Y Patenaude1, C Blais, C P Leduc. 1. Department of Radiology, Hôpital Ste-Justine, Montreal, Québec, Canada.
Abstract
RATIONALE AND OBJECTIVES: The World Health Organization recommends that the routine investigation of children with suspected pneumonia should not include lateral chest x-rays. However, the reliability of the frontal view alone in the diagnosis of pulmonary opacities has not been reported. METHODS: The authors studied prospectively 373 consecutive chest x-rays of children, ranging in age from 0 to 17 years, who were examined for suspected pneumonia. One radiologist interpreted the frontal view, and the diagnoses were compared with those of three radiologists who interpreted both frontal and lateral x-rays. RESULTS: William's index for rater reliability was 0.98 for all ages (95% confidence interval: 0.94, 1.00). Thus, a radiologist using only a frontal view would agree with members of an independent group using both frontal and lateral views as often as an isolated member of that group would agree with the other group members. CONCLUSIONS: Detecting a definite pulmonary opacity on the frontal view alone predicts its presence on frontal-lateral views. However, the interpretation of bronchial thickening and peribronchial alveolar confluences as opacities is a problem remaining on either frontal or frontal-lateral views.
RATIONALE AND OBJECTIVES: The World Health Organization recommends that the routine investigation of children with suspected pneumonia should not include lateral chest x-rays. However, the reliability of the frontal view alone in the diagnosis of pulmonary opacities has not been reported. METHODS: The authors studied prospectively 373 consecutive chest x-rays of children, ranging in age from 0 to 17 years, who were examined for suspected pneumonia. One radiologist interpreted the frontal view, and the diagnoses were compared with those of three radiologists who interpreted both frontal and lateral x-rays. RESULTS: William's index for rater reliability was 0.98 for all ages (95% confidence interval: 0.94, 1.00). Thus, a radiologist using only a frontal view would agree with members of an independent group using both frontal and lateral views as often as an isolated member of that group would agree with the other group members. CONCLUSIONS: Detecting a definite pulmonary opacity on the frontal view alone predicts its presence on frontal-lateral views. However, the interpretation of bronchial thickening and peribronchial alveolar confluences as opacities is a problem remaining on either frontal or frontal-lateral views.
Authors: Cynthia K Rigsby; Janet L Strife; Neil D Johnson; Harry D Atherton; William Pommersheim; Uma R Kotagal Journal: Pediatr Radiol Date: 2004-01-22