Intracardiac pressures in 274 patients with heart failure were measured using an implantable hemodynamic monitor (IHM) and were analyzed in a retrospective manner. An automated pressure change detection (PCD) algorithm was developed using the cumulative sum method. The performance characteristics of the PCD algorithm were defined in all patients who developed a heart failure–related event (HFRE); patients without HFRE served as controls. Optimal PCD threshold values were chosen using a receiver operator curve analysis. Each of the pressures measured with the IHM were evaluated using the PCD analysis. All had sensitivities ≥80 % and false-positive rates <4.7/patient-year; however, estimated pulmonary artery diastolic pressure (ePAD) had the best performance. An ePAD based on the optimized PCD threshold of 6.0 yielded a sensitivity of 83 % and a false-positive rate of 4.1/patient-year for detecting patients making the transition from compensated to decompensated heart failure. These performance characteristics were not significantly different for patients with an ejection fraction > vs. <50 % , estimated glomerular filtration rate > vs. <60 mL/min/1.73 m2, or age > vs. <60 years.
The automated PCD algorithm had high sensitivity and acceptable false-positive rates in detecting the development of decompensated heart failure before the patient developed worsening symptoms and required acute medical care. These data support the development of a prospective study to examine the utility of adding an automated PCD algorithm to IHM-based management strategies to prevent decompensated heart failure.