作者
Lei Li,Dan Zhang,Manrong Xu,Shaohua Li,Yun Shen,Shengdi Lu
摘要
Objective: This study aims to emulate the CONFIDENCE trial and compare combination finerenone plus empagliflozin with each monotherapy for albuminuria reduction in Chinese patients with chronic kidney disease (CKD) and type 2 diabetes (T2D). Design: This study employed modified target trial emulation. The primary contrast was the per-protocol effect, estimated by cloning, censoring, and inverse probability of censoring weighting, which is not susceptible to grace-period immortal time bias; a propensity-score-matched assignment analysis, the observational analogue of intention-to-treat, was secondary. Setting: Electronic health records from three Chinese tertiary (Grade A) hospitals, from January 2023 to September 2025, were utilized. Participants: A total of 2,895 adults (965 per strategy) with CKD and T2D met the CONFIDENCE eligibility criteria, selected by 1:1:1 propensity score matching from 3,822 eligible patients. Main outcome measures: Relative change in urine albumin-to-creatinine ratio (UACR) from baseline to day 180 was the main outcome measure (co-primary comparisons with Holm-Bonferroni correction). Results: < 0.001); this is not a true intention-to-treat effect and is supportive only. Hyperkalemia occurred in 12.7%, 12.8%, and 3.5%, respectively, acute kidney injury in 4.9%, 2.8%, and 2.3%, respectively, and symptomatic hypotension in 1.9%, 0.6%, and 0.4%, respectively. The findings were consistent across 10 sensitivity analyses and six subgroups. Conclusions: Combination finerenone and empagliflozin was associated with greater albuminuria reduction than either monotherapy, directionally consistent with CONFIDENCE. Because the emulated strategy permitted sequential initiation within 90 days whereas CONFIDENCE randomized simultaneous initiation, this is a modified emulation rather than a direct replication. Albuminuria is a surrogate endpoint and follow-up was 180 days, so hard kidney and cardiovascular outcomes remain unstudied. Despite loss to follow-up of 29.6% to 32.7%, the finding was robust in a tipping-point analysis. Hyperkalemia was frequent with finerenone, so biochemical monitoring remains necessary. Findings apply to comparable Chinese tertiary-care populations.