INSULIN USE AND COSTS OF CARDIOVASCULAR DISEASE EVENTS (CVD) AMONG PATIENTS WITH TYPE 2 DIABETES
Author(s)
Nicole M. Engel-Nitz, PhD, Researcher1, Gabriel Gomez Rey, MS, Research Analyst1, Peter Sun, MD, PhD, Research Scientist2, Theresa Rayl, BA, Senior Programmer Analyst11i3 Innovus, an Ingenix Company, Eden Prairie, MN, USA; 2 Lilly Research Laboratories, Indianapolis, IN, USA
OBJECTIVES: To compare CVD event-related costs among patients with Type 2 diabetes (T2D) with and without insulin use. METHODS: Patients with T2D were identified from insurance claims (1998-2004) of a large US health insurance plan. Index date was date of first CVD event; for subjects without CVD events, index date was a medical claim date in the same calendar quarter as the index date for a matched subject in the CVD group. Baseline period was a minimum of 12 months prior to index date, with variable follow-up after index date. Filled prescriptions for insulin were identified from pharmacy claims. US Census data were used to identify distributions of ethnicity and income. Patients with and without insulin use were compared on inflation-adjusted CVD-event related medical costs, adjusted for other risk factors. Two-part models were estimated; the first part used logistic regression to estimate probability of any cost. The second part estimated mean cost among patients with costs >$0; results were compared for estimations using GLM vs. OLS on a log-transformed cost with heteroskedastic smearing estimator. RESULTS: Among 342,638 T2D patients, 4.3% used insulin. Average CVD event-related medical costs were $646 and $705 in insulin vs. non-insulin users (p=0.5590). Among non-insulin users, 6.0% incurred costs, vs. 4.8% of patients with insulin use (p<0.0001). Adjusted odds of incurring CVD event related costs were 25% lower for insulin-users than for non-insulin users (OR=0.75, 95% CI: 0.68, 0.82). Among patients with costs, adjusted mean costs were not statistically significantly different between insulin and non-insulin groups, regardless of estimation technique. CONCLUSION: T2D patients with history of insulin use were less likely to incur CVD-related medical costs than non-insulin users, despite adjustment for risk factors. However, average costs were not statistically significantly different between groups. Estimation technique had a substantial impact on range of predicted mean costs.
Conference/Value in Health Info
2007-10, ISPOR Europe 2007, Dublin, Ireland
Value in Health, Vol. 10, No. 6 (November/December 2007)
Code
PDB36
Topic
Economic Evaluation
Topic Subcategory
Cost/Cost of Illness/Resource Use Studies, Cost-comparison, Effectiveness, Utility, Benefit Analysis
Disease
Diabetes/Endocrine/Metabolic Disorders