Diagnostic EGD: Indications, Technique, and Systematic Mucosal Survey
Procedure Overview
Diagnostic esophagogastroduodenoscopy (EGD) is the cornerstone investigation for evaluating dyspepsia with alarm features, suspected upper GI bleeding, dysphagia, suspected malabsorption, and surveillance of premalignant conditions. A systematic mucosal survey from oropharynx through the second portion of the duodenum, supplemented by structured biopsy protocols (Sydney for H. pylori, Seattle for Barrett’s), maximizes diagnostic yield while minimizing missed lesions. Procedural success depends on adequate preparation, appropriate sedation depth, and disciplined photo documentation.
S1. Indications & Patient Selection
Diagnostic EGD is indicated when symptoms or laboratory findings carry sufficient pre-test probability for a clinically actionable mucosal lesion. Inappropriate use—particularly for uncomplicated functional dyspepsia in patients under 60 with no alarm features—drives healthcare cost and procedural risk without benefit. The American Society for Gastrointestinal Endoscopy (ASGE) and ACG concur that test-and-treat for H. pylori (non-invasive) is the preferred initial strategy below age 60 absent red flags.
Established indications
- Dyspepsia + alarm features: age ≥60 with new-onset dyspepsia, unintended weight loss >5%, dysphagia, odynophagia, persistent vomiting, iron deficiency anemia, palpable mass, family history of upper GI cancer in first-degree relative.
- Upper GI bleeding: hematemesis, melena, coffee-ground emesis, or unexplained iron deficiency anemia (after negative colonoscopy in patients ≥40).
- Dysphagia or odynophagia: structural vs motility evaluation; eosinophilic esophagitis screening (≥6 biopsies from proximal and distal esophagus regardless of endoscopic appearance).
- Suspected malabsorption: celiac serology positive (tTG-IgA), suspected Whipple’s, giardiasis—duodenal biopsy with ≥4 second-portion + 2 bulb specimens.
- Surveillance: Barrett’s esophagus (per Seattle protocol), gastric intestinal metaplasia (high-risk subgroups), atrophic gastritis (operative link OLGA/OLGIM staging), familial syndromes (FAP, Peutz-Jeghers, hereditary diffuse gastric cancer).
- Caustic ingestion: within 12–24 hours; classify Zargar grade for prognosis.
- Foreign body removal: button batteries (immediate), magnets, sharp objects, food impaction (within 6 hours if complete obstruction).
Relative and absolute contraindications
| Category | Specific scenario | Action |
|---|---|---|
| Absolute | Suspected free perforation, hemodynamic instability without resuscitation, unwilling competent patient | Postpone or cancel; resuscitate; obtain consent |
| Relative | Recent MI <30 days, severe coagulopathy (INR >2.5 or platelets <50k for biopsy), large symptomatic AAA, severe COPD | Risk-benefit consultation; reverse coagulopathy if biopsy planned |
| Cautious-go | Pregnancy (any trimester if urgent), advanced age, ESRD, chronic anticoagulation | Proceed if benefits outweigh risk; ASGE anticoagulation algorithm |
Anticoagulation management
Diagnostic EGD without biopsy is a low-risk procedure—continue antiplatelets and DOACs. With planned biopsy, hold DOACs 24–48h based on creatinine clearance, hold warfarin to INR <2.0 (or proceed if INR within therapeutic range and biopsy is small-caliber forceps only per ASGE 2022). Aspirin 81 mg can be continued throughout.
S2. Equipment & Pre-Procedure Setup
Patient preparation
- NPO: solid food ≥6 h, clear liquids ≥2 h (ASA fasting guidelines).
- Informed consent: discuss perforation, bleeding, aspiration, missed lesion risk, sedation reactions; document.
- IV access: 20-gauge or larger peripheral IV; supplemental oxygen 2 L/min nasal cannula; capnography mandatory for moderate/deep sedation.
- Position: left lateral decubitus with chin tucked; bite block; suction at hand.
- Time-out: confirm patient identity, procedure, allergies, anticoagulation status, prior endoscopy reports.
