Sign In

Remember Me

Development of a Personalized, Patient-Centered Glycemic Control Benchmarking Tool for T2DM

Presented at the 74th Scientific Sessions of the American Diabetes Association, San Francisco, June 15, 2014.


Authors: MARCIA A. TESTA, DONALD C. SIMONSON, Boston, MA

Glycemic control treatment targets for T2DM are intended to trigger clinical action in the management of patients; however, targets are usually benchmarked against averaged, population outcomes, and are not specific to individual patient characteristics. To explore treatment effectiveness heterogeneity (TEH) and to pilot test a personalized diabetes treatment benchmarking tool, we developed a causal model, statistical algorithm and a prototype interactive calculator using a pooled database of clinical, demographic and outcomes data from 19 diabetes RCTs (N=6870; 989 clinics) with rigorous protocols as the high quality benchmark standard. The prototype was developed from a database subsample of new or recently diagnosed patients with two treatment options using multiple logistic regression equations to obtain estimated high benchmark probabilities (HBPs) for achieving 12-wk, HbA1c < 7% and < 8.0%. The Excel-based calculator required input of individual patient pretreatment characteristics including HbA1c and FPG after 3 wks of diet and exercise only (D&E), sex, age, BMI, diabetes duration, race/ethnicity, prior treatment and planned treatment option (monotherapy or D&E). The HBPs for a White male, age 50 yrs, BMI = 30, FPG = 150 mg/dl and HbA1c = 9%, diabetes duration = 1 yr and treated previously with D&E were 0.51 and 0.94 for HbA1c < 7% and < 8% after 12 weeks on sulfonylurea monotherapy, and 0.06 and 0.48 if remaining on D&E. HBPs for a Black female, age 50 yrs, BMI = 36, FPG = 150 mg/dl, HbA1c = 9.5%, diabetes duration = 1 yr and treated previously with D&E were 0.41 and 0.78 on sulfonylurea, and 0.04 and 0.18 if remaining on D&E. Mean (SD) clinic-specific HBPs with case mix adjustment for personalized data were 0.35 (0.18) and 0.67 (0.17) for HbA1c < 7% and < 8% respectively indicating substantial TEH among clinics reflecting the variability in patient characteristics. Personalized benchmarking can provide a more equitable and patient-centered quality of care standard for T2DM.

By

Leave a Reply

Skip to toolbar