The lateral chest radiograph

We are very pleased that the important issue of lateral chest x-ray has been brought to our attention through Dr Lynch and his colleagues' study. Because this is so important, we asked a paediatric radiologist to provide his point of view. As you will see, the issue is unquestionably important but not closed! Noni E MacDonald MC FRCPC Editor-In-Chief In this issue of Paediatrics & Child Health, Lynch et al (pages 566–568) have attempted to perform a comprehensive review of the literature to date concerning the utility of the lateral chest radiograph in the evaluation of the child with respiratory symptoms in the setting of suspected ‘pneumonia’. The attempt is noble but there are significant difficulties with this approach that need to be considered. There are, apparently, only four articles that meet the authors' inclusion criteria. Definitive conclusions based on these four articles alone are difficult. The articles themselves have intrinsic problems with respect to their research methodologies. Lamme et al (1) presented a sample size of 179 children of which approximately 80% were normal. This sample problem was addressed in a letter to the editor, and subsequently answered by the authors who agreed that the applicability of their patient sample to other populations, where the incidence of abnormal radiographs is higher, may be limited. As well, they agreed with the criticism that the lateral view may provide information other than ‘pneumonia’, admitting that this was not ‘in the scope’ of their study. Furthermore, they excluded all children younger than one year of age or older than 10 years of age. Kennedy et al (2) stated that in approximately 2.2% of cases, the lateral view alone was diagnostic as ‘positive’ for pneumonia, yet in another 8.2% of cases, ‘additional information’ was obtained from the lateral view, although they claim that the additional information did not contribute to the child's therapeutic care. In the study by Kiekara et al (3), blinded methodology, where the frontal and lateral views were independently reviewed in a blinded fashion, is not described. As well, the authors themselves concluded that “lateral views are supplementary, yielding additional information about the extent and location of pneumonia”. Similarly, in the paper by Patenaude et al (4), the authors concluded that the more subtle findings such as bronchial wall thickening and ‘localized peribronchial confluence’ are difficult to appreciate and interpret, and that “it remains to be studied if the lateral view decreases this difficulty”. It is apparent that manipulation of statistics from studies performed by others can yield interesting results, but when the studies themselves are limited in scope or in their endpoints, there is little service to be had by the manipulation of those statistics. Lynch et al themselves admit that the papers included in their review have heterogeneous definitions of ‘pneumonia’ and heterogeneous age distributions. The problem of the utility of the lateral chest film is being tackled by those who are well meaning enough to attempt to decrease medical radiation and cost, both of which carry little significance to the individual child, yet are meaningful when we consider the vast number of these studies that are performed worldwide. The World Health Organization has recommended that, in the paediatric age range, if the frontal view is normal, then no further views are necessary (5). The determination of how the initial frontal view is to be reviewed before the decision to potentially proceed to a lateral view is made, is not addressed. On a worldwide basis, this approach will save radiation exposure and medical cost, allowing for the possibility that a small number of pneumonias will be missed. There are many problems in this area. The clinical and radiographic definitions of pneumonia can be variable, especially in children who do not manifest ‘classical’ signs. The sensitivity of the radiographic examination to detect pneumonia is heavily dependent on the clinical acumen and experience of the examining clinician who will provide a variable selection bias of patients to be examined. Alternate diagnoses or infectious complications may be appreciated with the addition of the lateral view. For example, Smuts et al (6) in a prospective review of 449 children, assessed for possible tuberculosis by frontal and lateral radiography and found a significantly higher appreciation rate of hilar adenopathy on the lateral view than on the frontal view. The level of training of those interpreting the images (which involves the level of experience and comfort of the reader with both the clinical and radiographic diagnosis of ‘pneumonia’) can be variable and may influence the true incidence of radiologically diagnosed ‘pneumonia’. Even the most experienced radiologists will freely admit that the ‘two view’ conventional chest radiograph is one of the most difficult diagnostic imaging procedures to interpret. It is no surprise that this topic has not been addressed adequately in the literature. An appropriate study of this topic will require parental consent to a study in which the child will not necessarily receive what is currently accepted as the standard of care. The methodology must include a very large series with a significant proportion of abnormal cases with consolidative pneumonias of varying severity. This in itself is difficult because most lower respiratory tract infections in children are viral in etiology. It must be performed prospectively, with all images reviewed in a blinded fashion. Physicians of multiple levels of training must be included to ascertain the influence of previous training and experience in the interpretation of paediatric chest films. A widely accepted definition of both clinical and radiographic criteria of pneumonia must be presented as a proper endpoint. Furthermore, care must be taken to include all other potentially important radiographic findings that may contribute to the care of the child. In other words, it is a quagmire.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.