Endoscope and accessories
| Equipment | Specification | Use |
|---|---|---|
| Standard gastroscope | Outer diameter 9–10 mm, 1 working channel (2.8 mm), 4-way deflection | Routine diagnostic exam |
| Therapeutic gastroscope | Larger channel (3.7–6 mm), dual channel for hemostasis | Bleeding intervention, large foreign body |
| Pediatric/transnasal scope | Outer diameter 4.9–5.8 mm | Pediatric, unsedated, severe stricture |
| Image-enhanced endoscopy | NBI (Olympus), BLI (Fujifilm), LCI, i-SCAN (Pentax) | Vascular pattern enhancement, dysplasia detection |
| Biopsy forceps | Standard or jumbo (2.8 mm) | Mucosal sampling |
| Chromoendoscopy dyes | Lugol’s 1.5–3% (esophagus), indigo carmine 0.4%, methylene blue, acetic acid 1.5% | SCC screening (Lugol), pit pattern (indigo), Barrett’s (acetic) |
Sedation tiers
| Tier | Agents (typical adult) | Monitoring |
|---|---|---|
| Unsedated / topical only | Lidocaine 10% spray 2–3 puffs pharynx | Pulse ox |
| Conscious sedation (moderate) | Midazolam 1–2 mg IV + fentanyl 25–50 mcg IV titrated | Pulse ox, BP, ECG, capnography |
| Deep sedation (MAC, anesthesiologist) | Propofol 10–40 mg IV bolus then 25–75 mcg/kg/min | Capnography, BIS optional |
| General anesthesia | Propofol/remifentanil + LMA or ETT | Full anesthesia monitoring |
Reversal agents at hand
Naloxone 0.04–0.4 mg IV q2–3min (max 2 mg) for opioid reversal; flumazenil 0.2 mg IV over 15 sec, repeat q1min to max 1 mg for benzodiazepine reversal. Avoid flumazenil in chronic benzodiazepine users (seizure risk).
S3. Technique: Systematic Mucosal Survey
A disciplined, reproducible exam pattern is the single greatest determinant of diagnostic yield. The British Society of Gastroenterology and ESGE quality framework recommend a minimum 7-minute inspection time for the gastric mucosa alone, with 28 photo-documented anatomical landmarks per ESGE 2016 systematic photo-documentation protocol.
Insertion technique
- Oropharyngeal entry: with patient swallowing on cue, gently advance under direct vision through the cricopharyngeus (UES, ~15–18 cm from incisors). Avoid blind pushing—Zenker’s diverticulum or anterior cervical osteophytes can deflect the scope.
- Esophagus: insufflate minimally, advance with the lumen centered. Note vascular pattern, mucosal breaks, hiatal hernia, Schatzki ring, and the Z-line position relative to the diaphragmatic pinch.
- GE junction documentation: record the distance of the Z-line, hiatal hernia length, and any tongues of salmon-colored mucosa (suspect Barrett’s).
- Stomach entry: pass through the cardia, perform gentle insufflation. Lake-suction excess fluid for clear visualization.
- Retroflexion in fundus: 180° J-maneuver to inspect the cardia, fundus, and lesser curvature from below. Critical for Cameron lesions, fundic varices, and proximal gastric cancers (10–15% of gastric tumors are missed without retroflexion).
- Body and antrum: inspect greater curve rugae, incisura angularis (a “blind spot” requiring deliberate angulation), and antrum.
- Pylorus and duodenum: traverse the pylorus en face, evaluate D1 bulb (ulcers, polyps), then advance through the superior duodenal angle to D2 to identify the major papilla.
Withdrawal and second-look
The withdrawal phase is where most lesions are detected. Re-inspect the duodenum, pylorus, antrum, body (both curvatures), incisura, fundus on retroflexion, cardia, and esophagus during withdrawal with deliberate insufflation/deflation to expose folds. Apply NBI or chromoendoscopy on any suspicious area. Photograph the 28 ESGE-recommended landmarks regardless of pathology presence—the photographic record establishes inspection completeness.
