Treatment persistence in the use of basal insulins in Poland and Germany


To compare short-term basal insulin therapy persistence and its predictors in Poland and Germany.

Persistence was defined as proportions of patients remaining on the initial basal insulin (analogs: Poland: n = 6,889, Germany: n = 454,067; neutral protamine Hagedorn (NPH) insulins: Poland: n = 50,761, Germany: n = 226,064) over 2 years based on nationwide prescription databases (LRx; IMS Health) in Poland and Germany from 2013 to 2015. Persistence was evaluated by Kaplan-Meier curves (log-rank tests). Risk of discontinuation of initial basal insulin was investigated using Cox regression models adjusting for age, sex, comedication with other glucose-lowering agents and baseline or comedication with antihypertensives, lipid-lowering drugs, antidepressants, and antiepileptics.

In Poland, 2-year persistence was 83.0% in analog insulin and 73.3% in NPH users (p < 0.001). In Germany, persistence was also higher in patients with analog insulins (92.6% vs. 79.0%; p < 0.001). Analog insulin users were less likely to discontinue basal insulin compared with NPH users (adjusted hazard ratio (95%CI): Poland: 0.73 (0.67 - 0.79); Germany: 0.27 (0.27 - 0.28)). Higher age (> 75 vs. ≤ 60 years: Poland: 1.24 (1.16 - 1.33), Germany: 1.09 (1.07 - 1.11)) and GLP-1 receptor agonist use (Poland: 2.76 (1.38 - 5.53), Germany: 1.21 (1.16 - 1.26)) were related to higher risk of discontinuation. Male sex, metformin, sulfonylurea, thiazolidinedione, and short-acting insulin prescriptions as well as antihypertensive, anti-epileptic, and lipid-lowering drug use were associated with lower risk of discontinuation in both countries (all p < 0.05).

This real-world study shows that both in Poland and Germany treatment persistence of newly-prescribed basal insulin is influenced by type of insulin (analog vs. NPH) and by glucose-lowering and other comedications.
.