Diagnostic accuracy of fine needle aspiration biopsy for detection of malignancy in pediatric thyroid nodules: protocol for a systematic review and meta-analysis

BACKGROUND: Fine needle aspiration biopsy (FNAB) is an accurate test commonly used to determine whether thyroid nodules are malignant in adults. However, less is known about its diagnostic accuracy for this purpose in children, where conduct of FNAB is less frequent, more technically challenging, and pre-test probabilities of malignancy are often higher. The purpose of this systematic review is to evaluate the diagnostic accuracy of FNAB for the detection of malignancy in pediatric thyroid nodules. METHODS: We will search electronic bibliographic databases (MEDLINE, EMBASE, the Cochrane Library, and Evidence-Based Medicine) from their date of inception, reference lists of included articles, proceedings from relevant conferences, and the table of contents of the Journal of Pediatric Surgery (January 2007–present). Two reviewers will independently screen titles and abstracts and identify diagnostic accuracy studies involving FNAB of the thyroid in children. We will include studies comparing FNAB to a reference standard of surgical histopathology or clinical follow-up for detection of malignancy in pediatric thyroid nodules. Two investigators will independently extract data and assess risk of bias using the Quality of Diagnostic Accuracy Studies-II tool. Pooled estimates of sensitivity, specificity, and positive and negative likelihood ratios will be calculated using bivariate random-effects and hierarchical summary receiver operating characteristic models. In the presence of between-study heterogeneity, we will conduct stratified meta-analyses and meta-regression to determine whether diagnostic accuracy estimates vary by country of origin, use of ultrasound guidance during FNAB, qualifications of the individuals performing/interpreting FNAB, adherence to the Bethesda criteria for cytology classification, length of clinical follow-up, timing of data collection, patient selection methods, and presence of verification bias. DISCUSSION: This meta-analysis will determine the diagnostic accuracy of FNAB for detection of malignancy in pediatric thyroid nodules and explore whether heterogeneity observed across studies may be explained by variations in patient population, FNAB technique or interpretation, and/or study-level risks of bias. This will be the first study to determine the accuracy of Bethesda cytological classification levels of FNAB (benign, atypical, follicular, suspicious, malignant). We expect that our results will help in guiding clinical decision-making in children with thyroid nodules. SYSTEMATIC REVIEW REGISTRATION: PROSPERO No. CRD42014007140