Diabetes medication persistence, different medications have different persistence rates

Michal Shani, Alex Lustman, Shlomo Vinker

Research output: Contribution to journalArticlepeer-review


Aim To assess the persistence of diabetic patients to oral medications. Methods The study included all type 2 diabetic patients over 40 years, members of one District of Clalit Health Services Israel, who were diagnosed with diabetes mellitus before 2008 and who filled at least one prescription per year during 2008–2010, for the following medications: metformin, glibenclamide, acarbose, statins, angiotensin converting enzyme inhibitors (ACEI) and angiotensin II receptor antagonists (ARBs). Purchase of at least 9 monthly prescriptions during 2009 was considered “good medication persistence”. We compared HbA1c and LDL levels, according to medication persistence, for each medication; and cross persistence rates between medications. Results 21,357 patients were included. Average age was 67.0 ± 11.0 years, 48.9% were men, and 35.8% were from low SES. Good medication persistence rates for ARBs were 78.8%, ACEI 69.0%, statins 66.6%, acarbose 67.8%, metformin 58.6%, and glibenclamide 55.3%. Good persistence to any of the medications tested was associated with a higher rate of good persistence to other medications. Patients who took more medications had better persistence rates. Conclusions Different oral medications used by diabetic patients have different persistence rates. Good persistence for any one medication is an indicator of good persistence to other medications. Investment in enhancing medication persistence in persons with diabetes may improve persistence to other medications, as well as improve glycemic control.

Original languageEnglish
Pages (from-to)360-364
Number of pages5
JournalPrimary Care Diabetes
Issue number4
StatePublished - Aug 2017


  • Diabetes mellitus
  • Family medicine
  • Medication persistence
  • Primary care


Dive into the research topics of 'Diabetes medication persistence, different medications have different persistence rates'. Together they form a unique fingerprint.

Cite this