P121 Delay in diagnosis of Crohn's disease: description over time and identification of associated factors. A 21-year population based study
C. Mickaël*1, H. Sarter2, M. Fumery3, L. Armengol-Debeir4, E. Laberenne5, F. Vasseur2, D. Turck6, B. Pariente1, L. Dauchet7, L. Peyrin-Biroulet8, C. Gower-Rousseau9, G. Savoye4
1Lille University Hospital, Gastroenterology, Lille, France, 2Lille University Hospital, Epidemiology, Lille, France, 3Amiens University Hospital, Gastroenterology, Amiens, France, 4Rouen University Hospital, Gastroenterology, Rouen, France, 5Seclin Hospital, Gastroenterology, Seclin, France, 6Lille University Hospital, Pediatrics, Lille, France, 7Lille University Hospital , Public Health, Lille, France, 8Nancy University Hospital , Gastroenterology , Nancy, France, 9Lille University Hospital , Biostatistics, Lille, France
Delay in diagnosis (DD) of Crohn's disease (CD) may be responsible for complications and a poor response to treatment. The description of DD over time in a population-based CD cohort and its determining factors are unknown and may help implementing corrective measures. The aim of this study was to report in a 21-year population-based cohort the variation over time of DD and to identify associated factors including sociodemographic factors.
Sociodemographic and clinical characteristics of all CD patients issued from the EPIMAD registry and diagnosed from 1990 to 2010 were included. DD was defined as the time between onset of symptoms and CD diagnosis (months). DD was considered as long when in the upper quartile of this time period. Sociodemographic data included the living area of each patient at the time of diagnosis (urban, periurban or rural), the distance between the patient's living area and the nearest gastroenterologist, and the deprivation index of the living area (FDep99). Clinical data were classified according to the Montreal classification. Univariate and multivariate analyses were performed using a logistic regression to identify the baseline characteristics of patients associated with a long DD.
8 704 patients with CD were recorded including 57% of females; 10% of them had < 17 years at CD diagnosis (A1), 67% between 18 and 59 years (A2) and 23% >60 years (A3). The majority (65%) of patients had an ileocolonic CD and 5% had perianal lesions; 17% had a stricturing and 10% a penetrating behaviour. During the whole study period, median DD was 3 months [Q1=1; Q3=7] with no change over time. A long DD was >7 months and was associated in univariate analysis with the absence of weight loss (p=0.04), the presence of extra-intestinal manifestations (p=0.02), pure ileal involvement (L1) (p<0.0001) and stricturing behaviour (p=0.002). In multivariate analysis, only the absence of weight loss (OR=1.16 [1.04-1.28]) and pure ileal location (OR=1.36 [1.18-1.56]) were associated with a long DD. Socioeconomic characteristics were not associated with a long DD.
In this 21-year population-based study of CD patients, median DD was 3 months with no change over time and no influence of socioeconomic baseline characteristics. Only two clinical features (absence of weight loss and L1 location) were associated with a long DD.