Inter‐Rater Disagreements in Applying the Montreal Classification for Crohn's Disease: The Five‐Nations Survey Study
Bibliographic record
Abstract
BACKGROUND: The Montreal classification has been widely used in Crohn's disease since 2005 to categorize patients by the age of onset (A), disease location (L), behavior (B), and upper gastrointestinal tract and perianal involvement. With evolving management paradigms in Crohn's disease, we aimed to assess the performance of gastroenterologists in applying the Montreal classification. METHODS: An online survey was conducted among participants at an international educational conference on inflammatory bowel diseases. Participants classified 20 theoretical Crohn's disease cases using the Montreal classification. Agreement rates with the inflammatory bowel diseases board (three expert gastroenterologists whose consensus rating was considered the gold standard) were calculated for gastroenterologist specialists and fellows/specialists with ≤ 2 years of clinical experience. A majority vote < 75% among participants was considered a notable disagreement. The same cases were classified using three large language models (LLMs), ChatGPT-4, Claude-3, and Gemini-1.5, and assessed for agreement with the board and gastroenterologists. Fleiss Kappa was used to assess within-group agreement. RESULTS: Thirty-eight participants from five countries completed the survey. In defining the Montreal classification as a whole, specialists (21/38 [55%]) had a higher agreement rate with the board compared to fellows/young specialists (17/38 [45%]) (58% vs. 49%, p = 0.012) and to LLMs (58% vs. 18%, p < 0.001). Disease behavior classification was the most challenging, with 76% agreement among specialists and fellows/young specialists and 48% among LLMs compared to the inflammatory bowel diseases board. Regarding disease behavior, within-group agreement was moderate (specialists: k = 0.522, fellows/young specialists: k = 0.532, LLMs: k = 0.577; p < 0.001 for all). Notable points of disagreement included: defining disease behavior concerning obstructive symptoms, assessing disease extent via video capsule endoscopy, and evaluating treatment-related reversibility of the disease phenotype. CONCLUSIONS: There is significant inter-rater disagreement in applying the Montreal classification, particularly for disease behavior in Crohn's disease. Improved education or revisions to phenotype criteria may be needed to enhance consensus on the Montreal classification.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".