P280 Endoscopic training leads to a significant increase in agreement of endoscopic scoring for inflammatory bowel disease: results of the IGIBDEndo educational project
M. Daperno*1, M. Comberlato2, F. Bossa3, L. Biancone4, A. Bonanomi5, A. Cassinotti6, R. Cosintino7, R. Mangiarotti7, G. Lombardi8, A. Papa9, R. Pica10, M. Rabagliati11, L. Grassano11, M. Gasparini11, G. Pagana11, R. D’Incà12, A. Orlando13, F. Rizzello14
1AO Ordine Mauriziano, Gastroenterology Unit, Torino, Italy, 2Ospedale di Bolzano, Gastroenterology Unit, Bolzano, Italy, 3IRCCS ‘Casa Sollievo della Sofferenza’ Hospital, Department of Gastroenterology, San Giovanni Rotondo, Italy, 4Tor Vergata University, Gastroenterology Department, Roma, Italy, 5AOU-Careggi, Gastroenterology, Firenze, Italy, 6Luigi Sacco University Hospital, Gastroenterology Department, Milano, Italy, 7S. Camillo-Forlanini Hospital, Gastroenterology Unit, Roma, Italy, 8AORN Cardarelli, Gastroenterology Unit, Napoli, Italy, 9Complesso Integrato Columbus, Internal Medicine, Roma, Italy, 10ASL Roma B, Ospedale Pertini, Gastroenterology Unit, Roma, Italy, 11ISMB and Politecnico di Torino, Torino, Italy, 12Azienda Ospedaliera di Padova, Department of Surgery, Oncology and Gastroenterology, Padova, Italy, 13AO Ospedali Riuniti Villa Sofia-Cervello, Internal Medicine, Palermo, Italy, 14AO S. Orsola-Malpighi & University, Internal Medicine, Bologna, Italy
Endoscopy is widely used as objective clinical and trial endpoint for inflammatory bowel disease (IBD). Endoscopic reporting is subjective, and scores were proposed to limit variability in reporting. The IGIBDEndo project is an educational effort focused at increasing homogeneity in scoring IBD endoscopy amongst peers developed by the Italian Group for Inflammatory Bowel Disease (IGIBD).
The aim of this study was to evaluate agreement, as well as to explore the effect of the educational project on agreement on endoscopic scores.
Physicians attending residential meetings on IBD endoscopic scoring systems independently scored a series of endoscopic clips (5 in 2013 and 6 in 2014, both for Mayo and Rutgeerts score). Scores were discussed when all participants ended their voting. A second round of scoring was collected after discussion. Interobserver agreement was measured by means of kappa or intraclass correlation coefficient (ICC) as appropriate, and effects of the learning programme were measured as significant changes in pre/post-educational meetings. A meta-analytical approach was used to summarise results obtained in meetings in different calendar years.
In total, 229 physicians attended at least 1 training during 14 residential meetings (7 in 2013, 5 clips evaluated and 7 in 2014, 6 clips evaluated). Values of kappa increased significantly both for Mayo (pre-meeting 0.44, 95% CI 0.35 to 0.53, and post meeting 0.63, 95% CI 0.54 to 0.71, p < 0.03) and Rutgeerts score (pre-meeting 0.41 95% CI 0.30 to 0.52, and post meeting 0.78–0.68 to 0.87, p < 0.0001). Intraclass correlation coefficient (ICC) was good both for CDEIS and for SES-CD. Table 1 reports detailed results for variations in kappa pre- and post-learning process (Delta Kappa).
Table 1 Variations (Delta) pre and post meeting, in kappa values (with 95% confidence interval)
|Scoring system||Δ Kappa 2013||Δ Kappa 2014||Δ Kappa 2013 + 2014|
|Mayo||0.27 (95% CI 0.06–0.47)||0.10 (95% CI 0.00–0.36)||0.18 (95% CI 0.02–0.35)|
|Rutgeerts||0.51 (95%CI 0.32–0.70)||0.23 (95% CI 0.15–0.30)||0.37 (95% CI 0.27–0.47)|
Intraclass correlation coefficient for SES-CD and for CDEIS endoscopic scores were good both in 2013 and 2014, but no second round of observation was possible for these scores because of time constrains. Table 2 reports intraclass correlation coefficient in 2013 and 2014 for CDEIS and SES-CD.
Table 2 Intraclass correlation coefficient (with 95% confidence interval) for Crohn’s disease endoscopic
|2013||0.64 (95%CI 0.37–0.94)||0.67 (95%CI 0.41–0.95)|
|2014||0.83 (95%CI 0.62–0.98)||0.86 (95%CI 0.68–98)|
This study showed that sharing a common educational programme might increase substantially peers’ performance in endoscopic scoring for IBD. These results might have relevant effects on clinical performance and clinical trial development.