About the Eosinophilic Oesophagitis Endoscopic Reference Score (EREFS)
Five features scored on the worst-affected segment: oedema 0–1, rings 0–3, exudates 0–2, furrows 0–1 and stricture 0–1, giving a maximum of 8. What matters more than the total is its composition. The inflammatory subscore — oedema plus exudates plus furrows, maximum 4 — is what responds to topical corticosteroids, proton pump inhibitors or dietary elimination. The fibrostenotic subscore — rings plus stricture, also maximum 4 — reflects established remodelling and often will not, so a score of 4 made entirely of rings and stricture means something quite different from a 4 made of exudates and oedema. A normal-looking oesophagus does not exclude the diagnosis, which is why biopsies are mandatory regardless.
Formula
EREFS = oedema (0–1) + rings (0–3) + exudates (0–2) + furrows (0–1) + stricture (0–1)
Inflammatory subscore = oedema + exudates + furrows (0–4)
Fibrostenotic subscore = rings + stricture (0–4)
Total maximum 8- Inflammatory features
- Oedema, exudates and furrows. These respond to topical corticosteroids, proton pump inhibitors and dietary elimination.
- Fibrostenotic features
- Rings and stricture. These reflect remodelling and often require dilation rather than anti-inflammatory therapy alone.
- Score the worst-affected segment for each feature, so that maximal disease burden is captured rather than averaged away.
- The original system grades furrows 0–1, giving a maximum total of 8. Some later studies grade furrows 0–2, and some report proximal and distal segments separately for a total up to 20 — always state which convention a score was recorded under.
- A score of zero does NOT exclude eosinophilic oesophagitis. A meaningful minority of patients with histologically confirmed disease have a normal-appearing oesophagus, which is why biopsies are mandatory whenever the diagnosis is considered.
- The subscore split carries more information than the total. The same total of 4 means different things depending on whether it came from inflammatory or fibrostenotic features.
- The rings grading is anchored to endoscope passage rather than to appearance alone, which is what makes grade 2 and 3 reproducible.
Interpreting the result
Read the two subscores before the total. An inflammatory-only pattern is the one most likely to respond to treatment, and the inflammatory subscore falling on a follow-up endoscopy is good evidence that therapy is working. A fibrostenotic-only pattern suggests established remodelling without much active inflammation, and is the pattern least likely to improve with medical therapy alone — here, dilation is often what changes a patient's dysphagia, and it is worth confirming histological activity before escalating anti-inflammatory treatment for a finding that will not respond to it. A mixed pattern needs both approaches, and importantly means residual dysphagia after successful anti-inflammatory treatment should not automatically be read as treatment failure: the inflammation may have resolved while the rings have not. A total of zero is not a negative test. Take biopsies anyway — at least six, from at least two levels — because endoscopic appearance and histology diverge often enough that relying on the former alone misses the diagnosis.
| Score | Band | What it means | Action |
|---|---|---|---|
| Total 0 | No endoscopic features | Does not exclude eosinophilic oesophagitis — a meaningful minority of histologically confirmed cases have a normal-appearing oesophagus | Take biopsies regardless; record the zero as a baseline |
| Inflammatory subscore > 0, fibrostenotic 0 | Inflammatory features only | Oedema, exudates or furrows without remodelling. The pattern most responsive to medical therapy | Topical corticosteroids, proton pump inhibitor or dietary elimination; expect the inflammatory subscore to fall |
| Fibrostenotic subscore > 0, inflammatory 0 | Fibrostenotic features only | Rings or stricture without active inflammation. Least likely to improve with medical therapy alone | Consider dilation; confirm histological activity before escalating anti-inflammatory treatment |
| Both subscores > 0 | Mixed inflammatory and fibrostenotic | Active inflammation on a background of established remodelling | Treat the inflammation and reassess before attributing residual dysphagia to treatment failure; dilation may still be needed |
What the EREFS (Eosinophilic Oesophagitis) needs (5 inputs)
- Oedema (0–1)
- 0 absent, with distinct vascular markings visible; 1 present, with loss of vascular markings. An inflammatory feature, historically described as mucosal pallor or decreased vascularity.
- Rings (0–3)
- 0 absent; 1 mild, subtle circumferential ridges; 2 moderate, distinct rings that a standard adult endoscope passes; 3 severe, distinct rings that it does not. A fibrostenotic feature, historically called concentric rings, corrugated oesophagus or trachealisation.
