Validation of Cardiomems Pulmonary Artery Pressure Monitor Using the Donovan Mock Tank

Abstract

Maintenance of HF symptoms requires patients to monitor changes in limb edema, sudden weight gain and edema as markers for progressing or worsening HF. However, these symptoms can occur suddenly and can lead to decompensation and hospitalization if not addressed in a timely manner. The CardioMEMS HF System was developed to monitor patient pulmonary artery pressures and catch worsening heart failure before symptoms of fluid overload manifest. Indeed, the CardioMEMS has been shown to reduce HF hospitalizations by up to 30% with great reliability and reproducibility. Recently, a case report at Banner University Medical Center Tucson highlighted potential pitfalls in treating patients with CardioMEMS based on faulty readings from the Smart Pillow home system, highlighting a need for further understanding of potential biases in CardioMEMS readings against a known pressure. This study aimed to characterize CardioMEMS Hospital System and Smart Pillow (home system) readings against the Donovan Mock Circulation Tank (Mock Tank) – a chamber used to test and validate mechanical circulatory support devices. A series of groups was devised to represent a spectrum of pulmonary artery pressures and readings were taken from the PAP chamber of the Mock tank and compared to readings from the CardioMEMS device using the Hospital System. Linear regressions confirmed great agreement between the two measurement methods (R squared >0.95, p<0.0001) and Bland-Altman plots revealed a negative proportional bias (-1.272mmHg, p<0.0001) that suggested the CardioMEMS underestimation was greater at higher PAP chamber pressures. Biases in CardioMEMS measurements should be considered and additional confirmations of fluid overload should be considered when changing the treatment plan of patients with HF.