BACKGROUND: Subtotal thyroidectomy is a widely accepted surgical procedure for Graves' disease. The purpose of this work is to evaluate functional long-term results and determine predictive prognostic factors of postoperative thyroid function. STUDY DESIGN: This is a retrospective study conducted on 202 patients with Graves' disease undergoing subtotal thyroidectomy during the period 1979-2002. Predictive prognostic factors of final thyroid status were investigated by logistic ordinal regression, and probability of hypothyroidism during the years of follow-up was obtained by the Kaplan-Meier method. RESULTS: Surgery controlled hyperthyroidism in 196 out of 202 patients (97%). The probability of hypo-, eu-, and hyperthyroidism at 5 years was 62.1%, 35.5%, and 2.4%, respectively. No statistical change in thyroid function occurred in the follow-up after 60 months. Multivariate analysis by a logistic ordinal regression analysis showed that weight of the remnant, age, and gender seemed to influence long-term thyroid function. The higher rates of euthyroidism were obtained when the remnant weight was between 6 and 8 g. No recurrence or persistence of hyperthyroidism occurred with remnant weights under 5 g. CONCLUSIONS: Subtotal thyroidectomy controlled hyperfunction symptoms in 97% of our patients. Cure (euthyroidism) of Graves' disease patients should be attempted by leaving a thyroid tissue remnant between 6 and 8 g. Even more significant, our results suggest that euthyroidism rates could be improved by leaving a smaller remnant in elderly women and greater remnants in young men.
BACKGROUND: Subtotal thyroidectomy is a widely accepted surgical procedure for Graves' disease. The purpose of this work is to evaluate functional long-term results and determine predictive prognostic factors of postoperative thyroid function. STUDY DESIGN: This is a retrospective study conducted on 202 patients with Graves' disease undergoing subtotal thyroidectomy during the period 1979-2002. Predictive prognostic factors of final thyroid status were investigated by logistic ordinal regression, and probability of hypothyroidism during the years of follow-up was obtained by the Kaplan-Meier method. RESULTS: Surgery controlled hyperthyroidism in 196 out of 202 patients (97%). The probability of hypo-, eu-, and hyperthyroidism at 5 years was 62.1%, 35.5%, and 2.4%, respectively. No statistical change in thyroid function occurred in the follow-up after 60 months. Multivariate analysis by a logistic ordinal regression analysis showed that weight of the remnant, age, and gender seemed to influence long-term thyroid function. The higher rates of euthyroidism were obtained when the remnant weight was between 6 and 8 g. No recurrence or persistence of hyperthyroidism occurred with remnant weights under 5 g. CONCLUSIONS: Subtotal thyroidectomy controlled hyperfunction symptoms in 97% of our patients. Cure (euthyroidism) of Graves' diseasepatients should be attempted by leaving a thyroid tissue remnant between 6 and 8 g. Even more significant, our results suggest that euthyroidism rates could be improved by leaving a smaller remnant in elderly women and greater remnants in young men.