- Exudates (0–2)
- 0 absent; 1 mild, involving under 10% of the surface area; 2 severe, involving over 10%. An inflammatory feature, historically called white spots or plaques.
- Furrows (0–1)
- 0 absent, 1 present. An inflammatory feature, also called vertical lines or longitudinal furrows. The original system grades this 0–1; some later studies use 0–2, which changes the maximum total.
- Stricture (0–1)
- 0 absent, 1 present. A fibrostenotic feature.
What it returns
- Total EREFS
- 0 to 8, summing all five features scored on the worst-affected segment.
- Inflammatory subscore (0–4)
- Oedema plus exudates plus furrows. The component expected to fall with successful medical therapy, and the one worth tracking across endoscopies.
- Fibrostenotic subscore (0–4)
- Rings plus stricture. Reflects established remodelling and may persist despite adequate anti-inflammatory treatment.
How it is calculated
The system began as a nomenclature problem rather than a scoring one. Hirano and colleagues noted that the same endoscopic findings in eosinophilic oesophagitis were being reported under a proliferation of different names, which made it impossible to compare studies or to say whether an individual patient had improved. They defined five features in standardised terms, attached severity grades where a meaningful gradation existed, and then tested interobserver agreement by having 21 gastroenterologists score video recordings from 25 patients. Agreement was good for the four major features after modification, which is what the paper set out to establish. The grading choices reflect what can actually be judged consistently: rings are anchored to whether a standard adult endoscope passes rather than to subjective severity, exudates to a surface-area proportion, and oedema, furrows and stricture kept binary because intermediate grades did not reproduce.
Facts & figures
| Feature | Group | Grades |
|---|---|---|
| Oedema | Inflammatory | 0 absent (vascularity visible) · 1 present (loss of vascular markings) |
| Rings | Fibrostenotic | 0 absent · 1 subtle ridges · 2 distinct, endoscope passes · 3 distinct, endoscope does not pass |
| Exudates | Inflammatory | 0 absent · 1 under 10% of surface · 2 over 10% |
| Furrows | Inflammatory | 0 absent · 1 present |
| Stricture | Fibrostenotic | 0 absent · 1 present |
Inflammatory maximum 4, fibrostenotic maximum 4, total maximum 8. Rings are graded by whether a standard adult endoscope passes, which is what makes grades 2 and 3 reproducible.
| Findings | Total | Inflammatory | Fibrostenotic | Implication |
|---|---|---|---|---|
| Oedema 1, exudates 2, furrows 1 | 4 | 4 | 0 | Should respond to medical therapy |
| Rings 3, stricture 1 | 4 | 0 | 4 | Established remodelling; consider dilation |
| Oedema 1, rings 3 | 4 | 1 | 3 | Mixed — treat inflammation, expect residual rings |
Three patients with an identical total of 4 face three different treatment decisions. This is the argument for recording both subscores rather than the total alone.
Evidence
Derivation and validation — Hirano et al.
2013 · n = 25Endoscopic video recordings from 25 patients with eosinophilic oesophagitis, reviewed by 21 gastroenterologists to assess interobserver agreement on a proposed classification and grading system for the major oesophageal features.
After modification, the grading system demonstrated good agreement for the four major features of eosinophilic oesophagitis — rings, furrows, exudates and oedema — with moderate to good kappa values, establishing common nomenclature and severity scores for assessing disease activity.
Diagnostic and treatment-response accuracy — Dellon et al.
2016Assessment of the EREFS for distinguishing eosinophilic oesophagitis from controls and for measuring response to treatment.
Confirmed that the score distinguishes patients with eosinophilic oesophagitis from controls and decreases with treatment, with histological responders scoring significantly lower than non-responders — supporting its use as a serial measure rather than a single-time-point one.
AGA and Joint Task Force clinical guidelines
2020AGA Institute and Joint Task Force on Allergy-Immunology Practice Parameters clinical guidelines for the management of eosinophilic oesophagitis.
Set out the treatment framework — topical corticosteroids, proton pump inhibitors and dietary elimination — within which the inflammatory subscore is the component expected to respond.
Application across paediatric age groups
2020Study applying the EREFS to children of different ages to assess whether the features and grading transfer to paediatric practice.
Supported use in children while highlighting that fibrostenotic features are less common in younger patients, so the inflammatory subscore carries proportionately more of the information in paediatric disease.
