Premium
An index with improved diagnostic accuracy for the diagnosis of Crohn's disease derived from the Lennard‐Jones criteria
Author(s) -
Reinisch S.,
Schweiger K.,
Pablik E.,
ColletFenetrier B.,
PeyrinBiroulet L.,
Alfaro I.,
Panés J.,
Moayyedi P.,
Reinisch W.
Publication year - 2016
Publication title -
alimentary pharmacology and therapeutics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 3.308
H-Index - 177
eISSN - 1365-2036
pISSN - 0269-2813
DOI - 10.1111/apt.13727
Subject(s) - medicine , gold standard (test) , logistic regression , diagnostic accuracy , gastroenterology , disease , ulcerative colitis , pathology
Summary Background The Lennard‐Jones criteria are considered the gold standard for diagnosing Crohn's disease ( CD ) and include the items granuloma, macroscopic discontinuity, transmural inflammation, fibrosis, lymphoid aggregates and discontinuous inflammation on histology. The criteria have never been subjected to a formal validation process. Aim To develop a validated and improved diagnostic index based on the items of Lennard‐Jones criteria. Methods Included were 328 adult patients with long‐standing CD (median disease duration 10 years) from three centres and classified as ‘established’, ‘probable’ or ‘non‐ CD ’ by Lennard‐Jones criteria at time of diagnosis. Controls were patients with ulcerative colitis (n = 170). The performance of each of the six diagnostic items of Lennard‐Jones criteria was modelled by logistic regression and a new index based on stepwise backward selection and cut‐offs was developed. The diagnostic value of the new index was analysed by comparing sensitivity, specificity and accuracy vs. Lennard‐Jones criteria. Results By Lennard‐Jones criteria 49% ( n = 162) of CD patients would have been diagnosed as ‘non‐ CD ’ at time of diagnosis (sensitivity/specificity/accuracy, ‘established’ CD : 0.34/0.99/0.67; ‘probable’ CD : 0.51/0.95/0.73). A new index was derived from granuloma, fibrosis, transmural inflammation and macroscopic discontinuity, but excluded lymphoid aggregates and discontinuous inflammation on histology. Our index provided improved diagnostic accuracy for ‘established’ and ‘probable’ CD (sensitivity/specificity/accuracy, ‘established’ CD : 0.45/1/0.72; ‘probable’ CD : 0.8/0.85/0.82), including the subgroup isolated colonic CD (‘probable’ CD , new index: 0.73/0.85/0.79; Lennard‐Jones criteria: 0.43/0.95/0.69). Conclusion We developed an index based on items of Lennard‐Jones criteria providing improved diagnostic accuracy for the differential diagnosis between CD and UC .