Diagnostic value of the chest radiograph in asymptomatic neonates with a cardiac murmur

AIM: To establish the diagnostic accuracy and diagnostic usefulness of the chest radiograph in asymptomatic neonates with cardiac murmurs. PATIENTS AND METHODS: The chest radiographs of 68 asymptomatic neonates with cardiac murmurs were analysed retrospectively. The radiographs were anonymized and then evaluated for the presence or absence of cardiac disease by six radiologists, three who regularly interpret neonatal chest radiographs and three who do so infrequently. The eventual diagnosis for each neonate and the impact of the chest radiograph and original report on patient management were established by review of the clinical case notes. RESULTS: The results for each observer were expressed in 2 x 2 contingency tables and statistical analysis was performed using Fisher's exact test. The radiologists who were experienced in reporting neonatal chest radiographs achieved statistically significant results (P = 0.003, P = 0.002 andP = 0.007) compared with those who were less experienced (P = 0.13, P = 0.16 andP = 0.09). Review of the case notes established that the chest radiograph and original report did not influence clinical management in any of the 68 cases studied. CONCLUSIONS: Radiologists who frequently report neonatal chest radiographs achieve high accuracy in differentiating cardiac from non-cardiac disease. However, inaccuracies are unavoidable as radiological evidence of cardiac disease is often not present. A false-positive result could cause undue anxiety while a false-negative report could result in the omission of further investigations. Furthermore, a chest radiograph is unlikely to provide the definitive diagnosis. Chest radiographs did not appear to influence patient management in this study and cannot be recommended in the initial evaluation of the asymptomatic neonate with a heart murmur. Oeppen, R.S. et al. (2002)