2017 Sec 1 Green Book

Otolaryngology–Head and Neck Surgery 154(5)

PotenƟally eligible parƟcipants (n = 257)

Excluded (n = 0)

Eligible parƟcipants (n = 257)

No FNAB (n = 0)

FNAB (n = 338)

Non-thyroid FNAB non-diagnosƟc (n = 20)

Non-thyroid FNAB posiƟve (n = 57)

Thyroid FNAB posiƟve (n = 35)

Non-thyroid FNAB negaƟve (n = 133)

Thyroid FNAB negaƟve (n = 83)

Thyroid FNAB non- diagnosƟc (n = 10)

Histopathology (n = 43)

Histopathology (n = 34)

Histopathology (n = 6)

Histopathology (n = 1)

Histopathology (n = 8)

Histopathology (n = 5)

HP posiƟve: 40 HP negaƟve: 3

HP posiƟve: 1 HP negaƟve: 7

HP posiƟve: 2 HP negaƟve: 4

HP posiƟve: 32 HP negaƟve: 2

HP posiƟve: 4 HP negaƟve: 2

HP posiƟve: 0 HP negaƟve: 1

No Histopathology (n = 1) Lymphoma (n = 1) No Histopathology (n = 14) ATB (n = 8) LVM (n = 4) Neck abscess (n =2) Figure 1. STARD flowchart of enrolled patients and FNAB. ATB, atypical tuberculosis; FNAB, fine-needle aspiration biopsy; HP, histo- pathology; LVM, lymphovenous malformation.

between the level of anesthesia groups. This method was used to account for the dependency of FNAB across encounters within patients. 17 The 2015 Standards for Reporting of Diagnostic Accuracy (STARD) guideline for reporting diagnostic accuracy studies was used. 18 Results Over the 7-year study period, 257 patients who underwent at least 1 FNAB were identified. By having multiple FNABs at 1 or multiple visits, there were 338 total FNABs, with 284 specific encounters ( Figure 1 ). Baseline charac- teristics of the patients are shown in Table 1 . The 169 patients in the nonthyroid group were younger (mean age, 7.1 vs 13.6 years, P \ .001) and had a lower percentage of females (38.5% vs 77.3%, P \ .001) than the 88 patients in the thyroid group. The mean (SD) size of the 189 masses with prebiopsy imaging was 2.4 (1.73) cm (range, 0.3-12.5 cm). The average time between FNAB and surgery was 73 days (range, 0-1183 days). The volume of FNABs per- formed per month steadily increased from 3.2 in 2007 to 7.2 in 2014 with no seasonal variation. Diagnostic Accuracy The overall pathologically confirmed sensitivity and specifi- city were 93.5% and 64.3%, respectively ( Table 2 ). If clini- cally and pathologically confirmed results are combined, sensitivity was 94.6% and specificity was 97.7% ( Table 2 ).

When rapid on-site interpretation yielded a preliminary diagnosis, it correlated with final cytopathology 99.1% of the time (n = 106) and final surgical histopathology 93.3% of the time (n = 30). The false positives and false negatives are presented in Table 3 . The clinical nonthyroid HNM sensitivity and specificity were 98.2% and 97.7%, respectively, representing 190 total FNABs ( Table 2 ). There were an additional 20 nondiagnos- tic specimens (9.5%). Table 4 presents the final diagnoses, the most common of which was benign reactive lymphade- nopathy (BLN), present in 99 patients. Malignancy was detected in 6 of the 169 patients (3.6%), including a meta- static medullary thyroid carcinoma that was not confirmed by thyroid FNAB prior to thyroidectomy. There were 118 thyroid FNABs, with a clinical sensitiv- ity of 88.6% and specificity of 97.6% ( Table 2 ). There were 10 nondiagnostic results (7.8%). The most common diagnosis was benign colloid nodule, found in 31 children. Fifteen malignancies were detected in the 88 patients under- going thyroid FNABs (17.0%), most commonly papillary thyroid carcinoma. The rates of malignancy by are pre- sented TBSRTC 4 category in Table 5 with comparisons to a large pediatric thyroid FNAB series. 19 The overall FNAB nondiagnostic rate was 8.9% (n = 30). All 10 of the thyroid nondiagnostic results were performed with image guidance and under general anesthesia (GA) or sedation. Of the 20 nonthyroid nondiagnostic results, there

192