Abstract
AIM: Newborn infants with critical aortic arch obstruction are often undiagnosed at discharge, despite screening. This study investigated if adding the perfusion index improved early detection. METHODS: We retrospectively studied 38 newborn infants with critical aortic arch obstruction, who were routinely screened in 2014-2019 by 13 Swedish hospitals using pulse oximetry and the perfusion index. They were identified through surgery records and national mortality databases. The controls were 512 healthy newborn infants from one hospital. Optimal perfusion index cut-offs were determined using frequency distribution analysis. RESULTS: The groups had similar median gestational ages and birth weight. No infants with critical aortic arch obstruction were diagnosed just because of a positive perfusion index result. However, the right-hand perfusion index was significantly higher in the cases than in controls (p < 0.001). A perfusion index of > 3%, or positive pulse oximetry or positive physical examination, yielded 76% sensitivity and 85% specificity, with an area under the receiver operating characteristic curve of 0.81 (range 0.73-0.89, p < 0.0001). Pulse oximetry and just neonatal physical examinations had a lower sensitivity (45%, p = 0.009). CONCLUSION: A high perfusion index in the right hand enhanced critical aortic arch obstruction screening and repeated measurements should be explored to minimise false positives.