Individualized target heart rate in critically ill atrial fibrillation patients: insights from echocardiography and physiological signal analysis.

Objective.Atrial fibrillation (AF) is common in critically ill patients and is associated with increased mortality. Two treatment strategies are used: rhythm control and heart rate (HR) control; however, the optimal HR target in the latter remains uncertain. This study aimed to develop and compare methods for determining the target HR range using transthoracic echocardiography and physiological-signal analysis.Approach.The study was performed in a level 3 medical ICU. A custom software tool (OIIM_Python) was developed to analyse ECG, arterial pressure, and respiratory waveforms; to compute HR (HR_py) and cardiac output (CO) (CO_py). These measurements were compared with echocardiography, which provided five consecutive left ventricular outflow tract velocity-time integrals (VTI_echo) and corresponding HR (HR_echo). In phase one, 7 patients in sinus rhythm were assessed, yielding 31 paired measurements under variable preload and afterload conditions. In phase two, 13 patients with AF contributed 101 paired measurements. For both methods, mean and maximal CO were calculated, and the target HR was defined as the HR at which CO reached its maximum.Main Results.In sinus rhythm, HR_py and CO_py showed minimal bias relative to echocardiography (-1.48 bpm and-0.0003 l min-1, respectively) with clinically acceptable agreement. In AF, maximal CO_echo was significantly higher than mean CO_echo (5.24 ± 1.30 vs 4.79  ± 1.42 l min-1; +9%,p < 0.001), and target HR was significantly lower than mean HR_echo (98  ± 19 vs 110 ± 15 bpm;-11%,p < 0.001). Physiological-signal analysis similarly showed higher maximal CO_py compared with mean CO_py (6.03 ± 1.59 vs 5.32 ± 1.41 l min-1; +12%,p < 0.001), and target HR was substantially lower than mean HR_py (89 ± 17 vs 114 ± 15 bpm;-22%,p < 0.001).Significance.This feasibility study demonstrates that target HR associated with maximal CO can be estimated in critically ill patients with AF using echocardiography and/or physiological-signal analysis. These methods may support individualized HR-targeted management but require validation in larger multicentre interventional studies.