Immediate Hemodynamic Effects of Pulmonary Artery Flow Restrictors in Univentricular Circulation: Hypothesis-Generating Insights from High-Fidelity Physiologic Streaming Data Using Machine-Learning

Hypoplastic left heart syndrome has been historically palliated with the surgical Norwood or hybrid procedures. Transcatheter placement of pulmonary flow restrictors (PFR) with prostaglandin to maintain an unobstructed ductus arteriosus has emerged as an option for single-ventricle patients that are poor surgical candidates. High-fidelity physiologic data has been used to predict outcomes in interstage patients. This study aimed to characterize the acute hemodynamic effects of transcatheter flow restrictors on pulmonary to systemic blood flow ratio using physiologic data. A single-center retrospective review of 10 patients who underwent transcatheter flow restrictor placement was performed. High-fidelity physiologic data (1-second resolution) from before, during, and after flow restrictor placement were utilized. Data were aggregated with the Sickbay Clinical Platform. An autoregressive integrated moving average model was used to compare variables of interest before, during, and after flow restrictor placement. A total of 10 patients were included in the study for a total of 123,509 patient seconds of monitoring. A statistically significant decrease in pulmonary-to-systemic blood flow ratio, systemic saturation, and renal oxygen extraction, and an increase in blood pressure, were observed during and after PFR placement. Flow restrictor placement is associated with an acute decrease in calculated pulmonary to systemic blood flow ratio and an increase in calculated renal oxygen extraction ratio (related to an increase in systemic oxygen delivery). More importantly, this study demonstrates that high-fidelity physiologic data platforms can be used for real-time assessment of the effect of pulmonary artery flow reducers. Eventually, clinical decision algorithms may be able to guide the pulmonary artery flow reducer strategy, both intra-procedurally and afterward.