(481) Exploring Anticoagulation Trends as Early Infection Markers in 64 Pediatric ECMO/CRRT Patients

Introduction: New infections in pediatric ECMO and CRRT patients pose significant morbidity and mortality, yet infection rates in this population are not well described. Traditional markers like fever, CRP and procalcitonin have limited sensitivity to detect infections on extracorporeal support. Given the known relationship of the inflammatory and coagulation pathways, this study aims to characterize infection rates and investigate if changes in anticoagulation management during ECMO and CRRT can serve as early signals of infections.

Methods: This retrospective cohort study at a tertiary pediatric hospital included ICU patients aged 0-18 who developed infections after 72 hours on ECMO, CRRT monotherapy, or ECMO/CRRT tandem therapy from 2017-2023. Patients with severe coagulopathy or clinically insignificant positive cultures were excluded. We assessed whether the number of anticoagulation dose changes are linked to new infections.

Results: Of 453 extracorporeal runs, 90 new infections were identified in 64 (14.1%) patients. Infections occurred in 14% of ECMO and 18% of CRRT runs. The median time to infection was 18 days for ECMO, 14 days for CRRT and 13 days for tandem therapy, with no significant difference (p=0.86). Urinary tract infections differed significantly by support type (p=0.01), from 39.3% (n=11) on CRRT, followed by tandem (24%, n=6) and ECMO (8.1%, n=3). Respiratory infections differed significantly by support type (p=0.001), from 37.9% (n=14) on ECMO followed by tandem (28%, n=7) and CRRT (0%, n=0). Anticoagulation included 51% of patients on bivalirudin, 36% on heparin, and 12% on citrate. In the 72 hours before an infection, patients on heparin or bivalirudin had a median of 3 dose adjustments, compared to 0.5 for those on citrate. This cohort’s mortality rate was 43.8%, notably higher than the institutional extracorporeal patient mortality rate of 28.9%.

Conclusions: This pediatric cohort had higher institutional mortality than the general extracorporeal patient population. Infections peaked around the second week and varied by support type, underscoring the need for infection prevention and early detection. The association between systemic anticoagulation dose changes and new extracorporeal infections suggests a potential early warning marker, warranting further investigation.