Umeå universitets logga

Background Mass screening for celiac disease (CD) as a public health intervention is controversial. Before implementation, a suitable screening strategy should be outlined, and the acceptability of the screening scrutinized. Also, the benefits of early detection and possible negative consequences should be explored and compared. The overall aim of this thesis was to evaluate different strategies for finding 12-year-olds with undiagnosed CD in the general population, and to explore the experiences of those receiving the diagnosis in a mass screening.

Methods A school-based CD screening of 12-year-olds was conducted in five study sites across Sweden. Out of 10041 children who were invited, 7208 had a blood sample analyzed for CD-marker tissue transglutaminase of isotype IgA (tTG-IgA) and 7161 for total serum IgA (s-IgA). If the s-IgA value was low, tTG-IgG was also measured. Additional analysis of endomysial antibodies (EMA) was performed if borderline values of tTG were found. In total, 192 had elevated CD-markers, 184 underwent a small intestinal biopsy and 153 eventually had CD diagnosed. Before receiving knowledge about their CD status, children and their parents filled in questionnaires regarding symptoms and CD-associated conditions. Questionnaires were returned by 7054 children (98%) and 6294 parents (88%). Later, all adolescents who had been diagnosed with CD more than one year ago (n=145), and their parents, were invited to a mixed-method follow-up study in which they shared their experiences in questionnaires, written narratives and focus group discussions. In total, we have information on 117 (81%) of these adolescents, either from the adolescents themselves (n=101) and/or from their parent/s (n=125). Data were analyzed using a combination of descriptive and analytical quantitative and qualitative methodologies.

Results We found that information on symptoms and CD-associated conditions were poor predictors for finding undiagnosed CD in the study population. Questionnaire-based case-finding by asking for CD-associated symptoms and conditions would have identified 52 cases (38% of all cases) at a cost of blood-sampling 2282 children (37% of the study population). The tTG-IgA test had an excellent diagnostic accuracy with the area under the receiver operating characteristic curve of 0.988. If using the recommended cut-off for tTG-IgA (>5 U/mL) 151 had fulfilled biopsy criteria and 134 CD cases had been identified. The strategy of lowering the cut-off to tTG-IgA>4 U/mL, and adding the EMA analysis in those with tTG-IgA between 2-4 U/mL, identified another 17 cases (a 12% increase) at the cost of performing 32 additional biopsies. Measuring total s-IgA in 7161 children discovered only two additional cases at the cost of performing 5 additional biopsies. The positive predictive value of our screening strategy was 80%. 

Results from the follow-up study of the screening-detected CD cases illustrated that 54% reported health improvement after initiated treatment, but also that these health benefits had to be balanced against social sacrifices. We also found that although the screening-detected diagnosis was met with surprise and anxiety, the adolescents and their parents were grateful for being made aware of the diagnosis. A majority of parents (92%) welcomed a future screening, but both adolescents and parents suggested that it should be conducted earlier in life.

Conclusion Obtaining information on symptoms and CD-associated conditions was not a useful step in finding undiagnosed CD cases in a general population. The serological marker tTG-IgA, however, had excellent diagnostic accuracy also when lowering the cut-off. The diagnosis had varying impact on adolescents’ quality of life, and their perceived change in health had to be balanced against the social sacrifices resulting from the diagnosis. Overall, CD mass screening seemed acceptable to most of those who were diagnosed and their parents.