Meta-analysis: High pooled positive predictive value of the European Society of Paediatric Gastroenterology, Hepatology and Nutrition no-biopsy approach for coeliac disease testing in children.

Abstract

The European Society of Paediatric Gastroenterology, Hepatology and Nutrition established guidelines in 2012 for a no-biopsy approach to diagnose coeliac disease in children. This guideline required symptoms suggestive of coeliac disease, positive human leukocyte antigen (HLA) DQ2/DQ8 haplotypes, tissue transglutaminase type-2 immunoglobulin A antibody titre at levels greater than 10 times the upper limit of normal, and positive endomysial immune-globulin A antibody test. An updated 2020 guideline excluded the need for symptoms and positive HLA. To assess the pooled positive predictive value (PPV) of the no-biopsy approach with small bowel biopsy (SBB) data as the reference standard for comparison. Database searches (October 2023) provided data that we combined using a random-effects meta-analysis to provide a pooled PPV, representing the probability that a positive test result means that an individual truly has the condition. We included 23 studies. Study sample sizes totalled 23,769 but only 3007 children had comparative SBB. The proportion of coeliac disease confirmed by the no-biopsy approach and SBB ranged from 79.2% to 100%, with an overall pooled PPV of 97.4% (95% confidence interval 96.0, 98.6). Sensitivity analysis showed higher PPV for the criteria that included HLA (98.5% vs. 96.8%; p = 0.017). Both no-biopsy criteria exhibit high PPV when compared to the reference standard. These results provide a consistent message of accuracy and feasibility to inform change and improve outcomes.