Thirteen prospective studies were included in the review: six in adults (n reported as 670, but table of characteristics totalled 660); and seven in children and teenagers (n=371). Studies of children and teenagers were of greater quality than studies of adults. Three adult studies had selection bias and four adult studies did not use ileal intubation or histology was not undertaken. Verification bias was evident in two adult studies. Three studies of children and teenagers and two studies of adults did not provide reasons for withdrawal. Three studies were not blinded to index test results.
For adults (six studies), pooled sensitivity was 0.93 (95% CI 0.85 to 0.97) and pooled specificity was 0.96 (95% CI 0.79 to 0.99). For children and teenagers (seven studies), pooled sensitivity was 0.92 (95% CI 0.84 to 0.96) and pooled specificity was 0.76 (95% CI 0.62 to 0.86). There was a statistically significant difference between the lower specificity of children and teenagers and that of adults (p=0.048).
Inflammatory bowel disease was confirmed in 32% (n=215) adults and 61% (n=226) of children and teenagers (32% and 61% pre-test probability). Screening by faecal calprotectin levels had a significant effect on both age groups. An abnormal faecal calprotectin test result in adults increased the probability of inflammatory bowel disease to 91% (95% CI 77% to 97%) and a normal test result reduced the probability to 3% (95% CI 1% to 11%). In children, an abnormal test result increased the probability to 86% (95% CI 78% to 92%) and a normal test result reduced the probability to 15% (95% CI 7% to 28%).
Data on subgroup analyses and further investigations were not presented in the review, but the authors stated that subgroup analyses by cut-off value and further investigations showed no difference between the age groups.
The authors did not present formal assessment of statistical heterogeneity, but reported considerable heterogeneity among studies.