P685 Characteristics of non-responders to self-reported questionnaires in a large Inflammatory Bowel Disease cohort study
E. Multone1, J.-P. Vader1, C. Mottet2, A. Schoepfer3, N. Fournier1, B. Burnand1, P. Michetti4, V. Pittet*1
1Institute of Social and Preventive Medicine, Healthcare Evaluation Unit, Lausanne, Switzerland, 2Hôpital Neuchâtelois, Service of Gastroenterology, Neuchâtel, Switzerland, 3Lausanne University Hospital, Department of Gastroenterology & Hepatology, Lausanne, Switzerland, 4Clinique La Source-Beaulieu, Crohn and Colitis Center, Lausanne, Switzerland
A major threat to the validity of longitudinal cohort studies is non-response to follow-up, which can lead to erroneous conclusions. The profile of chronic diseases non-respondents is therefore important, but rarely assessed. The objective of this study was to evaluate the number and profile of non-respondents to self-reported questionnaires among patients included in the Swiss Inflammatory Bowel Disease Cohort.
In this study, we included all patients enrolled between 1st November 2006 and 30th June 2011. Patients who returned their questionnaires and those who did not were compared according to age, sex, socio-demographic, clinical and psychosocial characteristics. Odds ratio for non-response to initial and one-year follow-up were calculated using logistic regression. Missing values of socio-demographic and psychosocial variables were imputed using multivariate chained equations before assessing non-response to follow-up questionnaires.
Among 1945 IBD patients who received the inclusion questionnaire, 340 (17.2%) did not respond (18.8% Crohn's disease [CD], 15.6% ulcerative colitis [UC]). In CD patients, complicated disease (penetrating disease behaviour [OR=2.95; p< 0.001], perianal involvement [OR=1.74; p=0.003]) was the main risk factor of non-response, whereas longer disease duration ( >16 years) prevented UC and CD patients from being non-responders. Out of 1605 patients who received the first follow-up patient questionnaire, 323 (24.0%) did not respond (26.0% CD, 21.4% UC). Main risk factors of non-response to the follow-up questionnaire among CD patients were complicated disease (penetrating [OR=3.72; p<0.001], stricturing [OR=2.43; p<0.001], perianal involvement [OR=2.72; p<0.001]), mild or moderate depression (OR=2.27; p<0.001 resp. OR=1.83; p=0.048), or recent disease activity (OR=1.58; p=0.015). Being aged more than 30 (OR=0.45; p<0.001), taking immunomodulator (OR=0.64; p=0.007) or biological therapy (OR=0.47; p<0.001), having already had surgery and being higher educated prevented CD patients from being non-responders to the follow-up questionnaire. History of resection surgery (OR=0.35; p=0.019), taking immunomodulator therapy (OR=0.44; p<0.001), being married (OR=0.58; p=0.005) or having a positive social support (OR=0.97; p=0.040) were preventive factors for non-response to follow-up among UC patients.
Our study showed that characteristics of non-responders highly differed between UC and CD. The risk of non-response to repetitive solicitations (longitudinal versus transversal study) seemed to decrease with age. Assessing non-respondents' characteristics is important to document potential bias in longitudinal studies.