2316-P: Predictors of Therapeutic Inertia in the Action to Control Cardiovascular Risk in Diabetes (ACCORD) Trial

Introduction and Objective: Therapeutic inertia (TI), or failure to intensify or de-intensify therapy when appropriate, is widespread in T2D management, but whether TI also occurs in clinical trials is unknown. This post hoc analysis examines whether TI was present and potential predictors of TI in the ACCORD clinical trial. Methods: In ACCORD, a multicenter trial that enrolled approximately 10,000 participants with T2D and CVD or its risk factors, participants were randomly assigned to intensive (HbA1c <6%) or standard (HbA1c 7.0-7.9%) glycemic targets, with a trial protocol dictating clear parameters for intensification of therapy. We defined visit-specific TI as an HbA1c above the treatment arm-specific threshold for action with no increase or decrease in insulin dose or number of oral glucose-lowering agents (for which doses were not available). We examined the association of age, sex, race and ethnicity, visit HbA1c, and hypoglycemia frequency with TI at 4-, 8-, and 12-month study visits among participants on insulin, for whom medication titration data were more complete compared to participants not on insulin, using mixed-effects logistic regression models adjusted for socioeconomic, trial-specific, and clinical covariates. Results: Among 5,341 participants on insulin across 13,657 visits, HbA1c was not at goal at 67% of visits, with TI present in 26% of these encounters. Age and sex were not associated with TI. Compared to HbA1c 8.0-8.9%, higher HbA1c was associated with decreased likelihood of TI (OR 0.53 [0.39-0.70] for HbA1c 9.0-9.9% and OR 0.64 [0.42-0.97] for HbA1c ≥10%). Hypoglycemia was associated with increased likelihood of TI (OR 1.06 [1.03-1.10] per 1 unit increase in weekly hypoglycemic events). Black race was associated with increased likelihood of TI, both in unadjusted and fully adjusted models (OR 1.45 [1.25-1.67]). Conclusion: TI was present in ACCORD and was associated with Black race, hypoglycemia frequency, and lower HbA1c despite protocols intended to minimize TI as part of a clinical trial. Disclosure R.M. Perez: None. P. Jan: None. D. Wexler: Other - Data Monitoring Committee; Ended; Novo Nordisk. Other - Data Monitoring Committee; Current; Amgen Inc. J. Seiglie: None. S. Cromer: Other - Spouse is employee of Depuy Synthes; Current; Johnson & Johnson.