Risk of over- and under- treatment with levothyroxine in primary care in Copenhagen, Denmark.

Abstract

OBJECTIVE: A decrease over time in thyroid stimulating hormone (TSH) levels when initiating levothyroxine (L-T4) therapy for hypothyroidism has been reported, where treatment most often is initiated with TSH levels below 10 mIU/L. The primary objective of this study was to investigate whether this lower TSH threshold resulted in an increased number of overtreated patients. DESIGN AND METHOD: Retrospective cohort study comprising inhabitants in Copenhagen had TSH measurements requested by general practitioners which led to a new prescription of L-T4 between 2001 and 2012. Over- and under- treatment were defined as TSH <0.1 mIU/L or above 10 mIU/mL, respectively, in three consecutive measurements. Data were analyzed by Aalen-Johansen estimators and Cox proportional hazards models. RESULTS: In total, 14 533 initiations of L-T4 were included in the study. The cumulative risk of being over- or undertreated was 4.7 and 7.4% after 10 years. The hazard of overtreatment was higher among women, younger adults, and with lower initial TSH levels. The hazard of overtreatment decreased over the time period from 2001 to 2012. Among overtreated individuals, the chance of returning to a normal TSH was about 55% after 10 years. In 18% of the cases, L-T4 therapy was initiated on TSH levels less than 5 mIU/L. CONCLUSION: Although a still decreasing threshold for initiating L-T4 therapy is known, the risk of overtreatment (and undertreatment) was low and lessened in the period 2001-2012 among Danish primary care patients. Nevertheless, as many as 18% were started on L-T4 with normal TSH levels.