Incidence of Newly Detected Atrial Arrhythmias via Implantable Devices in Patients With a History of Thromboembolic Events
Background and Purpose—Evidence of atrial tachycardia/atrial fibrillation (AT/AF) is often sought in patients with ischemic stroke or transient ischemic attack. We studied patients with previous thromboembolic events (TE) who were implanted with devices capable of continuous arrhythmia monitoring to comprehensively quantify the incidence and duration of newly detected AT/AF.
Methods—This study represents a subgroup analysis of the TRENDS trial, which included patients with clinical indications for pacemakers or defibrillators and ≥1 stroke risk factors (heart failure, hypertension, age 65 or older, diabetes, or previous TE). A history of AF was not required. All implanted devices were capable of continuously monitoring the cumulative time spent in AT/AF each day. This analysis focuses primarily on the incidence and duration of newly detected AT/AF (defined as ≥5 minutes of AT/AF on any day) in patients with previous TE, no documented history of AF, and no warfarin or antiarrhythmic drug use.
Results—A total of 319 patients had a history of TE and ≥1 day of device data. Patients with a documented history of AF (n=80), warfarin use (n=56), or antiarrhythmic drug use (n=20) were excluded from analysis. Of the remaining 163 patients, newly detected AT/AF was identified via the device in 45 patients (28%) over a mean follow-up of 1.1±0.7 years. AT/AF recurred infrequently, with only 12 patients experiencing AT/AF on >10% of follow-up days.
Conclusion—Newly detected episodes of AT/AF were found via continuous monitoring in 28% of patients with previous TE. Most episodes would not have been detected by standard intermittent monitoring techniques.
Evidence of atrial fibrillation (AF) is often sought in patients who survive an ischemic stroke or TIA because the presence of AF has important etiologic and therapeutic implications. However, patient symptoms are insensitive and nonspecific for identifying paroxysmal atrial tachycardia/atrial fibrillation (AT/AF),1,2 whereas rhythm monitoring with external systems is hindered by intermittent sampling3 and patient compliance issues.4 In contrast, continuous monitoring has been shown to be significantly more effective for identifying patients with asymptomatic or intermittent AT/AF recurrences,3,5 and the daily amount of AT/AF has been related to the risk of thromboembolic events (TE) in patients receiving cardiac rhythm devices.6
Previous studies7–10 have investigated the role of short-term (up to 30 days) external monitoring for the identification of undiagnosed AF in patients with previous stroke. Because these previous studies essentially show that the likelihood of finding AT/AF increases with longer durations of monitoring, we hypothesized that continuous monitoring would find even more AT/AF. Therefore, we studied a cohort of patients who required implantation of a cardiac rhythm device with a history of TE to quantify the incidence and duration of newly detected AT/AF (NDAF). We also examined NDAF in patients without a previous TE but with other risk factors for stroke.
Subjects and Methods
The study population represents a group of patients that was enrolled in a larger study (TRENDS). The design11 and the main results6 of the TRENDS study have been previously reported. In brief, TRENDS was a prospective, observational study that enrolled patients with a clinical indication for a pacemaker or implantable cardioverter defibrillator and at least 1 of the following stroke risk factors: congestive heart failure, hypertension, age 65 or older, diabetes, or previous stroke/TIA. Patients were not required to have a history of AF, and warfarin use was prescribed by the patient’s managing physician. Physicians were encouraged to follow published guidelines for the use of antithrombotic therapy.12 The TRENDS study reported that TE risk is a quantitative function of arrhythmia burden among the overall population of patients indicated for a cardiac rhythm device. The study protocol was approved by the Institutional Review Board of each participating center and all patients provided informed consent.
The purpose of the current retrospective subanalysis was to quantify the incidence and duration of NDAF through the use of continuous monitoring by implanted cardiac devices. This substudy focuses primarily on those patients enrolled in TRENDS who had a history of stroke or TIA; however, data are also presented on patients without previous TE (but with other stroke risk factors, including congestive heart failure, hypertension, age 75 or older, and diabetes) for comparison. Patients with a documented history of AF were excluded. Similarly, patients using warfarin or antiarrhythmic drugs were also excluded because this may indicate the existence of undocumented but diagnosed AF. Patients were also required to have at least 1 day of device-stored data available for analysis.
