2017 Sec 1 Green Book

Fig. 3. Identical images of a right peritonsillar abscess (PTA) with the image on the right labeled. The PTA is outlined by a dashed white line. The tonsil (white arrow heads) and submandibular gland (SMG) are once again identified.

been found on some of these patients. With this is mind, we may have overcalled the number of false positives. Our study design did not mandate exploration of every peritonsillar space or require CT on every patient. For statistical integrity, however, children with positive ultrasounds managed medically successfully were not discarded, but rather counted as false positives. Based on these data, we believe that smaller PTAs may resolve with medical therapy alone; therefore, our study results likely represent an underestimation of ultrasound’s spec- ificity for PTA. When comparing the sizes of the measured abscesses in the true- and false-positive groups, a statis- tically significant difference was found, with the false positives measuring smaller. Although US is capable of visualizing small PTAs in pediatric patients, the spec- trum of peritonsillar and intratonsillar findings may lead a radiologist to overcall an abscess. This conclusion assumes that all of the false positives indeed did not have purulence. If purulent cavities were in fact present on some of these patients, then smaller abscesses either respond to medical management alone and/or are diffi- cult to locate during procedural interventions. Based on these findings, we developed an algorithm for PTA management (Fig. 5), which is now utilized at

The sensitivity and specificity of US in this study were 100% and 76.5%, respectively. Our data are similar to that of other studies. Araujo Filho et al. reported sen- sitivity and specificity as 80% and 92.8%, respectively, for transcervical US, primarily in adult patients. 13 The differences in anatomy and amount of subcutaneous tis- sue between the adult and pediatric populations may account for some of the variation between these statis- tics; however, both studies highlight the utility of this modality in evaluating this infectious process. In this study, it is important to note that of the patients who had a negative US, none failed medical management. US is a very sensitive tool for identifying fluid col- lections within tissue planes; therefore, we were not sur- prised to have a number of false positives, as oftentimes small fluid collections on CT or US are reported. Several patients who were successfully managed medically were diagnosed with PTA both on physical exam and US. These patients, who responded promptly to initial medi- cal therapy prior to procedural intervention, or whose parents opted for a more conservative medical treatment in lieu of surgical drainage, were followed clinically. Given that the peritonsillar regions of all exclusively medically managed children were never opened, it is possible that small collections of purulence may have

Fig. 4. Ultrasound and computed tomography (CT) images from the same patient. The far left and center images are identical, the center image being marked to identify the structures. The irregularly marginated hypoechoic region (outlined by white dashed line) is consistent with a left peritonsillar abscess. Again the tonsil (white arrow heads) and submandibular gland (SMG) are labeled for reference. The CT image (far right) shows the corresponding abscess (black arrow).

Laryngoscope 125: December 2015

Fordham et al.: Transcervical US in Pediatric PTA

65