Magnetic resonance imaging findings in pediatric neck infections-a comparison with adult patients

Abstract

<p><br></p><p>Background</p><p>Differences in the functioning of the immune system and the anatomical proportions of the neck between children and adults lead to different manifestations of deep neck infections. Magnetic resonance imaging (MRI) may serve as an alternative to computed tomography (CT) as the primary imaging modality. <br></p><p>Objective <br></p><p>To study characteristic MRI findings and the diagnostic accuracy of MRI in pediatric deep neck infections.</p><p>Materials and methods</p><p>We retrospectively studied a cohort of pediatric patients who underwent a neck 3-tesla MRI study over a five-year period. Inclusion criteria were: 1) emergency MRI findings indicating an infection, 2) infection as the final clinical diagnosis, 3) diagnostic image quality verified by the radiologist reading the study and 4) age under 18 years. Patient record data, including surgery reports, were compared with the MRI findings.</p><p>Results</p><p>Data of 45 children were included and analysed. Compared to adults, children had a higher incidence of retropharyngeal infection and lymphadenitis, and a lower incidence of peritonsillar/parapharyngeal infection. MRI showed evidence of an abscess in 34 children. Of these 34 patients, 24 underwent surgery, which confirmed an abscess in 21 but no abscess in three patients. In addition, three patients underwent surgery without MRI evidence of abscess, and an abscess was found in one of these cases. The measures of diagnostic accuracy among the children were sensitivity 0.96, specificity 0.77, positive predictive value 0.89, negative predictive value 0.91 and accuracy 0.89. Compared with adults, children had lower C-reactive protein, but a similar proportion of them had an abscess, and abscess size and rate of surgery were similar.</p><p>Conclusion</p><p>Despite the differences in the infection foci, emergency MRI in children had equal diagnostic accuracy to that in adults.</p>