Personalized Sedation Level Monitoring in ICU Patients Using Heart Rate Variability

Abstract

Aim: To develop an automated system to monitor sedation levels in intensive care unit patients using heart rate variability (HRV).Methods: We developed an automatic sedation level prediction system using HRV as input to a support vector machine learning algorithm. Our data consisted of electrocardiogram recordings from a heterogeneous group of 50 mechanically ventilated adults receiving sedatives in an ICU setting. The target variable was the Richmond agitation-sedation scale score, grouped into four levels: "comatose" (-5), "deeply sedated" (-4 to -3), "lightly sedated" (-2 to 0), and "agitated" (+1 to +4). As input we used 14 features derived from the normalized-RR (NN) interval. We used leave-one-subject-out cross-validated accuracy to measure system performance.Results: A patient-independent version of the proposed system discriminated between the 4 sedation levels with an overall accuracy of 52%. A patient-specific version, where the training data was supplemented with the patient's labeled HRV epochs from the preceding 24 hours, improved classification accuracy to 60%.Conclusions: Our preliminary results suggest that the HRV varies systematically with sedation levels and has potential to supplement current clinical sedation level assessment methods. With additional variables such as disease pathology, and pharmacological data, the proposed system could lead to a fully automated system for depth of sedation monitoring IntroductionAccurate administration and optimization of sedatives dosing is important in intensive care units (ICU) [1]. Patients are often sedated in the ICU to increase their level of comfortness, tolerate several procedures and facilitate mechanical ventilation and analgesia [2,3]. Therefore, it is critical to maintain patient's optimal sedation level since both over-and under sedation could lead to adverse patient outcomes including prolonged mechanical ventilation and ICU stay, increased risk of pneumonia and delirium [4,5]. Currently, subjective methods such as the Riker sedation agitation scale (SAS), Richmond agitation sedation scale (RASS), Ramsay sedation scale are used to score the level of sedation in ICUs. These scoring systems rely mainly on patient's response to external and noxious stimulation, which are subjective and are less discriminative during over-sedated state [6].Several electroencephalogram (EEG) based methods have been developed to assess patient's level of consciousness during general anesthesia based on spectral and entropy measures [7,8]. However, several studies have demonstrated their ineffectiveness in monitoring patients sedation level in the ICU environment [9,10]. Therefore, a more robust and reliable objective method is crucial to assess the depth of sedation in the ICU.Heart rate variability (HRV) is a popular noninvasive method that can provide information about the functioning of the autonomous nervous system using electrocardiogram (ECG). We recently investigated the potential of HRV for automatically predicting level of sedation using machine l...