Methods
This study was approved by the Institutional Review Boards at Children’s National Hospital, Nationwide Children’s Hospital, and the Albert Einstein College of Medicine/Montefiore Medical Center. Healthy adult and pediatric cardiologists, including trainees, across three institutions volunteered to participate providing they had completed at least 1 week of an electrophysiology rotation demonstrating a basic clinical understanding of cardiac ablation procedures. A pediatric electrophysiology simulator designed to illuminate a bulb when catheter electrode contact is achieved with various targets within the vasculature [
17] was set up on a benchtop. The participants were consented and given up to 5 min to become comfortable using the simulator, catheters, and novel torque tool.
After the introductory period, participants were outfitted with surface electromyography (EMG) sensors (Noraxon USA Inc., Scottsdale, AZ, USA) on the muscle belly of the left abductor pollicis, left brachioradialis, left flexor carpi radialis, left flexor carpi ulnaris, left extensor carpi ulnaris, and left extensor carpi radialis (Fig.
3). These muscles were selected due to their involvement in pinching and torquing movements of the left hand. Muscle usage of the right hand was not evaluated as this hand was not directly impacted by the torque tool. The participant was asked to squeeze a stress ball, fully flex the wrist, fully extend the wrist, and apply pressure with the thumb against a stationary object to collect a maximum voluntary contraction (MVC) for all the muscles under investigation. Surgical gloves were worn during the study to mimic clinical conditions. All participants torqued the catheter body with and without the torque tool using their left hand and used their right hand to manipulate the catheter handle.
The participants were asked to insert a catheter into the heart through a femoral approach and simulate typical 3-dimensional mapping of the right atrium and surrounding structures by deliberately touching the catheter tip to 6 targets within the vena cava and on the septal wall of the right atrium, avoiding the AV node, and in the order of: superior vena cava (SVC), inferior vena cava (IVC), fossa ovalis (FO), coronary sinus (CS), typical area of the slow pathway of the AV node, and the posteroseptal tricuspid valve annulus where a right posteroseptal accessory pathway could be located. After successfully reaching the 6 targets, they then simulated creating an ablation lesion either in the area of the slow pathway of the AV node or the posteroseptal accessory pathway by holding the catheter tip on the target for 1 min, again avoiding contact with the AV node. The task was completed a total of four times. A bidirectional D-F curve 7-French catheter (EZ Steer® NAV Bi-Directional, Biosense Webster Inc., Irvine, CA, USA) was used for two trials, whereas a unidirectional B-curve or a unidirectional F-curve 7-French catheter (NaviStar®, Biosense Webster Inc.) was used for another two trials. Each catheter was used once with the torque tool and once without. The choice of catheter, use of the torque tool, and the lesion location for the first trial were each determined by an independent coin toss. The catheter and lesion location remained the same for the first two trials and then switched for the last two trials. The torque tool conditions switched between the first and second trials and then followed the opposite pattern for the third and fourth trials. All potential starting conditions are shown in Table 1. The two catheters and alternating torque tool conditions were used to reduce learning curve bias in the results. Each participant completed paired trials with and without the tool under investigation to limit data bias that may result from limited catheter experience in some participants.
The time to successfully reach the 6 targets, total time to complete the entire task (mapping and lesion creation), and number of attempts to complete a 60-s simulated ablation lesion were recorded. At the conclusion of the task, participants completed the NASA Task Load Index questionnaire (TLX), which has been utilized to assess subjective mental demand, physical demand, temporal demand, performance, effort, and frustration [18], including in the assessment of medical devices [19].
EMG data was collected for the duration of the study. The MVC dataset was post-processed with an amplitude normalization to peak value and smoothed by a root mean squared algorithm (Myo Muscle, Noraxon). Peak values for each muscle group were identified as the maximum voluntary contraction. For each task, the EMG data was smoothed by root mean squared algorithm and normalized to percentage of MVC. Maximum muscle activation, mean activation, and the area under the curve, indicating the total work of the muscle, for each muscle group from each task were identified. A sample wave form of a filtered wave form both with and without the torque tool for the abductor pollicis is shown in Fig. 4.
Statistical analysis of the data was performed in Prism 9.3.1 (GraphPad, San Diego, USA). A Wilcoxon matched pairs signed rank test was used to compare the non-torque tool trials to the torque tool trials for all data collected. Any significant differences were defined as p < 0.05.