The purpose of this study was to pilot test a screening tool that identifies children at high-risk for celiac disease which could be administered at the public school level by the school nurse. This instrument validation was designed as a case-control study and had three different sample groups (N=138): one case and two levels of control. The case group was 78 children with diagnosed celiac disease whose parents were recruited from mailing lists made available by their local celiac support groups. The control group was divided into two groups: 45 non-case children without diagnosed celiac disease whose parents were recruited from a published community phonebook; and 15 non-case children known to have tested negative for celiac disease by either blood test or biopsy, whose parents were purposively recruited from three different celiac support groups. The instrument was an original survey containing 16 questions related to the common clinical symptoms and conditions of celiac disease.

Results: the case group mean questionnaire score was 43% higher than the control group (t=-12.5, p=.001). An ANCOVA of independent variables for celiac disease identified four statistically significant correlations (r≥.4, p=.001): anemia with anxiety; anemia with short stature; stomach pain with intestinal pain; and older children and aches. A factor analysis of the case group identified six subsets of health symptoms and conditions, that when present together, increase a child's risk for celiac disease. The accuracy measurements indicate that the questionnaire has a high degree of predictive validity for celiac disease (82% sensitivity, 94% specificity, 94% PV+, 80% PV-). The reliability test-retest score was relatively high (r=.87).

The pilot test concluded that the questionnaire met the nine recommended policy criteria by the American Academy of Pediatrics for school screening programs, including validity and reliability.