Otolaryngology (ENT) AI scribe 2026: endoscopy findings, audiometry, sinus surgery, sleep surgery, and the specialty vocabulary

May 8, 2026 · 6 min read

Otolaryngology (ENT) generates documentation across five distinct visit types: office E/M with endoscopy findings, audiometry interpretation, sinus surgery (FESS), sleep surgery (UPPP, hypoglossal nerve stimulator), and head and neck cancer evaluation/treatment. Each has specialty vocabulary that general AI scribes miss.

The 2026 ENT-aware AI scribe stack handles four things general scribes miss: structured endoscopy findings (nasal, laryngeal, ear), audiogram interpretation with pure-tone average and speech discrimination, surgical procedure structures (FESS with anatomic detail, sleep surgery), and head and neck cancer staging documentation aligned with AJCC.

Endoscopy findings structure

The ENT-aware system prompt

You are documenting an ENT encounter.

OUTPUT structured note based on visit content:

For office E/M with endoscopy:
1. Subjective by anatomic system (nasal/sinus, ear/hearing, throat/voice/swallow, sleep)
2. Anatomic exam by region (external ear, otoscopy, anterior rhinoscopy, oral, neck nodes)
3. Endoscopy findings if performed (nasal/laryngeal) by anatomic structure
4. Audiometry interpretation if performed: pure-tone average, speech reception, word recognition, tympanometry
5. Imaging review (CT sinus, MRI, etc.)
6. Assessment by problem with ICD-10 specificity
7. Plan: medical / procedural / surgical with rationale
8. Patient education

For surgical procedure (FESS, septoplasty, tonsillectomy, parotidectomy, etc.):
- Standard op note structure
- Anatomic detail per procedure
- Complications by procedure-specific category
- Specimen pathology

For head and neck cancer:
- Tumor location with subsite
- Size and extent
- Imaging review (CT, MRI, PET-CT)
- AJCC staging (TNM with version 9)
- Multi-disciplinary plan
- Treatment phase tracking

For sleep surgery:
- Pre-op DISE findings
- Procedure performed
- Post-op AHI / oximetry follow-up

Cite transcript or imaging report for each fact. For audiogram, use precise pure-tone averages and speech metrics.

Audiometry integration

ENT practices use audiogram data heavily. The AI scribe should pull audiogram report from the audiology system or EHR and structure:

• Pure-tone average (4-frequency, 6-frequency variants)
• Speech reception threshold
• Word recognition score (and presentation level)
• Tympanometry (Type A/B/C)
• Acoustic reflexes if measured
• Trend across audiograms

The H&N cancer staging integration

Head and neck cancer documentation should structure AJCC TNM staging with subsite-specific rules (oral cavity, oropharynx HPV+/HPV-, larynx by subsite, hypopharynx, salivary, etc.). The AI scribe should produce clinically-precise stage with cited evidence (imaging, exam, biopsy).

Vendor and DIY paths

For ENT practices: native specialty EHRs (NextGen Office ENT, others) have growing ambient AI features. The DIY stack with custom prompt control works well for solo / small group ENT, especially for the H&N cancer subset where staging precision matters.

BAA chain

Practice + EHR + audiology system + imaging vendor + transcription vendor + LLM vendor.

When to start

For ENT practices doing more than 30 endoscopies per week, the structured endoscopy documentation captures billable detail consistently. For H&N cancer practices, the staging structure improves multi-disciplinary care coordination.