Conventional coagulation and thromboelastograph parameters and longevity of continuous renal replacement circuits.
OBJECTIVE: To determine the relationship between conventional and thromboelastograph (TEG) coagulation parameters and continuous renal replacement therapy (CRRT) circuit longevity. DESIGN: Conventional coagulation and TEG parameters were measured at the commencement of and during CRRT. Time to circuit cessation was measured and only circuits reaching a predetermined rise from baseline in the pressure gradient across the haemofilter were diagnosed as failing due to clotting. All other circuits were excluded from analysis. SETTING: A general critical care unit of a metropolitan tertiary hospital. PATIENTS AND PARTICIPANTS: Fourteen consecutive patients requiring CRRT were studied. The CRRT technique used was continuous veno-venous haemodialysis. INTERVENTIONS: Thromboelastograph measurements were made prior to the commencement of CRRT and daily thereafter for each circuit. The international normalised ratio (INR), activated partial thromboplastin time (APTT) and platelet numbers were measured at commencement and 8 hourly thereafter. Heparin was used for anticoagulation unless considered contraindicated. MEASUREMENTS AND RESULTS: Forty-seven circuits with a mean (SD) circuit life of 33.0 (30.2) h were entered. Twenty-five circuits fulfilled circuit clotting criteria; the mean circuit life was 30.8 (22.1) h. Heparin anticoagulation was found to prolong circuit life significantly despite adequate mean circuit life, 33.2 (35.7) h, in heparin-free circuits. The starting APTT and the TEG variables reaction time (R) and coagulation time (RK) were significantly correlated. The starting APTT, starting RK and mean time taken for the amplitude to increase from 2 to 20 mm (K) were predictive of circuit life. None of these variables predicted which patients would need heparin. CONCLUSION: While TEG variables more closely predicted circuit longevity than conventional coagulation variables, the clinical benefit of TEG monitoring of anticoagulation for CRRT would appear to be minimal.