How it compares
EREFS (Eosinophilic Oesophagitis) vs Oesophageal biopsy and eosinophil count
Histology remains the diagnostic standard — EREFS describes appearance and cannot diagnose or exclude the disease.
Diagnosis rests on oesophageal eosinophilia, conventionally 15 or more eosinophils per high-power field, in the appropriate clinical context. EREFS correlates with activity and falls with treatment, but a normal-appearing oesophagus occurs in a meaningful minority of confirmed cases, so the score can never substitute for biopsies. In practice they are used together: histology for diagnosis and remission, EREFS for the structural picture that histology cannot show.
EREFS (Eosinophilic Oesophagitis) vs Los Angeles classification
Different diseases in the same organ — LA grades reflux erosions, EREFS grades the features of eosinophilic oesophagitis.
The two are frequently in the same differential, since both present with dysphagia and heartburn, and proton pump inhibitors treat both. They are not interchangeable: a mucosal break is not an EREFS feature, and furrows and rings are not LA findings. An endoscopy assessing dysphagia should record whichever applies, and take biopsies if eosinophilic oesophagitis is plausible.
EREFS (Eosinophilic Oesophagitis) vs Chicago Classification v4.0
Complementary — EREFS describes what the oesophagus looks like, Chicago describes how it moves, and dysphagia often needs both.
Eosinophilic oesophagitis can produce motility abnormalities, and a patient with dysphagia and a normal endoscopy may need manometry. Conversely a patient with rings on EREFS has a structural explanation that manometry will not add to. The two answer different halves of the dysphagia question.
Pearls & pitfalls
- A normal-looking oesophagus does not exclude eosinophilic oesophagitis. A meaningful minority of confirmed cases score zero, which is why biopsies are mandatory whenever the diagnosis is considered.
- Record both subscores, not just the total. Three patients with a total of 4 can face three different treatment decisions depending on composition.
- Score the worst-affected segment for each feature, so maximal disease burden is captured rather than averaged out.
- The original system grades furrows 0–1 for a maximum of 8. Some studies use 0–2, and some score proximal and distal segments separately up to 20 — always state which convention you are using.
- Ring grades 2 and 3 are separated by whether a standard adult endoscope passes, not by how severe the rings look. Use the actual observation.
- Fibrostenotic features may not improve with medical therapy. Residual dysphagia after the inflammatory subscore normalises is not necessarily treatment failure.
- Take at least six biopsies from at least two oesophageal levels, targeting exudates and furrows where present — eosinophil density is patchy.
- The score grades appearance only and does not replace histology, which remains the diagnostic standard and the measure of remission.
- Repeat the score at follow-up. Its designed purpose is serial comparison, and a single value in isolation wastes most of its information.
Critical actions
- Take oesophageal biopsies at every endoscopy where eosinophilic oesophagitis is considered, regardless of the EREFS — including when it is zero.
- Record the inflammatory and fibrostenotic subscores separately in the endoscopy report.
- Where a stricture or grade 3 rings are present, consider dilation alongside anti-inflammatory therapy rather than instead of it.
- Repeat endoscopy with scoring after a treatment trial, and judge response on the inflammatory subscore and histology together.
- Before escalating anti-inflammatory treatment for a fibrostenotic-only pattern, confirm there is histological activity to treat.
- State which grading convention was used, since the furrows scale and segment reporting vary between studies.
- Assess for food impaction risk and counsel accordingly where rings or stricture are present.
Why this score exists
The paper's stated aim was interobserver agreement, not severity measurement, and that ordering explains the design. Before EREFS the literature contained at least four names for the same ring finding, which meant studies could not be pooled and a patient's endoscopy report could not be compared with the one from two years earlier. Fixing the vocabulary came first; grading was added only where a gradation could be judged consistently, which is why oedema, furrows and stricture stayed binary while rings earned four levels. The choice to anchor ring severity to whether a standard adult endoscope passes is the sharpest example of that discipline — it replaces a subjective impression of narrowing with something the endoscopist directly observes while doing the procedure anyway. The inflammatory and fibrostenotic split then fell out naturally, and turned out to carry more clinical information than the total it was derived from.
About the creator
First author, 2013 validation study
Led the development and validation of the EREFS classification and grading system, and co-authored the AGA/JTF management guidelines.
Co-author of the 2013 validation study.
Co-author of the 2013 validation study.
Limitations
- Grades endoscopic appearance only and cannot diagnose or exclude eosinophilic oesophagitis, since a meaningful minority of confirmed cases have a normal-appearing oesophagus.
