This study utilized advanced informatics techniques to compare procedure duration in patients undergoing radiofrequency atrial ablation treated with active esophageal cooling to those treated with traditional luminal esophageal temperature monitoring. Contextual inquiry, workflow analysis, and data mapping were utilized. The findings demonstrated reduced procedure time and variability with active cooling.
Various methods are utilized during radiofrequency (RF) pulmonary vein isolation (PVI) for the treatment of atrial fibrillation (AF) to protect the esophagus from inadvertent thermal injury. Active esophageal cooling is increasingly being used over traditional luminal esophageal temperature (LET) monitoring, and each approach may influence procedure times and the variability around those times. The objective of this study is to measure the effects on procedure time and variability in procedure time of two different esophageal protection strategies utilizing advanced informatics techniques to facilitate data extraction. Trained clinical informaticists first performed a contextual inquiry in the catheterization laboratory to determine laboratory workflows and observe the documentation of procedural data within the electronic health record (EHR). These EHR data structures were then identified in the electronic health record reporting database, facilitating data extraction from the EHR. A manual chart review using a REDCap database created for the study was then performed to identify additional data elements, including the type of esophageal protection used. Procedure duration was then compared using summary statistics and standard measures of dispersion. A total of 164 patients underwent radiofrequency PVI over the study timeframe; 63 patients (38%) were treated with LET monitoring, and 101 patients (62%) were treated with active esophageal cooling. The mean procedure time was 176 min (SD of 52 min) in the LET monitoring group compared to 156 min (SD of 40 min) in the esophageal cooling group (P = 0.012). Thus, active esophageal cooling during PVI is associated with reduced procedure time and reduced variation in procedure time when compared to traditional LET monitoring.
With a rise in the incidence of atrial fibrillation (AF) and an aging population, there is an increased demand for left atrial ablation to achieve pulmonary vein isolation (PVI) for the treatment of AF1. Optimizing procedure duration and minimizing variability are of increased interest among electrophysiologists and hospitals to meet population needs. During PVI procedures, one major risk is thermal injury to the esophagus due to the anatomical proximity of the left atrium to the esophagus2. Many methods exist to protect the esophagus from injury, including the current standard, luminal esophageal temperature (LET) monitoring, and other more recent developments including mechanical esophageal deviation and active esophageal cooling3.
Recent studies have found that LET monitoring may offer limited benefits over using no protection at all4,5,6. Additionally, LET monitoring necessitates frequent pauses of the procedure in response to luminal temperature alerts, which notify operators that the esophagus has reached dangerous temperatures. Recent data have shown that the distance between the temperature sensor and the radiofrequency (RF) catheter influences the sensitivity of LET monitoring, with greater than 20 mm of distance resulting in the absence of detection of significant temperature rises7. Moreover, large lag times (up to 20 s) in temperature rises and large gradients in temperature (up to 5 °C) across the esophageal wall exist, further challenging the ability of LET monitoring to detect temperature elevations quickly enough to avert tissue damage8. Depending on the electrophysiology lab, the use of LET monitoring also requires frequent fluoroscopy exposure to patients and staff to reposition the temperature probe. These additional burdens may prolong the procedure, as reported in a recent study of a community hospital system in which a reduction in procedure duration when using active esophageal cooling instead of LET monitoring was found9.The use of active esophageal cooling allows the placement of contiguous point-to-point ablation lesions in the left atrium without the need to pause radiofrequency ablation due to temperature alarms or heat stacking. As a result, procedural pauses are reduced, and the contiguity of lesions is enhanced. This effect allows a reduction in procedure time and fluoroscopy time, and an improvement in the long-term efficacy of the ablation in reducing the recurrence of arrhythmias9,10,11,12,13.
As practice in an academic setting can vary drastically from a community hospital lab due to the introduction of trainees performing procedures while undergoing their education, the impact of the esophageal protection method is less certain. Moreover, advances in human factors analysis to ensure the identification of clinical data structures documenting the critical steps of each ablation case can be leveraged to facilitate studies of this type. Multiple individuals representing various specialties are involved during an ablation, making contextual inquiry useful for understanding the clinical workflows and pairing key activities with electronic health record (EHR) data structures14,15. Consequently, this study aimed to leverage medical informatics with contextual inquiry to compare the procedural efficiency of PVI procedures conducted with active esophageal cooling to those performed with LET monitoring.
This research was performed in compliance with the institutional guidelines of the University of Texas, Southwestern Medical Center, approval number STU-2021-1166. Data were collected retrospectively through chart review, and thus the need for patient consent was waived.
1. User workflow analysis
2. Observation of placement and use of an esophageal cooling device
NOTE: The placement and use of the active esophageal cooling device have been previously demonstrated and can be seen in Zagrodzky et al.10.
3. Structured data extraction
4. Identification of data necessitating manual extraction
5. Manual data extraction
Patient characteristics
In this analysis, a total of 164 patients were identified who underwent radiofrequency PVI between January 2020 and January 2022. Patients were included regardless of whether they received only PVI or received additional lesions such as roof lines, floor lines, mitral isthmus lines, etc. LET monitoring was performed with a single-sensor temperature probe and was performed by the same teams and in the same labs as the cases with active cooling. There were 63 patients who received LET monitoring for their PVI during the study period and 101 patients who received active esophageal cooling for esophageal protection. There were similar proportions of AF type in both groups (Table 1).
