Bias in the Composite Outcomes of Kidney-Cardio Protective Trials in Chronic Kidney Disease: A Meta-Epidemiological...

Background/Objectives: Composite endpoints are commonly used in chronic kidney disease (CKD) trials to enhance statistical efficiency but may not reflect clinically meaningful outcomes. We assessed agreement between composite endpoints and key components using the bias attributable to composite outcome (BACO) index and explored determinants of variability. Methods: We performed a meta-epidemiological analysis of randomized controlled trials evaluating sodium-glucose cotransporter 2 inhibitors, glucagon-like peptide-1 receptor agonists, and non-steroidal mineralocorticoid receptor antagonists in CKD. BACO was defined as the ratio of the log-hazard ratio for the composite endpoint to that of the reference outcome (kidney failure or cardiovascular death), with variance estimated using the delta method. Determinants were analyzed using inverse-variance weighted mixed-effects meta-regression. Results: Eight trials comprising 38 composite endpoints were included. Higher reference-event rates were associated with higher BACO values overall (β: 0.06, 95% CI: 0.02; 0.10) and in kidney failure-referenced analyses (β: 0.07, 95% CI: 0.02; 0.12). Stronger composite treatment effects correlated with higher BACO (β: -1.07, 95% CI: -1.84; -0.30). The number of components and follow-up duration showed no significant association. In cardiovascular death-referenced models, BACO was associated with trial size (β: 0.12 per 1000 participants), mean age (β: -0.04 per 10 years), and female proportion (β: 0.09 per 10% increase). Conclusions: Agreement between composite endpoints and clinically relevant outcomes is driven by the relative frequency and treatment responsiveness of component events rather than endpoint complexity. Composite endpoints in which clinically important outcomes are infrequent may not reliably reflect treatment effects, underscoring need for clinically aligned endpoint strategies.