S4. Findings Documentation & Biopsy Protocols
Standard classification systems
| Finding | Classification | Application |
|---|---|---|
| Reflux esophagitis | Los Angeles A–D | A <5 mm break, B ≥5 mm, C bridging <75% circumference, D ≥75% |
| Barrett’s esophagus extent | Prague C&M | C = circumferential length, M = maximal tongue length |
| Bleeding peptic ulcer | Forrest Ia–III | Ia spurting, Ib oozing, IIa visible vessel, IIb adherent clot, IIc flat pigment, III clean base |
| Gastric polyp / superficial neoplasm | Paris classification | 0-Ip pedunculated, 0-Is sessile, 0-IIa slightly elevated, 0-IIb flat, 0-IIc depressed, 0-III ulcerated |
| Gastric ulcer | Sakita-Miwa A1–S2 | Active (A1/A2) → Healing (H1/H2) → Scar (S1/S2) |
| Atrophic gastritis severity | OLGA / OLGIM | Stage 0–IV, stages III–IV high-risk for gastric cancer |
| Esophageal varices | Baveno (small/medium/large) + red signs | Risk stratification for prophylaxis |
Biopsy protocols by indication
Sydney Protocol (H. pylori and gastritis)
Five biopsies in separate jars: 2 antrum (greater + lesser curve, 2–3 cm from pylorus), 2 corpus (greater + lesser curve, mid-body), 1 incisura angularis. Allows for OLGA/OLGIM staging and detects patchy H. pylori distribution missed by single-site sampling.
Seattle Protocol (Barrett’s surveillance)
Targeted biopsies of all visible mucosal abnormalities first, then 4-quadrant random biopsies every 1 cm (high-grade dysplasia) or every 2 cm (no dysplasia/low-grade). Each level in a separate jar. Sensitivity for detecting dysplasia is ~50% with 2-cm and ~65% with 1-cm intervals.
Eosinophilic esophagitis
≥6 biopsies divided between proximal and distal esophagus regardless of endoscopic appearance (rings, furrows, exudates may be absent). Histology threshold: ≥15 eosinophils per high-power field.
Celiac disease
≥4 biopsies from second-portion duodenum (descending) plus ≥1–2 from duodenal bulb (D1). Patchy villous atrophy is common; bulb-only or single-site sampling underdiagnoses celiac by 10–15%.
Suspicious gastric lesion or mass
Minimum 6–8 biopsies from edge and base of ulcer; for suspected linitis plastica, deep “well biopsy” technique (multiple biopsies from same site) or EUS-FNA. All gastric ulcers require biopsy regardless of benign appearance—malignancy rate 1–3%.
Specimen handling pearls
Orient esophageal biopsies on filter paper with mucosal surface up; submit each anatomical site in a separately labeled jar. For Barrett’s, label with “GE junction +X cm”. Inadequate orientation or pooled jars compromise pathologist’s ability to grade dysplasia or stage gastritis.
S5. Complications & Sedation Safety
| Complication | Incidence (diagnostic) | Management |
|---|---|---|
| Sore throat | 10–60% | Reassure; lozenges; resolves <48 h |
| Mild abdominal distension | 20–30% | CO₂ insufflation reduces; usually self-limited |
| Cardiopulmonary events (hypoxia, hypotension) | 0.5–2% | Stop procedure; supplemental O₂; reposition; reverse sedation if needed |
| Bleeding (post-biopsy) | 0.3% | Most self-limited; endoscopic clip, epinephrine injection if persistent |
| Perforation | 0.03% (cervical 0.01%, esophageal 0.03%, gastric <0.01%) | NPO, broad-spectrum antibiotics, CT with oral contrast, surgical/endoscopic closure |
| Aspiration pneumonitis | <0.5% | Suction airway, supplemental O₂, supportive care; antibiotics if pneumonia develops |
| Sedation-related death | 1 in 10,000–25,000 | Resuscitation per ACLS |
When to halt and reassess immediately
- SpO₂ <90% despite supplemental O₂ — pause, reposition, jaw thrust, consider reversal
- Persistent hypotension SBP <90 — fluids, vasopressor (phenylephrine 50–100 mcg IV), withhold further sedation
- New chest or back pain post-procedure — chest CT to exclude perforation/pneumomediastinum
- Crepitus in neck or chest wall — Hamman’s sign — emergent surgical evaluation
- Ongoing hematemesis — re-scope with hemostatic intervention
Pearl: CO₂ over air
CO₂ insufflation reduces post-procedural bloating by >50% and minimizes risk of pneumoperitoneum if a microperforation occurs. CO₂ is absorbed across the mucosa within minutes, while air can persist and propagate dissection planes if perforation is present. Standard of care for diagnostic and therapeutic upper endoscopy in 2026.
