3119 — Cardiovascular Disease in Type-2 Diabetes: Using the UKPDS Methodology to Determine Attributable Risk Variance Across 20 Community Clinics

Zeber JE (Verdict), Pugh JA (Verdict), Parchman ML (Verdict)

Objectives:
Cardiovascular disease (CVD) is a serious but preventable complication for patients with diabetes. Besides fixed patient characteristics (age, ethnicity, disease duration), important modifiable risk factors include glucose levels (A1c), blood pressure, lipids and smoking. These potentially mutable behaviors play significant roles in developing CVD, resulting in substantial disease burden, health utilization, and premature mortality. We examine the prevalence range of CVD risk across multiple community clinics and influence of modifiable factors.

Methods:
Patients presenting with established type-2 diabetes were recruited from twenty primary care clinics across South Texas. Patient information and CVD risk factors were collected by survey and chart abstraction. The baseline 10-year risk for fatal and non-fatal myocardial infarction was calculated using the validated UKPDS risk engine. Sensitivity analysis altered all four modifiable risk values to American Diabetes Association (ADA) recommended levels, re-calculating the estimated CVD risk at the patient and clinic level to obtain the attributable risk.

Results:
313 patients had sufficient data to calculate UKPDS risk levels. For each modifiable risk factor, fewer than half these patients were at recommended ADA levels and only 15 percent had good control of all factors. The mean 10-year baseline risk for any CVD event was 16 percent, of which 5 percent (one-third) was due to modifiable factors. Veterans experienced nearly twice the baseline risk as other patients, yet with little difference in the modifiable percentage. Across the 20 clinics, the range of mean CVD risk was 7-49 percent, with a four-fold variance in attributable risk. The primary variable driver of risk reduction was A1c levels, followed by smoking status and lipids.

Implications:
CVD risk for diabetic veterans is dangerously elevated with substantial clinic variations. Targeting modifiable factors can dramatically reduce this hazard, as a substantial fraction of baseline risk is addressable through attention to mutable behaviors. Given a 20 percent VA diabetes prevalence rate, this risk reduction translates into approximately 25,000 preventable CVD events and 10,000 deaths per year.

Impacts:
Quality diabetes care requires targeting modifiable factors substantially associated with CVD risk. Efforts to recognize high-risk populations can better tailor interventions, allocate VA resources, and improve the health status of chronically ill veterans.