Impact of Levothyrox formulation change on health and healthcare use in France
A pharmaco-epidemiology study in real life from the SNDS data.
An unprecedented number of pharmacovigilance notifications were reported in France following the introduction in 2017 of a new formulation of Levothyrox (the only levothyroxine product available in France by then), giving rise to a high media coverage. A toxicity of this new formulation was suspected.
The objective of this study was to measure the impact of Levothyrox new formulation (NF) on health and healthcare use, based on data of the French national healthcare databases.
The whole population aged 18–85 years treated with Levothyrox in 2016 and/or 2017 was randomly split into two independent groups. Patients of the ‘2017’ group, who initiated Levothyrox NF in April, May or June 2017, were individually matched to patients treated with Levothyrox old formulation (OF) in 2016 (‘2016’ group). Levothyrox NF first dispensing date was considered as the index date. Matching criteria included sociodemographic characteristics, characteristics of the thyroid treatment, and Levothyrox dispensing date.
Patients were followed from the index date up to December, 31 of the same year. Risks of hospitalization or death, sick leaves of ≥7 days among working‐aged patients (<65 years), outpatient visits to general practitioners (GP) and specialists, and various medications consumption were compared in 2017 versus 2016, using conditional Cox models adjusted for baseline health status indicators. Patients treated with oral antidiabetic drugs (OAD) in 2016–2017 were considered as a control population.
A total of 2,075,106 patients treated with Levothyrox were included (1,037,553 treated with Levothyrox NF in 2017 and 1,037,553 treated with Levothyrox OF in 2016) and followed during 7.5 months in mean. Patients were predominantly women (85.7%), aged 61.6 years in mean.
Compared to 2016, in 2017 the risk of all‐cause hospitalization was lower (aHR 0.95, 95% confidence interval [0.94–0.96]) and the risk of death was similar (aHR 0.98 [0.94–1.02]). The risk of sick leave was 2% higher in 2017 compared to 2016 (aHR 1.02 [1.01–1.04]); a similar increase was observed in patients with OAD. The mean number of outpatient visits increased by 2%, from 6.08 in 2016 to 6.20 in 2017 (P < 0.0001); this increase was limited to visits to GP and endocrinologists and to the period from August to October 2017. Renewals of chronic treatments including psychotropic, anti‐hypertensive and lipid‐lowering drugs were slightly more frequent in 2017.
Our results do not provide evidence of increased risks of serious health concerns associated with Levothyrox NF. Though, they show a substantial increase in healthcare use following Levothyrox formulation change.
Find the study report on the impact of Levothyrox formulation change on health and healthcare use in France.