Definitions and Device Capabilities
AT/AF burden was defined as the total cumulative duration of AT/AF detected by the device in a given day. AT/AF burden may consist of multiple episodes on a single day or a portion of a single episode that spans multiple days. The implanted devices were capable of continuously monitoring AT/AF burden for up to 14 months. Data were extracted from the devices via telemetry at office visits. Follow-up visits were scheduled with sufficient frequency to ensure data continuity. Previous studies utilizing devices with similar detection algorithms have shown >95% sensitivity and specificity for detection of atrial arrhythmia episodes and measurement of atrial arrhythmia burden.13
NDAF was defined as at least 5 minutes of atrial arrhythmia recorded by the device on any day during the study. In this analysis, only days with at least 5 minutes of AT/AF were considered, because this duration excludes most episodes of atrial oversensing leading to false-positive recordings of AF.14 Because our goal was to study the readily available diagnostic data within the device memory, we did not attempt to distinguish between episodes of atrial tachycardias, atrial flutter, and atrial fibrillation. Furthermore, intracardiac electrograms were not available for each episode because of device memory constraints.
We quantified the percentage of patients with previous TE, no documented history of AF, and no warfarin or antiarrhythmic drug use who experienced NDAF. The percentage of patients having ≥5 minutes, ≥6 hours, or ≥24 hours of AT/AF burden on a single day or 2 and 7 consecutive days of AT/AF were tabulated. Time to identification of NDAF was computed using Kaplan-Meier survival estimates and compared with the log-rank test. We also tabulated the number of patients with recurrences and the timing of recurrences among those with at least 1 episode of AT/AF to illustrate the difficulties with intermittent monitoring, which is defined as the use of an externally worn continuous cardiac rhythm monitor for up to 30 days. Demographics were compared among patients with previous TE who were found to have NDAF vs those with no AT/AF. The incidence of NDAF was also computed among patients without a previous TE (but with other stroke risk factors) and compared to patients with a previous TE.
Continuous variables are presented as means±SD or medians and interquartile range, as appropriate, and categorical variables are presented as percentages. Categorical variables were compared using the χ2 test and continuous variables were compared using a Wilcoxon rank-sum test. A significance level of 0.05 was used.
The authors had full access to the data and take responsibility for its integrity. All authors have read and agree to the manuscript as written.
A total of 3045 patients were enrolled in the TRENDS study from 116 centers. From this population, 319 patients had a history of previous TE and device data available for analysis. Patients with a documented history of AF (n=80), warfarin use (n=56), or antiarrhythmic drug use (n=20) were excluded from the analysis. Among the remaining 163 patients (followed-up a mean of 1.1±0.7 years), the average CHADS2 score was 4.1±0.8, reflecting their high risk for stroke recurrence. The median time from the previous TE to enrollment in the study was 39 (12–73) months. Additional characteristics for the 163 patients included in this subgroup analysis are presented in the Table. Patients with a previous TE who were found to have NDAF were more likely to have heart failure vs those with no detection of AT/AF.
Incidence of NDAF
NDAF was identified by the implantable device in 45 of 163 patients (28%) over the course of follow-up. Detection of NDAF occurred in pacemaker (n=20) and defibrillator (n=25) patients in a similar proportion to their prevalence in the study.
There were 1428 patients in the overall TRENDS study population without a history of previous TE, documented AF, warfarin, or antiarrhythmic drug use and who had device data available (Table). Over a mean follow-up of 1.1±0.7 years, NDAF was found in 432 patients (30%), which was essentially the same incidence rate as for those with a history of previous TE (P=0.49).
