Mayo Endoscopic Subscore Interpreter
Mayo Endoscopic Subscore Interpreter
The one Mayo item that carries the mucosal-healing endpoint — and the one observers agree about least. Friability sits at grade 1 in the original wording and grade 2 in the modern one, which flips the verdict.
Mayo endoscopic subscore
Grade + friability → readingErythema with a decreased vascular pattern, and friability seen
Two wordings of the same four grades
Trial wording in current use — 1 mild: erythema, decreased vascular pattern (no friability) · 2 moderate: marked erythema, lack of vascular pattern, ANY friability, erosions
Mucosal healing — subscore 0 or 1 in most protocols and in ECCO-ESGAR; subscore 0 alone by a 15-study meta-analysis
- friability moved between grades
- Schroeder’s grade 1 includes mild friability. The trial wording removes it, so friability forces grade 2, and LUCENT 1 writes its endpoint as “0 or 1 excluding friability”. A friable mucosa therefore changes mucosal-healing status depending on which wording the report followed
- 0 or 1, or 0 alone
- most trials and ECCO-ESGAR define mucosal healing as a subscore of 0 or 1. A systematic review and meta-analysis of 15 studies of steroid-free clinically remitted patients followed at least 12 months found relapse lower at 0 than at 1 (odds ratio 0.33, 95% CI 0.26 to 0.43, I-squared 13%) with no difference in hospitalisation or colectomy, and concluded that 0 should be the target. Both are printed here
- the middle grades are the unreliable ones
- 67 patients, one colonoscopy video each, three expert endoscopists. Weighted kappa between the three pairs was 0.80, 0.52 and 0.49 over the whole range, 1.0 at grades 0 and 3, and 0.715, 0.259 and 0.252 restricted to grades 1 and 2 — the two grades that decide whether a patient has healed
- why UCEIS exists
- the UCEIS was built to give the endoscopic assessment a reproducible structure, and in that same 67-patient study its intraclass correlation was 0.922 (95% CI 0.832 to 0.959). It is NOT implemented here: its developing paper defines Likert levels and a 0–100 visual analogue prediction rather than the additive 0–8 score in general use, that additive form lives in a second paper that could not be read, and the article page carries an explicit copyright notice covering AI use. Freedom from encumbrance could not be established, so it is named rather than reproduced
- worst segment, central reader
- the subscore is assigned on the worst affected area seen rather than averaged, and trials read it centrally precisely because the site endoscopist and the central reader disagree often enough to change a trial result
Worked example
Erythema with a decreased vascular pattern, and friability seen
The appearance without friability is grade 1 in both wordings
Schroeder's original grade 1 reads “mild friability, erythema, decreased vascular pattern”, so friability is inside grade 1 there
The trial wording in current use removes friability from grade 1 and places “any friability” in grade 2, so the same mucosa is grade 2 there
Under the commoner mucosal-healing definition of 0 or 1, this patient has healed on one wording and not on the other — from one endoscopy, with no disagreement about what was seen
The practical move is to record the FINDING (erythema plus friability) alongside the number, so a later reader can apply whichever wording they need
The four grades under both wordings
| Grade | Schroeder (original) | Trial wording in current use |
|---|---|---|
| 0 | Normal or inactive | Normal or inactive disease |
| 1 | Mild: erythema, decreased vascular pattern, MILD FRIABILITY | Mild disease: erythema, decreased vascular pattern |
| 2 | Moderate: marked erythema, absent vascular pattern, FRIABILITY, erosions | Moderate disease: marked erythema, lack of vascular pattern, ANY friability, erosions |
| 3 | Severe: spontaneous bleeding, ulceration | Severe disease: spontaneous bleeding, ulceration |
Agreement between observers, and why UCEIS was built
| Measure | Figure | Range covered |
|---|---|---|
| Weighted kappa, endoscopic subscore, pair A–B | 0.80 | All grades |
| Weighted kappa, pairs A–C and B–C | 0.52 and 0.49 | All grades |
| Weighted kappa, all three pairs | 1.0 | Grades 0 and 3 only |
| Weighted kappa, the three pairs | 0.715, 0.259 and 0.252 | Grades 1 and 2 only |
| UCEIS intraclass correlation | 0.922 (95% CI 0.832–0.959) | All three endoscopists |
One item, two wordings, and the endpoint that rests on it
The endoscopic subscore is a quarter of the full Mayo score and almost all of its scientific weight. Mucosal healing — the endpoint that moved ulcerative colitis trials away from symptoms — is defined on this item alone, which is why it is worth reading on its own and why the partial Mayo score cannot speak to it at all.
Two things make it harder than it looks. The first is that two wordings are in print and they disagree about friability. Schroeder’s original grade 1 is “mild friability, erythema, decreased vascular pattern”; the wording in current trial use is “erythema, decreased vascular pattern” with friability moved into grade 2, and the LUCENT 1 protocol defines endoscopic remission as a subscore of 0 or 1 “excluding friability” in as many words. So a mucosa with erythema and friability and no ulceration is grade 1 under one wording and grade 2 under the other, and since the commoner healing definition is 0 or 1, that single word decides whether the patient has healed. Two endoscopists who agree exactly about what they saw can report different numbers.
