The diagnostic accuracy of serologic tests for celiac disease: a systematic review

Clinicians are increasingly utilizing noninvasive serologic tests for the diagnosis and screening of celiac disease (CD). The aim of this study was to conduct a systematic review of the diagnostic performance of serologic tests for the diagnosis and screening of CD. Standard systematic review methodology was used. A literature search was conducted in MEDLINE (1966 to October 2003) and EMBASE (1974 to December 2003) databases. A weighted mean of the sensitivity and specificity along with 95% confidence intervals and summary receiver operating characteristic (ROC) curves were calculated. The pooled specificity of endomyseal antibody (EMA)-monkey esophagus (ME) or EMA-human umbilical cord (HU) was close to 100% in adults and children. The pooled specificity of transglutaminase antibody (tTG)-guinea pig (GP) and tTG-human recombinant (HR) were between 95% and 99%. IgA-EMA-ME demonstrated sensitivities of 96% and 97% in children and adults, respectively. EMA-HU demonstrated a similar sensitivity of 97% in children but 90% in adults. The pooled sensitivity of tTG-GP in adults and children was 90% and 93%, respectively. The sensitivity of tTG-HR was 98% and 96%, respectively. The performance of antigliadin antibody was inferior to that of EMA and tTG. EMA and tTG offer high sensitivity and specificity. The sensitivity of these tests appears to be lower than reported when milder histologic grades are used to define CD (below 90%). If true, the nearly perfect negative predictive value of these tests would drop. The positive predictive value of these tests is likely lower than reported when the tests are applied in low-prevalence populations.