- Grading conventions vary between studies — furrows scored 0–1 or 0–2, and single-segment or proximal-plus-distal totals — so scores from different sources are not directly comparable.
- Interobserver agreement, while good for the major features, was established among gastroenterologists reviewing selected videos rather than in unselected routine practice.
- Oedema depends on the visibility of vascular markings, which is affected by insufflation, light source and image processing as much as by disease.
- The fibrostenotic subscore is coarse: rings and stricture between them offer only five possible values, which limits sensitivity to change in remodelling.
- Does not capture crepe-paper mucosa or narrow-calibre oesophagus, both of which carry practical importance for dilation risk.
- Fibrostenotic features are uncommon in young children, so the score carries less information in paediatric disease.
If you are the patient
Eosinophilic oesophagitis is an allergic condition in which the gullet becomes inflamed, causing food to stick or difficulty swallowing. At endoscopy the doctor looks for five particular changes — swelling of the lining, ring-like ridges, white patches, vertical lines, and narrowing — and scores each one. Adding them up gives the EREFS, out of 8. The useful part is that these five split into two groups. Three of them reflect active inflammation, and these are the ones that usually improve with treatment such as steroid medicines that coat the gullet, acid-reducing tablets, or removing certain foods from the diet. The other two — rings and narrowing — reflect scarring that has built up over time, and these often do not improve with medicines; a stretching procedure may be needed instead. That distinction explains why swallowing can still be difficult after treatment has successfully calmed the inflammation. One important point: even when the gullet looks completely normal, small tissue samples are still taken, because a normal appearance does not rule the condition out.
Frequently asked questions
What is the EREFS?#
The Eosinophilic Oesophagitis Endoscopic Reference Score: oedema (0–1), rings (0–3), exudates (0–2), furrows (0–1) and stricture (0–1), summed to a maximum of 8 on the worst-affected segment.
What is the difference between the inflammatory and fibrostenotic subscores?#
The inflammatory subscore is oedema plus exudates plus furrows (maximum 4) and reflects active inflammation that responds to medical therapy. The fibrostenotic subscore is rings plus stricture (maximum 4) and reflects established remodelling, which often requires dilation instead. The same total can mean very different things depending on the split.
Does an EREFS of zero exclude eosinophilic oesophagitis?#
No. A meaningful minority of patients with histologically confirmed disease have a completely normal-appearing oesophagus. Biopsies are mandatory whenever the diagnosis is considered, regardless of the score.
Is the maximum score 8 or 20?#
The original system grades furrows 0–1 and gives a maximum of 8 for a single segment, which is what this calculator uses. Some later studies grade furrows 0–2, and some report proximal and distal segments separately for a total up to 20 — so always state which convention a score was recorded under.
How are ring severity grades distinguished?#
By whether a standard adult endoscope passes. Grade 1 is subtle circumferential ridges, grade 2 is distinct rings that the endoscope passes, and grade 3 is distinct rings that it does not. Anchoring to endoscope passage rather than appearance is what makes grades 2 and 3 reproducible.
Why is dysphagia sometimes unchanged after successful treatment?#
Because the inflammatory features may have resolved while the fibrostenotic ones have not. Rings and stricture reflect remodelling that anti-inflammatory therapy often does not reverse, so residual dysphagia with a normalised inflammatory subscore is not necessarily treatment failure — it may be an indication for dilation.
How many biopsies should be taken?#
At least six, from at least two oesophageal levels, targeting areas with exudates or furrows where present. Eosinophil density is patchy, so a small number of biopsies from one level risks a false-negative result.
Should the EREFS be repeated?#
Yes. It was designed for serial use, and most of its value lies in the change between endoscopies rather than in any single value. Record it even when the oesophagus looks normal, so that a later score has a baseline to be read against.
References
Original / primary reference
Validation and guidelines
- Dellon ES, Cotton CC, Gebhart JH, et al. Accuracy of the Eosinophilic Esophagitis Endoscopic Reference Score in Diagnosis and Determining Response to Treatment. Clin Gastroenterol Hepatol. 2016;14(1):31-39.
- Hirano I, Chan ES, Rank MA, et al. AGA Institute and the Joint Task Force on Allergy-Immunology Practice Parameters Clinical Guidelines for the Management of Eosinophilic Esophagitis. Gastroenterology. 2020;158(6):1776-1786.