Circadian rhythms in pump parameters of patients on contemporary left ventricular assist device support

Abstract

Background: Algorithms to monitor pump parameters are needed to further improve outcomes after left ventricular assist device (LVAD) implantation. Previous research showed a restored circadian rhythm in pump parameters in patients on HeartWare (HVAD) support. Circadian patterns in HeartMate3 (HM3) were not studied before, but this is important for the development of LVAD monitoring algorithms. Hence, we aimed to describe circadian patterns in HM3 parameters and their relation to patterns in heart rate (HR). Methods: 18 HM3 patients were included in this study. HM3 data were retrieved at a high frequency (one sample per 1 or 2 h) for 1–2 weeks. HR was measured using a wearable biosensor. To study overall patterns in HM3 parameters and HR, a heatmap was created. A 24-h cosine was fitted on power and HR separately. The relationship between the amplitude of the fitted cosines of power and HR was calculated using Spearman correlation. Results: A lower between patient variability was found in power compared with flow and PI. 83% of the patients showed a significant circadian rhythmicity in power (p < 0.001–0.04), with a clear morning increase. All patients showed significant circadian rhythmicity in HR (p < 0.001–0.02). The amplitudes of the circadian rhythm in power and HR were not correlated (Spearman correlation of 0.32, p = 0.19). Conclusions: A circadian rhythm of pump parameters is present in the majority of HM3 patients. Higher frequency pump parameter data should be collected, to enable early detection of complications in the future development of predictive algorithms.