2017 Sec 1 Green Book

TABLE IV. Radiographic Diagnoses of Patients Without Evidence of Peritonsillar Abscesses.

TABLE V. Multivariate Analysis Assessing Correlation Between Age, Sex, Otolaryngologist Clinical Diagnosis and US Finding With Presence/Absence of Peritonsillar Abscess.

Diagnosis

No.

Standard Error

Lower 95%

Upper 95%

Coefficients

t Statistic P Value

Unilateral tonsillitis Bilateral tonsillitis

11

9 2 1 1 2

2 0.011 0.028

Age Sex

0.009 0.009 0.884 .382

Normal tonsils

2 0.030 0.101 2 0.235 0.174 US finding 0.399 0.151 2.637 .012 0.092 0.705 ENT dx 0.229 0.144 1.589 .120 2 0.062 0.520 2 0.300 .765

Parapharyngeal phlegmon Reactive lymphadenopathy

Lymphadenitis

x 5 diagnosis; ENT 5 ear, nose, and throat; US 5 ultrasound.

either case. Interestingly, one of these patients had a CT scan from an outside facility suggestive of PTA. The remaining 24 patients were exclusively managed medi- cally without any treatment failures. After a negative needle aspiration on initial visit 2 days prior, one patient returned to the emergency room. This patient was not determined to have a PTA at the time of representation and was again managed medically without failure. Another patient was readmitted following operative inci- sion and drainage of a clinical PTA due to reaccumula- tion of infection. We compared the ages of patients undergoing drain- age and those not undergoing procedures. After perform- ing a Mann-Whitney test on these data, we discovered the P value comparing the ages was not statistically sig- nificant. The box plot in Figure 1 represents these data. Furthermore, we elected to perform a multivariate anal- ysis using age, sex, otolaryngologist clinical diagnosis, and US findings as independent variables to see if any correlated statistically with the presence or absence of purulence. These data can be viewed in Table V. US finding was the only independent variable found to be correlative with the presence or absence of PTA ( P < .05). The sensitivity and specificity (with 95% confidence intervals) of transcervical ultrasonography in the diag- nosis of pediatric PTA are 100% (86.8%-100%) and 76.5% (58.9%-89.2%), respectively. The positive and negative

predictive values (with 95% confidence intervals) are 52.9% (27.9%-77.1%) and 100% (66.4%-100%), respec- tively. When a Fisher exact test was performed, the P value was statistically significant ( P < .01), indicating an important correlation between a negative US and patients who were able to be managed medically. Exam- ples of US images obtained in these patients are present in Figures 2–4. DISCUSSION Peritonsillar abscesses are common in the pediatric population; however, diagnosis in this cohort is challeng- ing due to limitations imposed by smaller oropharyngeal anatomy and the potential for uncooperative patients. As a result of these clinical hurdles, CT is frequently uti- lized as a diagnostic tool. One study of pediatric emer- gency room visits reported that CT was ordered in 65% of patients in whom a PTA was suspected. 15 Other groups have reported algorithms that include exams of the oropharynx under anesthesia or trials of intravenous antibiotics followed by operative intervention. 16,17 This study aimed to investigate the role of transcervical US in diagnosing pediatric PTA.

Fig. 1. Box plot comparing the age of patients undergoing drain- age procedures and those not undergoing drainage procedures. There was no statistical difference in age of patients in the two groups.

Fig. 2. Example of a normal transcervical ultrasound of the tonsil and peritonsillar region. The tonsil has a striated appearance and is outlined by the black arrow heads. SMG 5 submandibular gland; T 5 tongue.

Laryngoscope 125: December 2015

Fordham et al.: Transcervical US in Pediatric PTA

64