Adherence, Persistence, and Switching Among People Prescribed Sodium Glucose Co-transporter 2 Inhibitors: A Nationwide Retrospective Cohort Study

Richard Ofori-Asenso,Danny Liew,Samanta Lalic,Mohsen Mazidi,Dianna J Magliano,Zanfina Ademi,J Simon Bell,Jenni Ilomaki
DOI: https://doi.org/10.1007/s12325-019-01077-3
Abstract:Introduction: Non-adherence and non-persistence to diabetes medications are associated with worse clinical outcomes. In this study, we aimed to characterise the 1-year switching, adherence, and persistence patterns among people with diabetes aged 18 years and older prescribed sodium-glucose co-transporter 2 inhibitors (SGLT2is) in Australia. Methods: Using data from Australia's national Pharmaceutical Benefits Scheme (PBS), we identified 11,981 adults (mean age 60.9 years; 40.5% female) newly initiated on SGLT2is (5993 dapagliflozin; 5988 empagliflozin) from September 2015 to August 2017. Adherence was assessed via the proportion of days covered (PDC), persistence was defined as the continuous use of SGLT2i without a gap of ≥ 90 days, and switching was defined as the first change from dapagliflozin to empagliflozin or vice versa. Generalised linear models (GLMs) were used to compare the adherence (PDC = continuous), logistic regression models were used to compare the likelihoods of being adherent (PDC ≥ 0.80), and Cox proportional hazard models were used to compare the likelihoods of persistence and switching between people prescribed empagliflozin and dapagliflozin. Results: Overall, 65.8% (7879/11,981) of people dispensed SGLT2is were adherent (PDC ≥ 0.80) and 72.1% (8644/11,981) were persistent at 12 months. The mean PDC was 0.79 ± 0.27. The use of empagliflozin was associated with higher adherence (PDC = continuous) [odds ratio (OR) 1.04, 95% confidence interval (CI) 1.03-1.05], being adherent (OR 1.39, 95% CI 1.29-1.51), and persisting for 12 months [hazard ratio (HR) 1.14, 95% CI 1.06-1.22] compared with dapagliflozin. Only 4.3% (509/11,981) of people switched between the SGLT2i. Compared with dapagliflozin, people initiated on empagliflozin were less likely to switch [HR 0.46, 95% CI 0.38-0.55]. Conclusions: A considerable proportion of Australians prescribed SGLT2is were non-adherent or non-persistent. However, empagliflozin was associated with better adherence and persistence rates and a lower likelihood of switching compared with dapagliflozin.
What problem does this paper attempt to address?