Background: Many patients with type 2 diabetes mellitus (DM) may have other major illnesses including heart and kidney disease. Patients often have their DM therapies intensified to reach the recommended hemoglobin A1C (HbA1C) target (<7%). However, few data indicate whether early intensification of treatment with insulin, versus treatment with an additional oral antidiabetic agent, benefits patients with cardiovascular disease (CVD) and/or chronic kidney disease (CKD). Patients who only have DM (but not CVD or CKD) will likely experience delays in the development of CVD, CKD, and death from early intensification with insulin compared with patients who receive intensification with oral agents. However, among patients with CVD and/or CKD, intensification with insulin may have no benefit, or even cause harm.
- To examine the impact of intensification of DM treatment with insulin on CVD and CKD outcomes in persons with and without these two co-morbidities:
- To compare time to CVD or death in patients with intensification of DM treatment with insulin vs. oral agents in combination, stratified by history of CVD and CKD
- To compare time to CKD or death in patients with intensification of DM treatment with insulin vs. oral agents in combination, stratified by history of CVD and CKD
- Among the subgroup of patients using insulin, and after an elevated HbA1C (primary treatment failure), to examine the impact of intensive insulin regimens compared to less intensive insulin regimens on CVD, CKD or death in persons with and without these two co-morbidities:
- To compare time to CVD or death in patients using intensive insulin regimens compared to less intensive insulin regimens
- To compare time to CKD or death in patients using intensive insulin regimens compared to less intensive insulin regimens
Study Design: Retrospective cohort design
- Data Sources – This project will use National Veterans Affairs (VA) databases, including administrative, clinical, laboratory and pharmacy data, and Medicare linked records.
- Participants - National cohort of US veterans who initiated use of a single oral antidiabetic drug (new-users) and had an HbA1C of 7.1 to 9.0% in the first year of treatment. Our cohort will include veterans aged ≥18 years who received their regular medical care and prescriptions within a VA facility during the time period from 10/01/2001 to 09/30/2008 with follow-up through 9/30/2010. Study subjects will be those who started on their first DM treatment with metformin or sulfonylurea and have an HbA1c between 7.1- 9% between day 91 and day 365 after initiating treatment.
- Interventions & Comparator Groups - We will study whether treatment intensification during this first year is beneficial for all patients. We will compare two groups: 1) therapy intensified during the first year with an additional oral agent; and, 2) therapy intensified during the first year with insulin. We will compare the effects of these intensification strategies on the risk of heart attack, stroke, or death, or decline in kidney function, using intensification with insulin as reference. Three exposures will be examined:
- therapy intensified with an additional oral agent
- therapy intensified with insulin (insulin + oral)
- switch to insulin alone
- Outcomes – a) development of CVD (acute myocardial infarction and stroke), CKD (including End Stage Renal Disease), and death from all causes in individuals without CVD and CKD; b) progression of CVD (acute myocardial infarction and stroke), CKD (including End Stage Renal Disease), and death from all causes in individuals with CVD and CKD. Potential confounders include gender, race, CVD, CKD, baseline and time dependent: HbA1C, serum creatinine, BMI, medication use, blood pressure, and lipids.
- Statistical Methods - Marginal structural Cox proportional hazard regressions will assess the association between treatment intensification exposures among patients with and without CVD and/or CKD at baseline. A similar analysis plan will be used to assess treatment differences among groups using intensive insulin and non-intensive insulin regimens after the occurrence of an elevated HbA1c.
Expected Outputs: Four scientific reports and associated publishable manuscripts.
Expected date of project completion: Winter 2013
EHC Priority Condition(s): Diabetes, Cardiovascular Disease