AUTHORS: Davide Manca, Federico Appiani, Giovanni Colombo
ABSTRACT: Recent years have seen a growing interest in studying new methods to predict the correct dose of levothyroxine (LT4) that is used as a synthetic drug for hypothyroidism treatment and supplements/substitutes the T4 hormone secreted by the healthy thyroid. The paper proposes a quantitative procedure for the optimal administration of LT4 through a model-based approach dependent on some individualized parameters of the patient to free the endocrinologist from the current rough guesses of LT4 dosage. A dedicated algorithm receives as input data the patient features along with the major indicator of thyroid status, i.e. the serum concentration of TSH, the thyroid-stimulating hormone. Then the procedure calculates the Residual Thyroid Function (RTF), which quantifies the thyroid efficiency status, and eventually the optimal LT4 dose. As a complement to the quantitative models available in the scientific literature, we also take into account the age of the patient to determine the optimal dose that can avoid any suboptimal prescriptions regarding hypothyroidism diagnosis for both young and elderly individuals.