Endocrine Abstracts

Objectives: The study of the real indications for thyroid ultrasound and in which pathological conditions the echography is of real use and not for abuse.

Material and methods: In the last 10 years, over 13 000 echographies were performed in 5136 consecutive patients for ‘thyroid disturbances’. In a previous paper we described 7 echographic patterns for thyroid ultrasound (9th ECE). The diagnosis of Hashimoto’s thyroiditis (HT) was considered if antithyroperosidase antibodies (ATPO) were above 34 IU/ml. Idiopathic myxedema (IM) was considered if hypothyroidism was not associated with HT. Graves–Basedow disease (GBD) was based on positive TSH receptor antibodies (TRAB).

Results: Diagnostic at onset: normal thyroid was found in 3489 cases, HT in 487 cases, and GBD in 94 cases from whom 65% had an overlap between ATPO and TRAB. Subacute thyroiditis occurred in 13 cases, IM in 34 cases, benign nodules in 1009 patients and malign nodules in 10 cases. The echographic description at onset revealed a high positive predictive value (PPV) of 92% for pattern 1 in the diagnosis of HT. The high negative predictive value for pattern 4 (82.73%) suggests that a macronodule is rarely associated with HT. In GBD, pattern 5 had a medium PPV (42%). Changing echographic pattern in evolution suggested that in HT, the pattern changed in only 12 patients (2.5%). In subacute thyroiditis, the echographic shape changed suddenly from day to day. In GBD, in 24 patients without Hashimoto association (75%), the echographic shape changed from pattern 5 to quasi-normal (pattern 7), after 3–5 years. In IM, no change was registered. Only 25 benign nodules (2.5%) improved and changed their shapes. No changes were observed for malign nodules during follow-up prior to surgery (1–6 months).

Conclusions: It seems that thyroid echography is performed 3–4 times more than needed.