Case 2

Recirculation is a phenomenon unique to VV ECMO in which oxygenated blood exiting the return cannula is drawn directly back into the drainage cannula without passing systemically to the patient. This fraction of blood does not participate in oxygen delivery and represents wasted flow. The three main determinants are: (1) pump speed—higher speed increases recirculation; (2) cannula proximity—closer cannula tips increase recirculation; and (3) native cardiac output—which affects the competition between native venous return and ECMO return flow for drainage cannula uptake.

Recirculation (%) = 100 × (Pre-oxygenator O₂ sat – Mixed venous O₂ sat) / (Post-oxygenator O₂ sat – Mixed venous O₂ sat). 

Using ScvO₂ of 62% as the mixed venous estimate: 

Recirculation = 100 × (82 – 62) / (98 – 62) = 100 × 20/36 = 55.6%. 

This is a very high recirculation fraction, meaning more than half of the pump flow is wasted.

Answer: C. Increasing pump speed from 5.0 to 6.0 L/min would increase recirculation. Higher pump speeds generate more negative drainage pressures, drawing a greater proportion of nearby oxygenated return blood back into the drainage cannula. Decreasing pump speed (A) reduces recirculation by lowering drainage pressures. 

Repositioning cannulas further apart (B) reduces recirculation by increasing the distance between return and drainage ports. Switching to a dual-lumen bicaval cannula (D) physically separates drainage and return ports and directs the return jet across the tricuspid valve, which reduces recirculation.

Not all recirculation requires intervention—some degree is inevitable. However, this patient has clinically significant recirculation (suboptimal SpO₂ of 83% with a rising pre-oxygenator saturation of 82%). Management: (1) First, attempt to reduce pump speed. By reducing drainage pressures, this can lower the recirculation fraction. However, the patient must tolerate the lower blood flow. (2) If imaging confirms cannula malposition, reposition the cannulas to increase separation between drainage and return tips. (3) In rare cases, consider switching configuration to a dual-lumen bicaval cannula with the return port directed across the tricuspid valve or into the pulmonary artery, which physically separates drainage and return and reduces recirculation.

On VV ECMO, oxygenated blood is returned into the right side of the heart and passes into the pulmonary circulation. Therefore, the SvO₂ measured from a pulmonary artery catheter is artificially elevated by the ECMO return blood and does not represent true mixed venous oxygen saturation. This makes the Fick equation unreliable for cardiac output estimation and the recirculation formula inaccurate if PA SvO₂ is used. The most clinically practical alternative is to use a central venous oxygen saturation (ScvO₂) from a central venous catheter with its tip positioned far from the return cannula, as in this case from the internal jugular CVC.