RESUMO
OBJECTIVES: To evaluate the diagnostic performance of frozen section in thyroid nodules classified as follicular neoplasm. METHODS: A diagnostic test meta-analysis was designed. Studies that assessed frozen section in patients with thyroid nodules and a fine-needle aspiration biopsy result of Bethesda IV were selected. The outcomes measured were the number of false- and true-positive and -negative results. We used the Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) instrument for methodological quality assessment and a bivariate mixed-effects regression framework and a likelihood-based estimation of the exact binomial approach. RESULTS: Forty-six studies from 1991 to 2018 were included. Most studies had moderate methodological quality. The overall sensitivity and specificity were 43% (95% confidence internal [CI], 0.34-0.53) and 100% (95% CI, 0.99-1.00), respectively. The hierarchic summary receiver operating characteristic curve showed an area under the curve of 0.91 (95% CI, 0.80-0.97). CONCLUSIONS: Frozen section demonstrates moderate diagnostic performance in patients with follicular neoplasm, and its utility for making intraoperative decisions is limited. Its routine use should be discouraged.