Metformin adherence and discontinuation among patients with type 2 diabetes: A retrospective cohort study

Aims: To describe discontinuation and adherence to metformin in the United Kingdom. Methods: This was a retrospective analysis of data from the Clinical Practice Research Datalink database of type 2 diabetes patients aged >= 18 years with >= 1 metformin prescription in 2013. Metformin use was assessed in new and ongoing users, defined, respectively, as not having or having a prescription for metformin in the baseline period. Discontinuation was assessed in all patients and adherence in patients who did not discontinue metformin. Factors predictive of discontinuation and adherence were assessed. Results: Discontinuation among new and ongoing users was 35.9% and 23.1%, respectively. Among the continuers of metformin treatment, the adherence rate was 40.5% and 44.3% among new and ongoing users, respectively. Among new users, baseline use of DDP-4 inhibitors (HR 1.276) and diabetes duration (HR 1.013) were associated with an increased risk of discontinuation, whereas increased age (HR 0.997), concomitant lipid-lowering therapy (HR 0.956), macrovascular disease (HR 0.952), and chronic kidney disease (HR 0.952) were associated with a decreased risk of discontinuation among ongoing users. Variables positively associated with adherence in both user groups were (HR values for all patients) age (1.021), smoking status (1.188), and baseline comorbidities: chronic kidney disease, depression, dementia, and chronic obstructive pulmonary disease (1.106, 1.192, 2.27, and 1.211, respectively), while obesity (0.936) and HbA1c 8.0-8.9% (0.862; reference < 6.5%) were negatively associated with adherence. Conclusions: About one-third of patients initiating metformin discontinued within 12 months and fewer than 50% of all patients are adherent to metformin.