Chest radiography and the evaluation of the neonate for congenital heart disease

• Published in: Pediatric cardiology 2005-07, Vol.26 (4), p.367-372
• Main Authors: Fonseca, B, Chang, R-K, Senac, M, Knight, G, Sklansky, M S
• Format: Article
• Language: English
• Subjects: Diagnosis, Differential; Echocardiography; Heart Failure - diagnostic imaging; Humans; Infant, Newborn; Observer Variation; Prognosis; Radiography, Thoracic; Retrospective Studies; Sensitivity and Specificity
• ISSN: 0172-0643; 1432-1971
• DOI: 10.1007/s00246-005-8649-z

We assessed the utility of the chest film as a screening tool for term and preterm neonates with suspected structural heart disease. Three independent observers, blinded to patient diagnosis, retrospectively evaluated the initial chest film for each of 128 consecutive neonates with suspected heart disease (gestational age, 23-42 weeks; birth weight, 500-4,621 g) who had undergone both chest radiography and echocardiography. These evaluations were subsequently compared with the results of the respective echocardiograms, read by board-certified pediatric cardiologists blinded to the chest film and chest film interpretation. Kappa statistics demonstrated moderate correlation (0.42-0.48) among different observers in their interpretation of the chest films but poor correlation (0.15-0.34) between chest radiography and echocardiography. The chest film had a low sensitivity for structural heart disease (26-59%), with a negative predictive value of 46-52%. Among neonates less than 2 kg or younger than 35 weeks of gestation, the chest film had still lower sensitivity for detecting heart disease. Despite agreement among observers in chest film interpretation, the chest film does not function as a screening test for neonates with suspected heart disease, particularly in small or premature neonates. In neonates with suspected heart disease, echocardiography should be considered, even in patients with chest films that do not suggest congenital heart disease.