Recently, Zaborek et al. raised a Poisson regression model for levothyroxine(LT4) dosing scheme to predict individual dose of LT4 after thyroidectomy: daily LT4 dose=e[2.02+0.01(W)-0.0037(A )-0.098(F)-0.01(B)+0.007(T)+0.108(I)-0.014(M), where W is the weight of the patient (Kg), A is the age of the patient (years), and F is the gender (for women 1, male is 0), B represents the patient's body mass index (BMI), T represents the preoperative TSH level, I represents whether the patient takes iron preparations (1, if not 0), M represents whether the patient takes multivitamins/minerals (1, if not 0). We demonstrated its value with our retrospective data in our center. Therefore, we intend to carry out this randomized controlled trial in order to further evaluate the model. The clinical significance of this method can provide a basis for the future use in clinical optimization of individualized dosing.](streamdown:incomplete-link)

Patients who underwent total/near-total thyroidectomy or completion thyroidectomy and require levothyroxine replacement therapy will be enrolled in this study. They will be randomized distributed into two groups. In one group, patients will be dosing LT4 according to weight-based algorithm, namely 1.6 μg /kg/d. In the other group, patients will be dosing LT4 according to the Poisson regression model algorithm. They will be followed up 3-6 weeks after surgery for the first time. During the follow up, thyroid hormone will be tested and LT4 dosing will be adjusted accordingly. Then, we compare patients in which group will achieve euthyroidism more quickly.

Inclusion Criteria:

• patients underwent total/near-total thyroidectomy or completion thyroidectomy; patients with benign disease, or low-risk differentiated thyroid cancer, or medullary thyroid cancer who don't need TSH suppressive therapy;

Exclusion Criteria:

• patients with heart disease or severe liver/kidney insufficiency; patients taking liothyronine in addition to levothyroxine;