| dc.description.abstract | A fine needle aspiration citology (FNAC) has allowed a better selection of candidates for
thyroidectomy providing a reduction in the number of operated patients and increased number of
malignant tumors surgically resected. However, there is a wide variation in results, which may
reflect differences in experience gained in its interpretation. It is believed that the efficiency of
the method is directly related to training professional. The critical parameter of the clinical
usefulness of a diagnostic test is its interobserver variability, and especially intraobserver. The
objectives of this study was to evaluate the influence of the examiner's experience in the
interpretation of cytological findings of thyroid and to correlate the cytologic findings issued by
pathologists with and without experience, with the histological examinations of surgical
specimens and to evaluate the influence of a broad definition of malignancy (BDM) instead of
using the narrow definition of malignancy (NDM) on the calculations of measures and validation
of diagnostic prediction. The results revealed consistent intraobserver reproducibility after the
lapse of six months, both for the experienced pathologist (K = 0.82 [0.71 to 0.94], Kp = 0.79, p =
0.988) and for non-experienced (K = 0.78 [0.64 to 0.91], Kp = 0.71, p = 0.291). There was no
difference for the inexperienced and experienced pathologist, respectively, in terms of rate:
insufficient material (p = 0.72, p = 0.48), follicular neoplasm (FN) (p = 1.0, p = 0,09), falsepositive (p = 1.0, p = 0.72), false-negative using the NDM (p = 0.55, p = 0.60), false-negative
using the BDM (p = 0.72, p = 0.72), proportion of malignant tumors identified using the NDM (p
= 1.0, p = 0.59) and BDM (p = 1.0, p = 0,66). The analysis of interobserver reproducibility
showed a significant difference in both the first moment (K = 0.55 [0.38 to 0.72], Kp = 0.50, p =
0.011) or the second afther six months (K = 0,54 [0.37 to 0.71], Kp = 0.44, p = 0.021). There was
also a difference in the number of unsatisfactory results in both the first (p = 0.001) and second
examination (p = 0.021), and FN in the second analysis (p = 0.015). However, no difference in
the proportion of malignant tumors identified in the first (p = 0.79 [NDM], p = 0.81 [BDM]) and
the second analysis (p = 0.78 [NDM], p = 0,50 [BDM]), as well as false-positive (p = 0.62, p =
0.21) and false-negative in the first analysis (p = 0.55 [NDM], p = 0.72 [BDM]) and the second
analysis (p = 0.60 [NDM], p = 0.72 [BDM]). Conclude that the intraobserver reproducibility in
the interpretation of FNA cytology is consistent and independent of experience to the survey,
however, the interobserver reliability is influenced. The experience of the examiner influences
significantly the proportion of unsatisfactory results and cases compatible with FN. The use of
the NDM should be avoided when interpreting the case compatible with FN, in view of the high
rate of malignancy observed in these samples, even when the reports are issued by experienced
pathologists. Surgical indication of thyroid nodules should take into account the high rate of
false-negative results observed at the FNAC, which occurs even when performed by experienced
pathologist, as well as the high rate of unsatisfactory results, mainly due to the high rates of
malignancy observed in these samples. | |