The second is reproducibility, and the figures are worse than the instrument’s reputation. Sixty-seven patients were scored from colonoscopy video by three expert endoscopists. Over the whole range, weighted kappa between the three pairs was 0.80, 0.52 and 0.49. At grades 0 and 3 it was 1.0 — nobody mistakes a normal colon for a bleeding ulcerated one. Restricted to grades 1 and 2 it fell to 0.715, 0.259 and 0.252. The grades where agreement collapses are exactly the grades that decide mucosal healing. That is the reason UCEIS was developed, and in the same study its intraclass correlation was 0.922. UCEIS is named here rather than built: its developing paper defines Likert levels and a 0 to 100 visual analogue prediction rather than the additive 0 to 8 score in general use, the additive form lives in a second paper that could not be read, and the article page carries an explicit copyright notice covering text and data mining and AI. Freedom from encumbrance could not be established, so no part of it is reproduced.
Even the healing definition is unsettled. Most protocols and ECCO-ESGAR use a subscore of 0 or 1. A systematic review and meta-analysis of 15 studies, restricted to steroid-free clinically remitted patients followed for at least 12 months, found relapse lower at grade 0 than at grade 1 — odds ratio 0.33, 95% confidence interval 0.26 to 0.43, I-squared 13% — with no difference in hospitalisation or colectomy, and concluded that grade 0 should be the target. Both numbers are in current use and a trial reporting “mucosal healing” means one or the other. This index scores symptoms, not inflammation. It can be high with a normal mucosa — bile-acid diarrhoea, a stricture, bacterial overgrowth, coexisting irritable bowel syndrome — and low with active endoscopic disease. That mismatch is why treat-to-target moved off symptom indices and onto objective markers. The faecal calprotectin interpreter and the CRP unit converter are what sit alongside this item between endoscopies: median calprotectin by subscore ran 94, 168, 428 and 780 micrograms per gram for grades 0 to 3 in a pooled analysis, which is an ordering and not a scale. A score is not a diagnosis and a figure from a cohort is not a probability for one patient. This states what the number meant in a named study; the clinician in front of the patient decides what follows. Thresholds here are the published ones; reference intervals and assay units are method- and laboratory-dependent and your own laboratory’s interval takes precedence.
Frequently asked questions
Does friability make the subscore 1 or 2?
It depends which wording the report followed, and the answer changes the patient’s mucosal-healing status. Schroeder’s original grade 1 includes mild friability. The trial wording in current use removes it from grade 1, so any friability forces grade 2, and the LUCENT 1 protocol defines endoscopic remission as a subscore of 0 or 1 excluding friability. Record the finding as well as the number.
Is mucosal healing a subscore of 0, or of 0 to 1?
Both are in use. Most trial protocols and ECCO-ESGAR use 0 or 1. A meta-analysis of 15 studies of steroid-free clinically remitted patients followed at least 12 months found relapse lower at 0 than at 1 — odds ratio 0.33, 95% CI 0.26 to 0.43 — with no difference in hospitalisation or colectomy, and argued that 0 should be the target. A paper reporting mucosal healing has chosen one of these.
How reproducible is the endoscopic subscore?
Perfect at the ends and poor in the middle. In 67 patients scored from video by three expert endoscopists, weighted kappa between pairs was 0.80, 0.52 and 0.49 across all grades, 1.0 at grades 0 and 3, and 0.715, 0.259 and 0.252 for grades 1 and 2 alone. The grades that decide healing are the ones observers agree about least, which is why trials read endoscopies centrally.
Why is UCEIS not calculated here?
Because freedom from encumbrance could not be established, and because the authoritative point table was not read. The 2012 developing paper scores three descriptors on Likert levels and predicts severity on a 0 to 100 visual analogue scale rather than defining the additive 0 to 8 total in general use; that additive form was published in a second paper which could not be read. The article page also carries a copyright notice covering text and data mining and AI training. UCEIS is therefore named and its reliability figures quoted, and none of it is reproduced.
Which part of the colon is scored?
The worst affected area seen, not an average over the colon, and in acute presentations often only the rectum and sigmoid because the examination is a limited unprepared flexible sigmoidoscopy. A subscore from a limited examination is not the same observation as one from a full colonoscopy, and the report should say which it was.
Related calculators
References
- Belvis Jiménez M, Castro Laria L, Maldonado Pérez B, et al. Index of the Mayo Endoscopy and Ulcerative Colitis Endoscopy Index of Severity: are they equally valid? Rev Esp Enferm Dig. 2020;112(11):821–5.
- Viscido A, Valvano M, Stefanelli G, et al. Systematic review and meta-analysis: the advantage of endoscopic Mayo score 0 over 1 in patients with ulcerative colitis. BMC Gastroenterol. 2022;22(1):92.
- Pfizer. Statistical Analysis Plan, A3921139 (OCTAVE Open), Appendix 2.1: Mayo scoring system. NCT01470612, clinicaltrials.gov.
- Eli Lilly. Clinical Protocol I6T-MC-AMAM (LUCENT 1), endoscopic remission defined as endoscopic subscore 0 or 1 excluding friability. NCT03518086, clinicaltrials.gov.
- Mosli MH, Zou G, Garg SK, et al. C-reactive protein, fecal calprotectin, and stool lactoferrin for detection of endoscopic activity in symptomatic inflammatory bowel disease patients: a systematic review and meta-analysis. Am J Gastroenterol. 2015;110(6):802–19.
Not medical advice. For healthcare professionals and education. Reference intervals vary by laboratory and assay — always use your own laboratory's. Never base a dose or a treatment decision on this page alone. Full disclaimer at calcengines.com/disclaimer/