Methods:

Researchers analyzed data from patients in the OptumLabs Data Warehouse, a large database of patients across the United States enrolled in Medicare Advantage and commercial insurance plans, from January 1, 2008–June 30, 2019. All patients included were aged ≥18 years and had filled a prescription for generic levothyroxine with any of the three most common generic preparations (Mylan, Sandoz, or Lannett) during the study period. For ≥3 months prior to the “index date,” all were on stable doses of generic levothyroxine and had ≥1 normal TSH level during that time frame. After the “index date,” all patients continued on the same levothyroxine dose and preparation from the same manufacturer until the first TSH test result obtained 6 weeks to 12 months after the “index date.” The index date was within 12 months after the first filled generic levothyroxine prescription and was: 1) the date the patient switched to a different generic levothyroxine prescription, or 2) for patients who remained on the same preparation and dose, a randomly assigned date between the first fill date and 12 months later. Patients were excluded for pregnancy, hypopituitarism, or for conditions or medications that could affect TSH level (including use of thyroid extract or triiodothyronine).

The primary outcome was the proportion of patients with first TSH level 6 weeks to 12 months after the index date that was normal (0.3-4.4 mIU/L), abnormal (<0.3 or >4.4 mIU/L), or markedly abnormal (<0.1 or >10 mIU/L). Researchers also evaluated mean TSH levels and assessed whether TSH levels differed among patients with endogenous hormone production versus those without.