AT/AF Burden in NDAF Patients
NDAF patients often experienced individual or consecutive days with considerably greater amounts of AT/AF than the minimum threshold of 5 minutes (Figure 1). For example, 25 patients (56% of NDAF patients with previous TE) had at least 6 hours of AT/AF on a single day. Over the entire study period, the average daily AT/AF burden in NDAF patients with previous TE was 1.8±4.0 hours, whereas the median daily AT/AF burden was 0.08 (interquartile range [IQR], 0.01–1.23) hours. The day with maximal AT/AF duration was greater among NDAF patients with previous TE (median, 10.8; IQR, 2.4–23.9 hours) compared to NDAF patients without previous TE (median, 5.9; IQR, 1.2–20.4 hours), but the difference did not reach statistical significance (P=0.29). Figure 2 shows the distribution of the percentage of follow-up days in which at least 5 minutes of AT/AF was detected by the device among NDAF patients. For example, 73% of NDAF patients with previous TE experienced episodes of AT/AF on <10% of follow-up days. Only 11% of NDAF patients with previous TE experienced AT/AF on a majority of follow-up days, a frequency needed to be reliably detected by random intermittent monitoring.3
Time to Detection of NDAF
The time from implantation of the continuous monitoring device until first detection of NDAF is presented in the survival curve in Figure 3. The median time from enrollment in the study to identification of NDAF was 1.7 (IQR, 0.4–6.7) months among previous TE patients with NDAF and 2.0 (IQR, 0.3 to 5.4) months among nonprevious TE patients with NDAF (log-rank, P=0.56). The continuous arrhythmia data from the device allowed us to simulate the percentage of patients who would have been identified as having NDAF if this rhythm information had only been available on the subset of days monitored by traditional methods. Only 3% of patients would have been identified via a single 24-hour Holter monitor if performed immediately at the time of enrollment, whereas only 4% would have been identified via 48-hour Holter monitoring performed at the time of enrollment in the study. Increasing the initial monitoring duration to 7, 21, or 30 consecutive days would have identified 6%, 9%, or 11% of previous TE patients, respectively. In contrast, continuous arrhythmia monitoring over the follow-up period resulted in significantly greater identification of NDAF (28% of patients) compared to these common intermittent monitoring regimens (P<0.001).
Detection of AF in patients at high risk for stroke will strongly influence anticoagulation decisions.12 The present prospective study, using continuous arrhythmia monitoring, detected NDAF in 28% of patients with previous TE over a mean follow-up of 1.1 years. The majority of NDAF was identified >30 days after study entry. In a small percentage of patients, NDAF was present during most of the follow-up period but the majority had infrequent episodes. The latter episodes of NDAF would have been missed entirely, even with prolonged durations of traditional ECG monitoring.
Incidence of NDAF
Whereas we found a substantially higher percentage of previous stroke patients having NDAF overall compared to other studies, our rates of detection were similar when comparing equivalent monitoring periods. Jabaudon et al8 identified NDAF in 5.0% of patients via 24-hour Holter monitoring, whereas we identified NDAF in 3.1% of patients within the same 24-hour period. Bansil et al9 found NDAF via 48-hour Holter in 4.0% of patients compared to 4.3% in our study over the same period. Jabaudon et al8 also found NDAF in 8.1% of patients via Holter recordings within the first 7 days compared to 6.1% in the present study.
Two previous studies evaluated the incidence of occult AF in patients with cryptogenic stroke. Using 21-day7 and 30-day10 external event monitors, NDAF was detected in ≈20% of patients when inpatient telemetry and electrocardiography did not detect AF. However, the majority of AF episodes detected in these studies were <30 seconds in duration, thereby failing to satisfy the definition of an AF “episode” in recently published AF consensus statements.15 In the present study, 60% of NDAF patients were identified beyond the initial 30 days of the study, potentially adding to the value of long-term arrhythmia monitoring.
In this analysis, we found a similar rate of NDAF in patients with stroke risk factors regardless of the presence of previous TE. This may indicate that the presence of any stroke risk factors is associated with a high likelihood of finding previously unknown AT/AF, particularly in a population with cardiac conditions necessitating device implantation. However, there was a tendency (not significant) for the duration of maximal daily AT/AF burden to be longer among those with a previous TE.
Several previous studies have examined the relationship between AT/AF burden and outcome in cardiac rhythm device patients. A substudy of the MOde Selection Trial (MOST) reported that the composite end point of nonfatal stroke and death was significantly higher in patients having at least 5 minute episodes of high rate atrial arrhythmias.16 This finding, in part, contributed to our selection of the 5-minute threshold for the definition of NDAF.
The TRENDS study recently reported that an arrhythmia burden ≥5.5 hours on any of 30 previous days appeared to double the risk of TE compared to no burden, based on a 20-second definition for AT/AF.6 In a subanalysis of the present patients with previous TE, we found that 58% of NDAF patients met the ≥5.5-hour criterion for elevated TE risk. Taken together, these findings indicate that clinically important amounts of AT/AF occur in a substantial proportion of patients with previous TE but no previous diagnosis of AF.
