External Control Augmentation Increases Estimates Precision for Finerenone plus Sodium-Glucose Cotransporter-2 Inhibitors

<h2>Abstract</h2><h3>Background</h3> The FIDELITY pooled analysis of the FIDELIO-DKD and FIGARO-DKD trials showed a complementary benefit of finerenone and sodium-glucose co-transporter-2 inhibitors (SGLT-2is) in patients with chronic kidney disease (CKD) and type 2 diabetes (T2D). This study used US electronic health record (EHR) data to create an external control arm (ECA) to augment the FIDELITY comparator group and improve the precision of finerenone's treatment effect estimates. <h3>Methods</h3> We identified eligible patients from EHR data who met adapted FIDELITY criteria. These ECA patients were matched (1:1) to the pool of SGLT-2i users within the FIDELITY population using a linear programming method based on baseline demographics and clinical characteristics. We then calculated the treatment effects of finerenone on cardiovascular and kidney composite endpoints, hospitalization for heart failure (HHF), and all-cause mortality, incorporating the augmented ECA data. <h3>Results</h3> Eligible ECA patients (n = 877) were matched to FIDELITY SGLT-2i users (n = 877), yielding a median (Q1–Q3) absolute standardized mean difference of 0.000 (0.000–0.004). ECA augmentation improved the precision of the trial estimates; data resembled the original FIDELITY estimates but with narrower 95 % confidence interval (CI) ranges. For the cardiovascular endpoint and HHF, a significant benefit for finerenone + the SGLT-2i subgroup vs the augmented SGLT-2i controls was observed. <h3>Conclusion</h3> Our findings demonstrate that an ECA can effectively augment underrepresented study populations in cardiorenal trials, enhancing the statistical precision of treatment effect estimates when there is sufficient homogeneity between the internal and external control groups.