
ADHERENCE TO INSULIN IN SINGAPOREAN PEDIATRIC TYPE 1 DIABETES PATIENTS AND ITS IMPACT ON GLYCEMIC CONTROL AND HEALTH-CARE UTILIZATION
Author(s) -
Brandon Chua,
Xin Yan Lim,
Kar Men Poh,
Jamie Stephanie Caleb,
McVin Hua Heng Cheen,
Sheow Lei Lim,
Ngee Lek
Publication year - 2019
Publication title -
asian journal of pharmaceutical and clinical research
Language(s) - English
Resource type - Journals
eISSN - 2455-3891
pISSN - 0974-2441
DOI - 10.22159/ajpcr.2019.v12i12.35558
Subject(s) - medicine , glycemic , diabetes mellitus , confidence interval , poisson regression , logistic regression , type 2 diabetes , medical prescription , insulin , type 1 diabetes , pediatrics , emergency medicine , endocrinology , environmental health , population , pharmacology
Objective: Optimizing glycemic control is challenging with insulin non-adherence. This study aimed to characterize the prevalence of non-adherence among Singaporean pediatric patients with type 1 diabetes mellitus (T1DM) and investigate its associated outcomes.
Methods: Singaporean patients with T1DM aged ≤18 years old with ≥1 year of insulin prescription between 2012 and 2016 were included in this retrospective, single-center longitudinal study. Patients on insulin pumps were excluded from the study. Non-adherence was defined as medication possession ratio (MPR) <100%. Glycemic control was defined using mean hemoglobin A1c (HbA1c) within the study period. Health-care utilization was defined as the number of outpatients, inpatient, and emergency visits. The t-test, Chi-square test, logistic regression, and Poisson regression were used to analyze means, proportions, factors associated with non-adherence, and association of non-adherence and health-care utilization, respectively. Sensitivity analyses were performed for MPR thresholds of 80% and 95%.
Results: A total of 206 patients were included in this study. Non-adherent patients were older, had a longer duration of diabetes since diagnosis and shorter duration of follow-up. Gender, race, financial class, and number of concurrent medications were comparable between groups. The prevalence of non-adherence was 34.0% (95% confidence interval [CI]: 27.9–40.7%). Non-adherent patients had a higher average HbA1c (non-adherent: 9.6% [2.1] vs. adherent: 8.6% [1.3], p 5 years of diabetes were more likely to be non-adherent.
Conclusion: Non-adherence defined as MPR 5 years of diabetes to improve their adherence to insulin therapy.