Agreement and reproducibility of radiological signs in NEC using The Duke Abdominal Assessment Scale (DAAS)

PURPOSE: Necrotizing enterocolitis (NEC) is associated with high morbidity and mortality. Abdominal radiography is currently an imaging modality of choice in NEC. Recently, a numeric scale of radiological signs in NEC—The Duke Abdominal Assessment (DAAS) was introduced. The aim of this study was to measure the intra- and inter-observer agreement on the radiological signs of NEC according to DAAS to access the feasibility of this scale. MATERIALS AND METHODS: We have retrospectively analyzed 87 radiographs performed in a group of 43 high-risk neonates with suspected NEC. Radiographs were assessed by 6 independent observers: two pediatric radiologists, two radiology residents, and two neonatologists. Data were analyzed using κ statistics as a measure of intra- and inter-observer agreement. RESULTS: Fair-to-good intra-observer agreement was noted for all but one of observers. However, with the wide range in κ values, we found only fair inter-observer agreement detecting signs of NEC according to DAAS. There was a higher intra-group agreement in radiology practitioners, with the highest among experienced pediatric radiologists. CONCLUSION: However, with high observer variability in interpretation of all radiologic signs, we did not confirm that Duke Abdominal Assessment Scale could reliable facilitate reporting of abdominal radiographic findings in neonates with suspected NEC. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1007/s00383-016-4022-y) contains supplementary material, which is available to authorized users.