Abstract 13151: Can Depth of Chest Compressions Be Estimated From Compression-Induced Artifacts in the Thoracic Impedance or ECG Signal?

Abstract

Objective: To explore the feasibility of estimating chest compression depth (CD) by analyzing the compression-induced artifacts in the thoracic impedance (TI) or ECG signals from out-of-hospital cardiac arrest (OHCA) patients who receive manual CPR. Methods: Data from 641 OHCA patients was recorded by Philips MRx defibrillators with Q-CPR meters in 2011 and 2012. For each of the 94,566 extracted segments (6 seconds long) with chest compressions, the CD from the Q-CPR meter and the peak-peak value of TI signal (PPV-TI) were obtained. Univariate linear regression was performed between PPV-TI and CD for the whole population and the coefficient of determination ( R 2 ) was calculated. Linear regression was also performed in each of the individual patients who received at least 5 minutes of chest compressions (n=450). A subset of segments (n=608) were analyzed from asystolic patients (n=215), where the signals in the ECG were assumed to be compression-induced artifacts. In these segments, the CD and the peak-peak value of ECG signal (PPV-ECG) were obtained and linear regression was performed. Results: CD was 47 ± 11 mm (mean ± standard deviation). The R 2 between PPV-TI and CD was 0.05 (Figure 1A). For the segments with the 5% smallest PPV-TI, 18% of the compressions were deeper than 50 mm. The mean of R 2 between PPV-TI and CD in individual patients was 0.25 (Figure 1B), and 35.3% of the patients had an R 2 value smaller than 0.1. The R 2 between PPV-ECG and CD was 0.001 (Figure 1C). Conclusions: The linear correlation between PPV-TI and CD was low for the entire population. There was almost no correlation between PPV-ECG and CD. These results suggest that it is unreliable to estimate compression depth from the compression-induced artifacts in either the TI or ECG signal. Within individual patients, the R 2 of PPV-TI varied and was below 0.1 in >35% of the patients, which suggests that estimating the trend of compression depth from the change of TI is not feasible in many patients.