Adherence to Premixed Insulin in a Prefilled Pen Compared With a Vial/Syringe in People With Diabetes in Singapore

Background

The real-world clinical effectiveness of exogenous insulin is limited by nonadherence. Various insulin delivery systems have been developed to help improve adherence, with prefilled pens gaining popularity among adult Singaporeans with diabetes. However, adherence to insulin in people with diabetes in Singapore and most of Asia has not been studied.

Objectives

This study aimed to compare adherence to premixed insulin formulated in a prefilled pen versus a vial/syringe and to identify predictors of adherence in 955 patients managed at the outpatient clinics of the largest acute care hospital in Singapore.

Methods

In this retrospective longitudinal study, electronic medical and pharmacy refill records were used to determine adherence to insulin over 24 months, measured in terms of compliance and persistence. Compliance is expressed as the medication possession ratio (used as continuous and categorical variables), and persistence is reported as a dichotomous variable with a permissible refill gap of 30 days before discontinuation of therapy is considered. Multivariate linear or logistic regression analysis was used to identify predictors of adherence.

Results

Compared with prefilled pen users, vial/syringe users were older (mean SD] age, 64.1 10.6] vs 62.4 11.9] years; P = 0.032), and more were undergoing polypharmacy (69.6% vs 54.1%; P < 0.001). The mean (SD) medication possession ratio was comparable in vial/syringe versus prefilled pen users (83.8% 26.9%] vs 86.0% 23.2%]; P = 0.266). Prefilled pen users were more persistent with therapy compared with vial/syringe users (odds ratio = 1.36; 95% CI, 1.01–1.86) after adjusting for sociodemographic and clinical covariates. Median time to discontinuation of therapy was comparable (vial/syringe vs prefilled pen: 409 vs 532 days; P = 0.076). Being managed by an endocrinologist and not receiving government subsidies were significant predictors of persistence.

Conclusions

Compared with other studies that found strong associations between adherence and insulin devices, the findings of this study suggest that persistence but not compliance varies by insulin device. The willingness to pay for health care, in addition to affordability, may affect insulin therapy adherence. Patients with uncontrolled diabetes due to suspected nonadherence may benefit from referral to a multidisciplinary care team comprising endocrinologists, diabetes nurse educators, dietitians, and allied health professionals. These findings are applicable to outpatients with similar demographic features managed at other acute care hospitals in Singapore. The impact of insulin devices on glycemic control needs to be investigated in future larger studies.