Pitfall: Skipping retroflexion
Failure to retroflex in the gastric fundus misses 10–15% of upper gastric neoplasms, large hiatal hernias, fundic varices, and Cameron lesions in chronic IDA workup. Retroflexion adds <30 seconds and is mandatory regardless of indication.
S6. Post-Procedure Care & Reporting
Recovery
- Monitor in recovery 30–60 min; discharge criteria: alert, baseline vitals, ambulating with assistance, no significant pain, voided.
- NPO until gag reflex returns (usually 1 h post-topical anesthetic).
- No driving, alcohol, or major decisions for 24 h after sedation.
- Responsible adult escort required.
Discharge instructions
- Light diet for 24 h; resume normal diet thereafter.
- Resume regular medications (including anticoagulation per pre-procedure plan).
- Return precautions: chest pain, severe abdominal pain, fever >38.5°C, hematemesis, melena, or shortness of breath—immediate ER evaluation.
- Biopsy results discussed at follow-up visit (typically 1–2 weeks).
Standardized report content (ESGE/ASGE quality indicators)
| Element | Required content |
|---|---|
| Indication | Specific clinical question (e.g., “iron deficiency anemia workup, post-negative colonoscopy”) |
| Sedation | Type, doses, response, complications |
| Extent of exam | Distal landmark reached (e.g., “second portion of duodenum”) |
| Gastric inspection time | Documented in minutes (≥7 min target) |
| Findings | Anatomical site, size, classification (LA/Forrest/Paris), photos referenced |
| Interventions | Biopsies (number, sites, jars), hemostatic therapy, dilation, etc. |
| Diagnosis | Endoscopic impression with differential |
| Recommendations | Pathology pending, follow-up interval, surveillance schedule |
When to refer / escalate
- Suspected malignancy on visual inspection — refer to surgical oncology pending histology
- Barrett’s with dysplasia — referral for endoscopic eradication therapy (RFA, EMR, ESD)
- Recurrent or refractory bleeding — interventional radiology or surgery
- Caustic injury Zargar grade 3a/3b — ICU admission and burn/surgical consult
- Failed cannulation, intubation, or technical difficulty — refer to high-volume tertiary center
Post-procedure reporting is the part that closes the loop between what experienced clinicians saw and what the next clinician does. The report template forces five fields: scope reach (D2 vs limited), mucosal description by zone (esophagus / GE junction / fundus + retroflexion view / body / incisura / antrum / pylorus / D1 / D2), retained content (fluid, food, bile), biopsy site map (drawn or annotated), and a discrete recommendation (surveillance interval or repeat indication). Seasoned clinicians also document negative findings explicitly — “fundic rosette intact, no varices, normal Z-line at 40 cm” — because absence of a finding documented carries diagnostic weight, whereas a silent absence reads as an unobserved field.
Same-day discharge criteria after diagnostic EGD: fully recovered from sedation (Aldrete ≥9), tolerating oral fluids, ambulatory with escort, vital signs stable for 30+ minutes, no abdominal pain disproportionate to expected, and written post-sedation precautions (no driving for 12-24 hours, no major decisions, no alcohol). Red-flag symptoms requiring return: severe persistent chest or abdominal pain, hematemesis, melena, fever, neck swelling or crepitus, or worsening dyspnea. Pathology turnaround typically 5-10 business days — experienced practitioners close the loop with a personal phone call for any malignant or high-grade dysplasia diagnosis, never delegating that to portal messaging.
Procedural Experience: The Search Discipline That Catches Hidden Lesions
Retroflexion Is Where Cancer Hides
The cardia and high lesser curvature require deliberate retroflexion with full insufflation and a slow torque sweep. Experienced practice tells trainees that an EGD without a real retroflexion view (long enough to inspect the entire cardiac rosette circumferentially) is incomplete, regardless of what the report says. Seasoned clinicians look specifically for subtle pallor, mucosal nodularity, type IIc depressions, and absence of the normal rosette folds; any of these triggers narrow-band imaging zoom and targeted biopsy.
