*2.4. Statistical Analysis*

Mean and standard deviation or quartiles were used for descriptive purposes, as appropriate. The diagnostic ability of the TIRADS systems in distinguishing thyroid nodules that required or not FNA was evaluated using cytology as reference standard (<TIR3A/III vs. ≥TIR3A/III) in all 480 nodules. The results of the TIRADS classifications and cytology were also compared to the histopathological/follow-up findings on 401 nodules. Sensitivity, specificity, and positive and negative predictive values (PPV and NPV) were calculated alongside their 95% confidence intervals (CI). The McNemar test was considered for the comparison of the performances (two-sided test, α = 0.05). A decision tree was applied to the cytological classes and to the single TIRADS ecographic components of each thyroid nodule to explore whether and which US features were relevant in malignancy detection. The risk of malignancy was calculated as the rate of prevalence. All of the statistical analyses were performed using the open-source R software v.3.6.0 (R Foundation for Statistical Computing, Vienna, Austria).
