognized preprint servers, science agencies, scientific conferences and established science news outlets.