[Should all patients with thyroid nodules > or = 1 cm undergo fine-needle aspiration biopsy?]

The prevalence of thyroid nodules > or = 1 cm is high in a previously iodine-deficient area. Under the hypothesis, that all patients with such nodules undergo fine-needle aspiration biopsy (FNAB) and that sensitivity and specificity of cytology are calculated with 85%, the positive predictive value of pathologic cytologic finding will reach 1.5% only according to Bayes-theorem. This is clinically unacceptable, as resection will be the consequence in all cases with suspect cytology. Even implementation of a second, independent test (e. g. moleculargenetic testing of thyreocytes, sensitivity to detect mutation 50%, specificity 95%) and application of sequential Bayes-theorem the positive predictive value of combined pathologic findings will increase to 13% only. Nevertheless, 58% out of all thyroid cancer remain undetected by such a sequential algorithm. As a consequence , pre-selection of thyroid nodules for FNAB is required to increase the pretest-probability to at least 5-10%. A combination of sonographic criteria and scintigraphy, even in patients with normal TSH-levels, is suited to selected thyroid nodules for FNAB.