J Cardiothorac Vasc Anesth. 2026 Jun 10:S1053-0770(26)00490-8. doi: 10.1053/j.jvca.2026.06.013. Online ahead of print.
ABSTRACT
OBJECTIVES: To test whether renal dysfunction modifies the prognostic value of 24-hour lactate clearance for survival to discharge after extracorporeal cardiopulmonary resuscitation (ECPR).
DESIGN: Retrospective cohort study with a 24-hour landmark approach.
SETTING: Single tertiary academic medical center.
PARTICIPANTS: Of 158 patients screened for ECPR, 99 with calculable lactate clearance and available 24-hour creatinine comprised the analytic cohort.
INTERVENTIONS: None (observational study).
MEASUREMENTS AND MAIN RESULTS: Effect modification was tested using multivariable logistic regression with an interaction term (lactate clearance × elevated 24-hour creatinine >2.0 mg/dL), adjusting for age, cardiopulmonary resuscitation duration, shockable rhythm, and baseline lactate. After baseline lactate adjustment, the interaction odds ratio (OR) was 0.63 per 10% increase in clearance (95% confidence interval, 0.45-0.87; p = 0.005), indicating attenuated prognostic value in patients with elevated creatinine. Stratified analyses confirmed that higher clearance was associated with survival only in patients with normal creatinine (OR, 1.30; area under the receiver operating characteristic curve [AUROC], 0.654), whereas no meaningful association was observed with elevated creatinine (OR, 0.98; AUROC, 0.516). The interaction remained robust after excluding patients with chronic kidney disease (OR, 0.59, p = 0.005) and when modeling creatinine continuously (OR, 0.91, p = 0.028).
CONCLUSIONS: Among patients undergoing ECPR who survive to 24 hours, elevated creatinine defines an interpretation boundary for percentage-based lactate clearance. Clinicians should interpret this metric cautiously when renal dysfunction is present and consider absolute lactate values as complementary information. These findings should be considered hypothesis-generating, given the single-center retrospective design, modest events-per-variable ratio, and the inability to fully disentangle renal-mediated metabolic factors from baseline lactate distribution differences; prospective multicenter validation is required before clinical implementation.
PMID:42401501 | DOI:10.1053/j.jvca.2026.06.013