Radiological Findings
To define clinically relevant categories for describing X-ray findings, we conducted a rapid literature review in PubMed, Web of Science, Cochrane, and Scopus databases. Medical subject heading terms (MESH-term) were used for the search, and only articles in Spanish or English were included. Three pairs of authors carefully reviewed all selected publications.
Based on the insights gathered from the literature review, a group consisting of pediatricians, epidemiologists, and radiologists collaborated to develop patterns of radiological findings (as depicted in Figure 2). The categories for these radiological findings were established using the definitions provided in the Glossary of terms for chest imaging suggested by the Fleischner Society and the International Expert Consensus Statement on Chest Imaging in Pediatric COVID-19 Patient Management .
[INSERT FIGURE 2 ABOUT HERE]
A comprehensive database was constructed for the study, and the radiologists’ performed evaluations on various aspects of the chest X-rays. Specifically, they assessed the presence or absence of interstitial or focal opacities, the distribution and affection of the lungs, as well as the presence of any complications related to COVID-19, such as pleural effusion or consolidations.
The initial group consisted of four radiologists with varying clinical experience. Their experience as radiologists ranged from 8 to 22 years, while their experience as pediatric radiologists ranged from 5 to 22 years. To ensure an unbiased evaluation, this group was divided into two subgroups, and half of the X-ray images were randomly assigned to each subgroup for assessment. In addition to the initial group, an additional radiologist with 30 years of experience independently reviewed all the X-ray images and was considered the gold standard for the study. This expert’s evaluation served as a reference for comparison with the assessments made by the initial group of radiologists.
In order to maintain consistency and establish a standardized interpretation, the images were categorized according to their level of quality, dividing them into three groups: ”excellent”, ”optimal” and ”suboptimal” quality. These categories were defined based on previously agreed criteria supported by the literature, specifically the European Commission guidelines on quality criteria for diagnostic radiographic image . In performing the classification, several technical aspects were considered, including image alignment and rotation, adequate penetration, correct anatomical coverage and the presence of motion artifacts. A chest radiograph that complied with all the technical aspects was classified as ”excellent quality”; if it showed an alteration in at least one of the characteristics, it was classified as ”optimal”; and if it showed more than two alterations, it was designated as ”suboptimal”.
This standardization ensured that the quality of the images did not impact the accuracy of the radiological interpretations. Here are some examples (Figure 3) of X-ray classifications to provide visual representations of the radiological findings and classifications.
[INSERT FIGURE 3 ABOUT HERE]