Roberto would certainly be a good candidate for VV ECMO if he needed it, he just doesn’t need it, at least yet. He has excellent cardiac function, no respiratory acidosis, a reasonable hemoglobin, a clear CXR and a normal lactate, confirming that he is tolerating the current saturations just fine. While it can be frustrating to not act on a saturation this low, it is currently not harming him.

Roberto was accepted directly to the cardiothoracic surgery ECMO service with pulmonary critical care consultation and assistance. We were all very suspicious of the numbers, and were contemplating a VQ scan. A TEE was suggested to look for a missed intracardiac shunt, but his saturations were normal preoperatively and any new ASD, VSD etc would shunt left to right, and not cause cyanosis. We were able to easily wean the FiO2 to 50% with no change in his arterial saturation. Having seen this type of extreme intrapulmonary shunting with nitrate vasodilators like NTG and Nipride (and iNO), we acted on a hunch. Since is cardiac funtion was really excellent, we cut the Milrinone in half, and then stopped it, while monitoring his cardiac parameters with the Swan-Gantz catheter. Within an hour his arterial saturation was 90% on FiO2 of 40%, and within 3 hours it was 98%. He was extubated the next day and transferred back to his original hospital where his subsequent post-operative course was unremarkable.

This is an example of a very rare side effect of a common medication that you will likely never witness. The take-home message is not to worry about Milrinone, but to avoid reflexively overtreating hypoxemia with ECMO when it is actually not causing a problem for the patient, and wait until it is - either directly with inadequate DO2 and rising lactate, or indirectly with the hemodynamic impact of escalating ventilator settings.