Procedure duration and procedure variability
Procedure duration was defined as the time from the first sheath placed to the time of the last sheath removal. The mean procedure time in patients that underwent LET monitoring was 176 min ± 52 min. In the actively cooled group, the mean procedure time was 156 min ± 40 min, representing a 20 min overall reduction in procedure duration (P = 0.012). The median procedure time was 172 min (interquartile range [IQR] = 144 to 198 ) in the LET monitored group, and 151 min (IQR = 129 to 178 ; P = 0.025) in active esophageal cooling group. Overall, there was a median reduction of 21 min (FIGURE 1). Other than differences in the operator, no other factors differed between the groups other than the type of esophageal protection utilized. As such, the difference in procedural duration is believed to be due entirely to the pauses required with LET monitoring, reacting to temperature elevations, as well as the need to reposition repeatedly while ablating around the pulmonary veins. Although a long-term efficacy analysis has not yet been performed at this clinical site, data from elsewhere have shown improved efficacy with cooling when compared to LET monitoring. This is believed to be due to the improved point-to-point lesion sequencing that can be completed without interruption from local overheating alarms.
In the context of the technique described here, these results highlight the utility of the technique of workflow analysis, human factors analysis, and contextual inquiry to facilitate uncovering and analyzing data that can provide important insights into clinical practice. Traditional analyses of this type often rely on the manual extraction of large quantities of data, adding time and cost burdens to clinical investigations while reducing reliability and consistency. Incorporating advanced informatics techniques as described here opens new avenues for investigation without requiring extensive time and funding.
Esophageal Protection | |||
Active Esophageal Cooling (n=101) | LET Monitoring (n=63) | ||
Patient age (years), mean (SD) | 67.9 ± 11.3 | 64.5 ± 11.6 | |
Gender | Male | 66 | 46 |
Female | 35 | 17 | |
AF Type | Paroxysmal Atrial Fibrillation | 55 | 36 |
Persistent Atrial Fibrillation | 38 | 23 | |
Long-standing Persistent Atrial Fibrillation | 8 | 4 |
Table 1: Patient characteristics, including age, gender, and type of atrial fibrillation treated.
Figure 1: Histogram comparing the procedure times of both groups. The green bars show patients receiving LET monitoring; the blue bars show patients receiving active esophageal cooling. Abbreviation: LET = luminal esophageal temperature. Please click here to view a larger version of this figure.
Supplementary File 1: REDCap data storage extraction form. An example of the data extraction form used for this protocol, demonstrating the specific data elements recorded. Please click here to download this File.
This investigation demonstrates the use of advanced informatics techniques, including contextual inquiry, workflow analysis, and the pairing of key activities with electronic health record (EHR) data structures, to analyze the impact on procedural times of two different esophageal protection methods utilized during cardiac ablation. This is the first study of the effects of esophageal cooling on procedure time and variability to be performed in an academic medical center, where trainees (fellows) receive clinical training in electrophysiologic procedures and perform many of the procedures as a part of this training while being overseen by experienced electrophysiologists. The main finding of this study is that the use of active esophageal cooling was associated with shorter procedure times and less variability around procedure times. Leveraging expertise from trained informaticists ensured accuracy in data identification and facilitated the data acquisition.
The reduction in procedure time and the variability around procedure time offers several benefits. Better predictability of procedure duration improves hospital scheduling, and reducing procedure times may allow additional cases to be scheduled, further improving hospital operations. More importantly, patient risk is reduced as procedure time is shortened. Increased operative duration, in general, increases the risk of complications such as surgical site infections, venous thromboembolism, bleeding, pneumonia, urinary tract infections, renal failure, and hematoma formation18. The likelihood of developing a complication increases with increasing operative time increments (i.e., 1% for every 1 min, 4% for every 10 min, 14% for every 30 min, and 21% for every 60 min increase in operative time)18. In the case of left atrial ablation, access time in the left atrium is the most significant procedural variable for the risk for post-operative cognitive dysfunction19.
A previous study in a community medical center also found procedural time savings associated with the use of active esophageal cooling during left atrial ablation for the treatment of atrial fibrillation9. The mechanism behind this effect relates to the elimination of frequent pauses from overheating that result in ablations and the temperature alarms that are utilized in LET monitoring. As active cooling eliminates overheating and therefore, the need for temperature alarms, it allows electrophysiologists to proceed without pauses20,21,22.
The critical steps in this protocol include properly identifying the individuals and their roles in the procedure to accurately record real-time field observations, probing to uncover any unconscious behaviors involved in experts' workflows, and identifying specific elements of interest related to outcomes to determine where these variables are recorded and located in the Epic chronicles database. With careful completion of these steps, similar analyses can be undertaken for countless outcomes of interest.
The limitations of this analysis include the non-randomized allocation of patients and the retrospective collection of data recorded as the standard of care in the EHR. Although non-randomization introduces the potential for unmeasured confounders to influence results, no secular changes to treatment protocols occurred during the time period investigated in this analysis. Likewise, the use of data recorded as the standard of care in the hospital EHR may reduce the potential for bias in the data.
In conclusion, using contextual inquiry, workflow analysis, and data mapping to analyze procedural timing, this study demonstrated reduced procedure time and variability with active cooling when compared to traditional LET monitoring.
The authors have nothing to disclose.
The authors would like to acknowledge the staff of the UT Southwestern Department of Electrophysiology: Cheryl Thomas RN, Roma Alfonso RN, Eileen Dwyer RN, Anish Varghese RN, Josey George RCIS, Pam Harrison RCIS, and Carolyn Carlson RN. Data are available upon request from the authors.
Blanketrol III hyper-hypothermia system | Gentherm Medical, Cincinnati, OH | Model 233 | Programmable heat exchanger for temperature regulation |
ensoETM | Attune Medical, Chicago, IL | ECD02A | Active esophageal cooling device |
EPIC Clarity | Epic System Corporation, Verona, WI | Electronic Health Record reporting database | |
REDCap | Nashville, TN | Secure web application for building and managing online surveys and databases, including compliance with 21 CFR Part 11, FISMA, HIPAA, and GDPR |