Impact of C-reactive Protein on Anticoagulation Monitoring in Extracorporeal Membrane Oxygenation
Abstract
Objective
To evaluate the impact of inflammation on anticoagulation monitoring for patients supported with extracorporeal membrane oxygenation (ECMO).
Design
Prospective single-center cohort study.
Setting
University-affiliated tertiary care academic medical center.
Participants
Adult venovenous and venoarterial ECMO patients anticoagulated with heparin/
Measurements and Main Results
C-Reactive protein (CRP) was used as a surrogate for overall inflammation. The relationship between CRP and the partial thromboplastin time (PTT, seconds) was evaluated using a CRP-insensitive PTT assay (PTT-CRP) in addition to measurement using a routine PTT assay. Data from 30 patients anticoagulated with heparin over 371 ECMO days was included. CRP levels (mg/dL) were significantly elevated (median, 17.2; interquartile range [IQR], 9.2-26.1) and 93% of patients had a CRP of ≥5. The median PTT (median 58.9; IQR, 46.9-73.3) was prolonged by 11.3 seconds compared with simultaneously measured PTT-CRP (median, 47.6; IQR, 40.1-55.5; p < 0.001). The difference between PTT and PTT-CRP generally increased with CRP elevation from 2.7 for a CRP of <5.0 to 13.0 for a CRP between 5 and 10, 17.7 for a CRP between 10 and 15, and 15.1 for a CRP of >15 (p < 0.001). In a subgroup of patients, heparin was transitioned to argatroban, and a similar effect was observed (median PTT, 62.1 seconds [IQR, 53.0-78.5 seconds] vs median PTT-CRP, 47.6 seconds [IQR, 41.3-57.7 seconds]; p < 0.001).
Conclusions
Elevations in CRP are common during ECMO and can falsely prolong PTT measured by commonly used assays. The discrepancy due to CRP-interference is important clinically given narrow PTT targets and may contribute to hematological complications.