Assessment of correlation between effluent fluid ionized calcium and post-filter ionized calcium in critically ill patients undergoing regional citrate anticoagulation in continuous renal replacement therapy (RCA-CRRT)

BACKGROUND: Regional citrate anticoagulation (RCA) is standard in continuous renal replacement therapy (CRRT). Post-Filter ionized calcium (Post-Filter iCa) is the gold standard for monitoring anticoagulation efficacy, but it requires blood sampling. Effluent fluid iCa can be measured without blood sampling. This study assessed the correlation and agreement between Effluent Fluid iCa and Post-Filter iCa in critically ill patients undergoing RCA-CRRT. METHODS: We conducted a prospective non-interventional study from June to November 2024 at West China Hospital of Sichuan University. Thirty-six critically ill patients receiving CVVHDF with RCA were enrolled, yielding 216 paired samples. Ionized calcium was measured in Pre-Filter, Post-Filter, and Effluent Fluid samples using a point-of-care analyzer. CRRT was performed using the Prismaflex system with ST150 filters and Oxiris tubing. Post-filter blood and effluent fluid samples were drawn simultaneously to ensure direct comparability. RESULTS: This study analyzed 216 paired samples from 36 critically ill patients on CVVHDF with RCA to assess the relationship between Post-Filter and Effluent Fluid ionized calcium. A moderate correlation was found (r = 0.551 [CI 0.45,0.64], p < 0.001). Bland-Altman analysis showed a mean difference of - 0.024 mmol/L (95% CI: -0.033 to - 0.015), with limits of agreement from - 0.153 (95% CI: -0.1682 to - 0.1378) to 0.105 mmol/L (95% CI: 0.0900 to 0.1204). Variance component analysis demonstrated that only 13.3% of variability (ICC = 0.134) was attributable to between-patient differences, while 86.7% reflected within-patient temporal variation. ICC for Post-Filter iCa reliability over time was 0.269, and for Effluent iCa was 0.146, indicating substantial dynamic changes. Agreement between methods was moderate at early time points (ICC 0.514-0.628 at 2-24 h) but declined at 48-72 h (ICC 0.424-0.435). Cluster analysis of serial calcium measurements identified three distinct profiles among 36 patients: a dominant cluster (66.7%, n = 24) with target-range levels; a high-calcium cluster (19.4%, n = 7), and a low-calcium cluster (13.9%, n = 5). CONCLUSION: Effluent Fluid iCa demonstrates moderate correlation with Post-Filter iCa but substantial within-patient variability limits its use as a direct substitute. While it may offer some supplementary information for tracking anticoagulation trends, it cannot replace Post-Filter sampling at present. Cluster analysis identified distinct calcium phenotypes associated with filter lifespan, though further multicenter validation would be needed to explore potential clinical utility.