I read with interest the systematic review and meta-analysis by Samaniego-Laguna et al.,¹ which addresses whether intraoperative hemoadsorption during cardiopulmonary bypass (CPB) reduces postoperative complications. The reported reduction in acute kidney injury (AKI; risk ratio [RR], 0.75; 95% confidence interval [CI], 0.59-0.96; p = 0.020) is potentially important because AKI was one of the primary outcomes. I believe that the analysis may raise concerns about the robustness and reproducibility of this renal signal. First, the investigators may have inadvertently omitted an apparently eligible contemporary randomized trial. The investigators searched databases through January 2025 and used forward snowballing to identify articles that had been missed or published while the review was being conducted.¹ The OXICARD trial was published online on January 21, 2025, 3 days after registration of the review protocol and almost 1 year before the meta-analysis was published online.² OXICARD randomized adults undergoing scheduled cardiac surgery with anticipated CPB longer than 90 minutes to intraoperative oXiris hemoadsorption or standard care, a comparison that appears to be compatible with the review’s eligibility criteria. AKI occurred in 14 of 33 patients receiving oXiris and 12 of 35 controls, corresponding to an unadjusted RR of approximately 1.24.² The absence of OXICARD warrants clarification because it may have affected the conclusions of the meta-analysis. Indeed, a subsequent updated synthesis incorporating OXICARD identified 9 randomized trials reporting cardiac surgery–associated AKI (947 patients) and found no significant benefit with an inverse-variance random-effects model (RR, 0.80; 95% CI, 0.63-1.03; p = 0.08; I² = 40%) and very low-certainty evidence.³ Its leave-one-out analysis identified the Abou-Arab trial as an influential study.^2,3 This observation does not establish that OXICARD alone explains the difference from Samaniego-Laguna et al.,¹ because study eligibility and model specification also differed between the reviews, but it does demonstrate that the inference is sensitive to incorporation of contemporary evidence and to analytically defensible model choices.