Michael Baad1, Jorge Delgado2, Jillian S Dayneka3, Sudha A Anupindi3, Janet R Reid3. 1. Department of Radiology, Weill Cornell Medicine, 525 E 68th Street, Starr 8A-37, New York, NY, 10065, USA. mib9181@med.cornell.edu. 2. Department of Radiology, Massachusetts General Hospital, 55 Fruit Street, Boston, MA, 02114, USA. 3. Department of Radiology, The Children's Hospital of Philadelphia, University of Pennsylvania Perelman School of Medicine, 3401 Civic Center Blvd, 3NW13, Philadelphia, PA, 19104, USA.
Abstract
PURPOSE: We aim to evaluate the diagnostic performance and relationship between clinical characteristics, imaging findings, and final diagnosis for the neonatal contrast enema (CE). METHODS: Retrospective 10-year review of all neonatal CEs including imaging findings, clinical information, indication, and final diagnosis from discharge summaries, surgical reports, and pathology (reference standard). Two blinded pediatric radiologists reinterpreted 366 CEs for obstruction, microcolon, rectosigmoid index (RSI), serrations, meconium, ileal cut-off, transition zone, diagnosis, and level of confidence. CE diagnostic performance was calculated versus reference standard. RESULTS: Diagnoses included Hirschsprung disease (HD) (15.8%), small left colon syndrome (14.8%), small intestinal atresia/colonic atresia (SIA/CA) (12.6%), meconium ileus (MI) (4.4%), and normal (48.9%). CE had a moderate specificity (87.7%) and low sensitivity (65.5%) for HD; abnormal RSI and serrations showed high specificities (90.3%, 97.4%) but low sensitivities (46.6%, 17.2%). CE showed high specificity (97.4%) and low sensitivity (56.3%) for MI blinded to cystic fibrosis status. Microcolon was specific (96.6%) but not sensitive (68.8%) for MI. CE showed highest PPV (73.1%) (specificity 95.6%, sensitivity 82.6%) for SIA/CA. Microcolon with an abrupt cut-off was specific (99.1%) but not sensitive (41.3%) for atresias. CONCLUSION: Neonatal CE demonstrates high specificities and low to moderate sensitivities across all diagnoses, with lowest performance in HD. CLINICAL TRIAL REGISTRATION: None.
PURPOSE: We aim to evaluate the diagnostic performance and relationship between clinical characteristics, imaging findings, and final diagnosis for the neonatal contrast enema (CE). METHODS: Retrospective 10-year review of all neonatal CEs including imaging findings, clinical information, indication, and final diagnosis from discharge summaries, surgical reports, and pathology (reference standard). Two blinded pediatric radiologists reinterpreted 366 CEs for obstruction, microcolon, rectosigmoid index (RSI), serrations, meconium, ileal cut-off, transition zone, diagnosis, and level of confidence. CE diagnostic performance was calculated versus reference standard. RESULTS: Diagnoses included Hirschsprung disease (HD) (15.8%), small left colon syndrome (14.8%), small intestinal atresia/colonic atresia (SIA/CA) (12.6%), meconium ileus (MI) (4.4%), and normal (48.9%). CE had a moderate specificity (87.7%) and low sensitivity (65.5%) for HD; abnormal RSI and serrations showed high specificities (90.3%, 97.4%) but low sensitivities (46.6%, 17.2%). CE showed high specificity (97.4%) and low sensitivity (56.3%) for MI blinded to cystic fibrosis status. Microcolon was specific (96.6%) but not sensitive (68.8%) for MI. CE showed highest PPV (73.1%) (specificity 95.6%, sensitivity 82.6%) for SIA/CA. Microcolon with an abrupt cut-off was specific (99.1%) but not sensitive (41.3%) for atresias. CONCLUSION: Neonatal CE demonstrates high specificities and low to moderate sensitivities across all diagnoses, with lowest performance in HD. CLINICAL TRIAL REGISTRATION: None.