While David has multiple underlying medical problems, his surgery was technically a success, but his LV was too weak from the sepsis, the surgery, possible emboli, and cardioplegic arrest to support his circulation. He did have two arrests requiring CPR, but has been heavily sedated since then prohibiting any valuable neurological exam other than his reactive pupils and lack of seizure activity, and he cannot get a head CT because of his hemodynamic instability. Many times, when we lack the information we want, we give the patient the benefit of the doubt and proceed with an attempt at rescue with VA ECMO, emptying and resting the heart, removing the catecholamines that increase myocardial oxygen demand, and with mild systemic cooling to protect the brain as much as possible.

In David’s case it appeared to pay off. His lungs cleared quickly and his heart gradually recovered significant function over 3 days on full flow and milrinone afterload reduction. He was treated with heparin as well as amicar because of his recent surgery, and the possible RV thrombus, PE. By day 5 he was ready to come off ECMO.

His sedation was stopped with initiation of ECMO and throughout the run his pupils remained normally reactive, he had no witnessed seizures. rigidity, myoclonic jerking, or other findings consistent with an ischemic insult. But, while his eyes were open, he remained unresponsive and had no spontaneous movements. After no change by ECMO Day 4, he was transported on ECMO for a CCT which showed infarcts in both cerebral hemispheres and the right cerebellum. The neurology consult confirmed that the prognosis was very poor, but the family declined withdrawal of care or a DNR order.

He was taken off ECMO the next day, had a tracheostomy about a week later and subsequently transferred to a long-term care facility in the same unresponsive state about 4 weeks post ECMO. The etiology and timing of the strokes is difficult to establish since he had multiple risks factors (endocarditis with vegetations, surgery and CPB, CPR, possible endocardial thrombus, and VA ECMO. Unfortunately, because patients are, by definition, critically ill when ECMO is considered, usually heavily sedated and frequently paralyzed, and rarely stable enough for a trip to the CT scanner, this is a frequent “blind spot” that is part of ECMO management.