The Incisura Is the Single Highest-Yield Site
The angularis incisura sits at the tangent of forward and retroflex views. Experienced practitioners deliberately look at incisura twice — once on forward antrum view with the tip angled up, once on retroflexion from the body. Anything more than minor erythema gets NBI zoom and biopsy. They count this as one of their top three “always do” steps.
Wash, Don’t Hurry Past Mucus
Adherent mucus and bile on the gastric body fold often hide flat lesions; experienced practitioners now routinely instill 200-500 mL of dimethicone-water (sometimes with N-acetylcysteine) via the channel, then re-inspect. The first pass experienced clinicians see is often dirty mucosa; the second pass after washing is where the real diagnosis is made. Experienced practice will rinse and re-inspect a single suspicious field three times if needed — the cost is 60 seconds, the gain can be a 2 cm differentiated adenocarcinoma they would otherwise have walked past.
Biopsy Map, Not Just Random Bites
For Helicobacter and atrophic gastritis surveillance experienced practice uses the updated Sydney protocol modified to OLGA/OLGIM scoring: two antrum (greater + lesser), one incisura, two body (greater + lesser). Seasoned clinicians label each jar in real time and dictate the map into the report so the pathologist knows which fragment is which. For any focal lesion, they take targeted biopsies first, then a separate jar for surveillance background mapping. Random sampling without anatomic labeling is data lost; mapped sampling makes follow-up scope and risk stratification meaningful.
Illustrative Scenario
A patient in their fifties was referred for iron deficiency anemia after a “normal” EGD performed elsewhere two months prior. Initial review of outside images showed limited retroflexion and no incisura tangent view. On the repeat scope the treating clinicians deliberately spent time on the high lesser curvature retroflex view, washed thoroughly with 300 mL dimethicone, then re-inspected. A 12 mm pale flat depressed lesion on the lesser curvature near incisura emerged after washing — Paris IIc, slightly nodular margin. Targeted NBI zoom showed irregular microvascular pattern with demarcation; biopsy returned well-differentiated tubular adenocarcinoma confined to mucosa. ESD achieved en bloc R0 resection with curative criteria. Five-year endoscopic surveillance was negative.
Practice Pattern Refinements vs Standard Technique
- 9-12 minute floor for diagnostic EGD inspection time. Endoscopist quality data correlate detection of high-grade IM and early neoplasia with inspection time. The internal minimum is 9 minutes between intubation of esophagus and full withdrawal for diagnostic-intent procedures — never a stopwatch race, never under 7 minutes for any inspection scope.
- Mandatory five-point retroflexion sweep. Cardia 360°, fundic rosette, lesser curvature high, incisura tangent, antral hood — each is a separate deliberate maneuver, not a single retroflexion turn. Photo documentation of each view in the report.
- Dimethicone wash on any mucus-coated field. If the first pass shows adherent mucus, experienced practitioners instill defoamer and re-inspect; they do not biopsy “through” mucus or skip the field.
- OLGIM-style biopsy mapping with named jars. Antrum greater + lesser, incisura, body greater + lesser — five jars labeled by anatomic site. Adds 30 seconds to documentation and gives the pathologist the topographic context that pure “random gastric biopsies × 5” wastes.
S7. Frequently Asked Questions
- Endoscopic Mucosal Resection: Colonic Polyp and Gastric Lesion Technique
- Flexible Bronchoscopy: Indications, BAL, Transbronchial Biopsy, and EBUS
- Capsule Endoscopy: ESGE and ACG Technique with Patency Capsule Indications and Small Bowel Bleeding Crohn Surveillance
- Endoscopic Submucosal Dissection (ESD): Technique, Indications, and Complications for Early GI Cancer
- Endoscopic Ultrasound (EUS): Indications, Technique, FNA, and Therapeutic Applications
- Lumbar Puncture Technique: Indications, Atraumatic Needle Selection, Opening Pressure, and Post-LP Headache Prevention
Primary guideline documents this article draws on. Links resolve via DOI to the publisher; access may require a subscription.