(706) The Clinical Implications of Discordant aPTT and Anti-Xa Results in Pediatric Patients on ECMO

Introduction: Extracorporeal membrane oxygenation (ECMO) is a lifesaving therapeutic approach for critically ill patients that require mitigating conflicting risks. Monitoring of anticoagulation parameters such as activated partial thromboplastin time (aPTT) and anti-Xa are necessary to minimize the risk of bleeding and clotting. The consequences of their concordance vs discordance remain uncertain.

Methods: We completed a secondary analysis of the BATE study dataset of 514 infants and children on ECMO. Target ranges were defined as 60-90 sec for aPTT and 0.4-0.6 for anti-Xa. Both aPTT and anti-Xa were indexed as z-scores to evaluate the distance between them. Bleeding and clotting were assessed on the day following the tests. Statistical analyses involved Fisher exact test to compare proportions, area under the receiver operating characteristic curve (AUROC) to determine test performance, and logistic regression models to adjust for weight and heparin dose.

Results: Of 4660 ECMO days, 2512 recorded both aPTT and anti-Xa alongside following-day bleeding and clotting events. High aPTT and low anti-Xa occurred on 17% (437/2512) of days, low aPTT and high anti-Xa on 1% (26/2512), and concordant results on 82% (2049/2512) of days. Subsequent-day bleeding occurred in 22% (95%CI 18-26%) of days following discordant results (101/463), compared to 28% (95%CI 26-30%) for concordant results (573/2049, p=0.007). Subsequent-day clotting occurred in 16% (95%CI 13-19%) of days following discordant results (74/463), compared to 25% (95%CI 23-27%) for concordant results (522/2049, p< 0.001). Adjusting for weight and heparin dose, discordance was independently associated with bleeding (p=0.03) and clotting (p < 0.001). The AUROC for the distance between aPTT and anti-Xa z-scores, in relationship with bleeding was 0.55 (95%CI 0.53-0.58, p< 0.001) and 0.56 for clotting (95%CI 0.53-0.58, p< 0.001).

Conclusions: Unexpectedly, discordant coagulation tests were independently associated with more favorable bleeding and clotting outcomes. It is possible these anomalies prompted more clinical interventions or anticoagulation adjustments, resulting in better outcomes the following day. Further research is required to corroborate these findings and delineate the mechanism for this association.