Differences in risk stratification for worsening heart failure in patient with reduced versus preserved ejection fraction using diagnostic parameters available in implantable cardiac monitors

Abstract

Abstract Background Ambulatory management of heart failure (HF) patients require the ability to identify when patients are at highest risk of impending HF events (HFE). The risk stratification ability based on diagnostic parameters measured by implantable cardiac monitors may differ depending on the left ventricular ejection fraction (LVEF) of HF patients. Objective The purpose of this study was to evaluate the difference in risk stratification performance of individual diagnostic parameters currently available in ICMs in identifying patients at increasing risk of HFE in patients with reduced versus preserved LVEF. Methods Patients with history of HFE who were implanted with an ICM was identified from the aggregated and de-identified electronic health record (EHR) database during 2007-2021. The device collected data were merged with the EHR data to create a de-identified database of real-world patients. Currently approved ICMs measure nighttime heart rate (NHR), heart rate variability (HRV), atrial fibrillation (AF) burden, ventricular rate during AF, and activity duration (ACT) daily and stores and transmits this information for the lifetime of the device. Each diagnostic parameter was categorized into high, medium, or low risk using a combination of features predefined in earlier studies. HFEs were defined as inpatient stay with a primary diagnosis of HF or HF related observation unit or emergency visit with IV diuresis administration. Patients were divided into two cohorts of patients with (1) median LVEF ≤ 40% (HFrEF), and (2) median LVEF > 40% (HFpEF). Performance evaluation was performed by simulating monthly follow-ups which consisted of looking at the individual parameter risk states in the last 30 days and evaluating the occurrence of HF events in the following 30 days. A marginal Cox proportional hazards model was used to compare HF events for different risk groups (high, medium, and low) for each individual diagnostic parameter. Results There was a total of 267 patients with median LVEF ≤ 40% (64±13 years old, 67% male) and 622 patients with LVEF>40% (69 ±12 years old, 47% male). For the reduced LVEF cohort, there was a total of 4,867 follow-up months with 65 monthly evaluations (1.3%) had an HF event in the next 30 days. For the preserved LVEF cohort, there was a total of 11,057 follow-up months with 224 monthly evaluations (2.0%) had an HF event in the next 30 days. The event rate comparisons between the risk groups for each individual diagnostic parameter are shown in Figure. As compared to ICM patients with HFrEF, ICM patients with HFpEF had higher absolute event rate across all risk groups, thus lowering the relative risk stratification between risk groups. Conclusion ICM patients with HFpEF showed a higher event rate and hence a higher absolute risk for each risk group, however relative risk between risk groups and hence the risk stratification performance for identifying risk for worsening HF was better in patients with HFrEF.