An audit of thyroid surgery in a general surgical unit.

A total of 143 patients undergoing thyroid surgery in a general surgical unit over an 8-year period were reviewed. In only two patients did thoracic inlet views or thyroid function tests alter clinical management. Fine-needle aspiration failed to detect one well-differentiated follicular carcinoma (false-negative rate 1.1%). The sensitivity for malignancy of fine-needle aspiration, ultrasound and radioisotope scan were 94%, 53% and 24%, respectively. The corresponding specificity was 59%, 72% and 58% and accuracy 65%, 70% and 49%, respectively. The specificity of fine-needle cytology for detecting neoplastic disease (adenoma or carcinoma) was 86% and accuracy 91%. Combinations of fine-needle cytology, ultrasound and radioisotope scanning increased the sensitivity for malignancy, so that fewer tumours were missed, but at the cost of reduced specificity, positive predictive value and accuracy. Hence, ultrasound was only recommended when fine-needle aspiration was inconclusive. Overall perioperative morbidity was 6.3% (one case of postoperative bleeding, two wound infections, four cases of prolonged hypocalcaemia). There were two proven cases of transient, but no permanent, recurrent laryngeal nerve injuries as a result of surgery. Thyroid surgery may be performed satisfactorily by general surgeons with an interest in thyroid disease. Fine-needle cytology is the most informative preoperative investigation. Although aspiration cytology, ultrasound, and scintigraphy all have appropriate indications and limitations, there is no single test or group of tests that can substitute for careful clinical assessment and follow-up.