Postcardiotomy Cardiogenic Shock: Initial Results of Oklahoma Hub-and-Spoke Model

Purpose Morbidity and mortality in postcardiotomy cardiogenic shock (PCCS) patients are still high. Recent data from the Society of Thoracic Surgeons (STS) shows an increased number of patients undergoing coronary artery bypass grafting (CABG) in low-volume facilities. To address this, we established the Oklahoma Shock Network to deliver advanced therapy to these patients. We envisioned that transferring these patients to and managing them in hub centers, where highly-trained staff and advanced facilities are available, could improve their outcomes and help low-volume facilities better manage this sick cohort of patients. Methods Oklahoma Shock Network is a hub-and-spoke model that was initiated in September 2014 to manage patients with advanced cardiogenic shock/advanced respiratory failure. As of September 2018, 194 patients were transferred to our facility, of which 50 with severe PCCS were included in this retrospective review. Results All 50 patients were placed on VA ECMO (mean age: 58 ±14; male: 41; mean duration of support: 10 days ± 8). Mean distance traveled was 56 miles ± 59, with a range of 1 to 190 miles. ECMO was initiated in the OR, secondary to difficult weaning from CPB in 28 patients. In 22 patients, ECMO support was initiated after surgery, with mean time from surgery to ECMO initiation of 4 days. Prior to ECMO support, 1 patient was on Impella, while 21 (42%) had intra-aortic balloon pumps. Twenty-four patients underwent CABG (48%). Sixteen patients had other procedures in addition to CABG. One patient underwent VSD repair. Two patients each of MV replacement, AV replacement, and pericardiectomy. Three patients required aortic root replacement. There were no transport-related complications noted. Bleeding and tamponade were the most common complication (14 patients; 28%) followed by pneumonia in 11 patients (22%). Continuous renal replacement therapy was performed in 24 patients (48%). Twenty-nine patients (58%) were successfully weaned, of whom 20 (40%) were discharged from the hospital. Two were bridged to durable MCS, of which 1 patient was discharged (LVAD) and 1 deceased (TAH). Support was withdrawn in 17 patients (34%). Two patients required transition to Bi-VAD, but both died. Conclusion Our data demonstrates the safety of transferring early PCCS patients. The exceptional results justify expansion of the hub-and-spoke model to improve outcomes in PCCS patients.