
Abstract
Purpose:
This studywas designed to optimize the latestgeneration venovenous (vv)-extra-corporeal membrane oxygenation(ECMO)-circuit configuration andsettings based on the evaluation ofblood oxygenation and CO 2 removaldeterminants in patients with severeacute respiratory distress syndrome(ARDS) on ultraprotective mechani-cal ventilation.
Methods:
Bloodgases and hemodynamic parameterswere evaluated after changing one ofthree ECMO settings, namely, circuitblood flow, FiO2ECMO (fraction ofinspired oxygen in circuit), or sweepgas flow ventilating the membrane,while leaving the other two parame-ters at their maximum setting.
Results:
Ten mechanically venti-lated ARDS patients (mean age44 ± 16 years; 6 males; meanhemoglobin 8.0 ± 1.8 g/dL) onECMO for a mean of 9.0 ± 3.8 days)receiving femoro–jugular vv-ECMOwere evaluated. vv-ECMO blood flowand FiO2ECMO determined arterialoxygenation. Decreasing the ECMOflow from its baseline maximumvalue (5.8 ± 0.8 L/min) to 40 % less(2.4 ± 0.3 L/min) significantlydecreased mean PaO 2 (arterial oxy-gen tension; 88 ± 24 to 45 ± 9 mmHg; p \ 0.001) and SaO 2 (oxygensaturation; 97 ± 2 to 82 ± 10 %;p \ 0.001). When the ECMO flow/cardiac output was [60 %, SaO 2 wasalways [90 %. Alternatively, the rateof sweep gas flow through the mem-brane lung determined blooddecarboxylation, while PaCO2 (arte-rial carbon dioxide tension) wasunaffected when the ECMO bloodflow and FiO2ECMO were reduced to\2.5 L/min and 40 %, respectively.In three additional patients evaluatedbefore and after red blood cell trans-fusion, O2 delivery increased aftertransfusion, allowing lower ECMOflows to reach adequate SaO2 .
Conclusions:
For severe ARDS patientsreceiving femoro–jugular vv-ECMO,blood flow was the main determinantof arterial oxygenation, while CO2elimination depended on sweep gasflow through the oxygenator. AnECMO flow/cardiac output [60 %was constantly associated with ade-quate blood oxygenation and oxygentransport and delivery.