Author: Nadali. Prepared with AI assistance.
What was published
On November 6, 2026, JAMA Network Open published a study of 209,009 encounters at 14 Boston clinics from May 2025 to May 2026. AI scribe use was associated with more notes signed on the encounter day when clinicians and patients shared a language; findings for interpreter-mediated visits were inconclusive. This single-system observational evidence establishes neither causation nor documentation quality.
Who benefits
Clinic owners and operations managers testing or considering AI help with notes. Nadali’s takeaway: a pooled pilot average should not replace separate checks of different communication settings.
This is a new research publication, not a product launch. It does not establish which tools your clinic can access or how well they work in Ukrainian. Nadali has not conducted its own tests.
What to change in your pilot
These are editorial suggestions for organizing checks, not the study protocol or medical advice.
- Define separate groups before comparing results: a shared clinician–patient language, telephone interpretation, video interpretation and in-person interpretation. Do not collect unnecessary personal data.
- For each group, record encounter counts, the share of reviewed notes signed that day, and correction time. Make differences in sample size visible.
- Assign a responsible professional to assess content separately. Signing a note quickly does not itself confirm that all important details were captured correctly.
- Expand only after evaluating the group where you intend to use the tool. Do not transfer a shared-language result to interpreter-mediated visits.
For broader measurement planning, see our discussion of speed and workload in clinic messaging. This is a different workflow, not validation of note-taking results.
Availability and limits
The primary source was read on November 8, 2026. Before a real pilot, the clinic must separately check data-handling terms, supported languages and note-review arrangements in the selected system. This news item grants no permission to record consultations or dispense with an interpreter.
The cover is an original conceptual AI illustration, not a photograph of a clinic or the study and not evidence of testing.
