05.11.2019 | Original Research | Ausgabe 5/2020
Transpulmonary thermodilution before and during veno-venous extra-corporeal membrane oxygenation ECMO: an observational study on a potential loss of indicator into the extra-corporeal circuit
Journal of Clinical Monitoring and Computing
- Alexander Herner, Tobias Lahmer, Ulrich Mayr, Sebastian Rasch, Jochen Schneider, Roland M. Schmid, Wolfgang Huber
Haemodynamic monitoring before extra-corporeal membrane oxygenation (ECMO) might help to optimize the effectiveness of ECMO. However, there are concerns that pulmonary arterial and trans-pulmonary thermodilution (TPTD) might be confounded by a loss of indicator into the ECMO-circuit, resulting in an overestimation of volumetric parameters. Since there is a lack of data on indicator dilution techniques during ECMO, we compared TPTD-measurements before and during ECMO. TPTD-derived parameters before and after initiation of ECMO were compared in 14 intensive care unit-patients with veno-venous ECMO and TPTD-monitoring (PiCCO®). Eight patients had a jugular and six patients a femoral central venous catheter (CVC). Cardiac index, global end-diastolic volume index (GEDVI) and extra-vascular lung water index (EVLWI) before ECMO as well as the ECMO-flow were comparable in patients with jugular and femoral CVC. Pre-ECMO, cardiac index (CI) was not significantly different compared to values during ECMO (4.5 ± 1.7 vs. 4.4 ± 2.1 L/min/m2; p = 0.43). By contrast, GEDVI (791 ± 179 vs. 974 ± 384 mL/m2; p = 0.04) and EVLWI (21 ± 9 vs. 28 ± 11 mL/kg; p < 0.01) were higher during ECMO than before. Increases in GEDVI (36 ± 210 vs. 378 ± 247 mL/m2; p = 0.02) and EVLWI (3 ± 2 vs. 11 ± 8 mL/kg; p = 0.06) were substantially more pronounced in patients with femoral compared to jugular indicator injection. In multivariate analysis, femoral indicator injection was independently associated with larger increases in GEDVI (p < 0.01) and EVLWI (p = 0.04) during ECMO. However, CI and haemodynamic parameters not derived from TPTD, but from pulse contour analysis (systolic and diastolic arterial pressure, stroke volume variation and pulse pressure variation) were not affected by the start of ECMO. Our study demonstrates marked increases in GEDVI and EVLWI after the onset of ECMO. These increases were more pronounced for femoral compared to jugular indicator injection. CI and haemodynamic parameters not derived from TPTD were not affected by the extra-corporeal circuit.