Real-World Effectiveness of Type 2 Diabetes Medications

This poster presents findings from an AIM-AHEAD Research Fellowship Program project evaluating the real-world comparative effectiveness of tirzepatide versus semaglutide among adults with type 2 diabetes (T2D). We developed a scalable target trial emulation framework using an active-comparator, new-user design and multi-institutional electronic health record (EHR) data. The framework was developed for four EHR data sources (TriNetX, OCHIN, MedStar Health, and UTHealth Houston), with the initial implementation conducted in TriNetX. Inverse probability of treatment weighting and inverse probability of censoring weighting were used to address confounding and differential follow-up. At approximately 40 weeks, tirzepatide was associated with greater reductions in hemoglobin A1c (HbA1c) and body mass index (BMI) compared with semaglutide. The estimated between-group differences were −0.54 percentage points for HbA1c (95% CI, −0.62 to −0.47) and −1.37 kg/m² for BMI (95% CI, −1.58 to −1.17). Tirzepatide was also associated with higher glycemic target attainment. These findings were directionally consistent with the SURPASS-2 randomized trial and demonstrate the feasibility of scalable target trial emulation using real-world EHR data to support future individualized treatment effect estimation and data-driven treatment decisions.