By defining NDAF as a minimum of 5 minutes of device-detected AT/AF on a given day, it is possible that we may have underreported the true incidence of NDAF. Implantable devices typically require atrial arrhythmias to persist for ≈30 seconds for detection to occur. Relying solely on device detection of any AT/AF as the threshold for diagnosing NDAF would have increased the incidence of NDAF among those with previous TE to 36% in the present analysis. However, such short durations of AT/AF are more likely to include false-positive detections.14
Previous studies have shown that the probability of detecting AT/AF via intermittent monitoring decreases as the amount of AT/AF present decreases.3 Without continuous arrhythmia monitoring, identification of patients with low AT/AF burden may be delayed. In the present study, we found that the majority of NDAF patients (73%) had AT/AF on <10% of follow-up days, making it highly unlikely to be detected by standard monitoring techniques. Whereas we obtained continuous heart rhythm monitoring using implanted cardiac devices, small subcutaneous devices are now available to provide similar heart rhythm surveillance. These may facilitate the detection of AT/AF in high-risk patients, such as those with cryptogenic stroke, potentially leading to improved prevention of recurrent stroke.
This study was conducted primarily in patients with a remote previous TE who had a clinical indication for an implantable pacemaker or defibrillator; therefore, these results may not be representative of all patients with previous TE and, in particular, may not apply to patients with cryptogenic stroke in whom monitoring is typically performed in close proximity to the TE. However, the rates of NDAF detected over the initial few days or weeks in this study were similar to previous studies in nondevice patients. Another limitation is that some patients included in our analysis may have had AF that was known but not documented prior to study enrollment. However, our exclusion of patients who were already using warfarin or antiarrhythmic drugs likely excluded the majority of patients with known previous AF. It is also possible that AF may have been detected clinically during follow-up, thereby reducing the incremental diagnostic value of continuous monitoring. However, because only 11% of patients had AT/AF on the majority of follow-up days, this is unlikely except for those with the greatest AT/AF burden. A final limitation is that the absence of electrograms for all episodes precluded us from verifying the precise diagnosis of the atrial arrhythmias. Previous reports, though, suggest that most arrhythmia oversensing that can result in false-positive detection of AT/AF is eliminated by excluding episodes <5 minutes,14 which was the duration criteria for the current study.
In a cohort of patients with previous TE, no known AF, and requiring a cardiac rhythm device, 28% of patients were identified as having NDAF via a device with continuous monitoring capabilities. Whereas AT/AF occurred on <10% of follow-up days in 73% of NDAF patients, it also persisted for at least 6 hours on at least 1 day in the majority of NDAF patients. These finding highlight not only the difficulty of detecting AT/AF with traditional methods but also the importance of identifying AT/AF in this population.
Sources of Funding
This study was funded by Medtronic.
P.D.Z., J.L.K., and C.E.H. are employees and stockholders of Medtronic. T.V.G. is a consultant for Medtronic and a speaker for Medtronic, St. Jude Medical, and Boston Scientific. E.G.D. is a consultant for Medtronic and a speaker for Medtronic, St. Jude Medical, and Boston Scientific. D.G.W. is a consultant for Medtronic and Biotronik. D.E.S. is a consultant for Medtronic. M.D.E. is a consultant for Medtronic.
- Received October 27, 2009.
- Accepted November 5, 2009.
Page RL, Wilkinson WE, Clair WK, McCarthy EA, Pritchett EL. Asymptomatic arrhythmias in patients with symptomatic paroxysmal atrial fibrillation and paroxysmal supraventricular tachycardia. Circulation. 1994; 89: 224–227.
Tayal AH, Tian M, Kelly KM, Jones SC, Wright DG, Singh D, Jarouse J, Brillman J, Murali S, Gupta R. Atrial fibrillation detected by mobile cardiac outpatient telemetry in cryptogenic TIA or stroke. Neurology. 2008; 71: 1696–1701.
Jabaudon D, Sztajzel J, Sievert K, Landis T, Sztajzel R. Usefulness of ambulatory 7-day ECG monitoring for the detection of atrial fibrillation and flutter after acute stroke and transient ischemic attack. Stroke. 2004; 35: 1647–1651.
Glotzer TV, Daoud EG, Wyse DG, Singer DE, Holbrook R, Pruett K, Smith K, Hilker CE. Rationale and design of a prospective study of the clinical significance of atrial arrhythmias detected by implanted device diagnostics: the TRENDS study. J Interv Card Electrophysiol. 2006; 15: 9–14.
Fuster V, Rydén LE, Asinger RW, Cannom DS, Crijns HJ, Frye RL, Halperin JL, Kay GN, Klein WW, Lévy S, McNamara RL, Prystowsky EN, Wann LS, Wyse DG, Gibbons RJ, Antman EM, Alpert JS, Faxon DP, Fuster V, Gregoratos G, Hiratzka LF, Jacobs AK, Russell RO, Smith SC Jr, Klein WW, Alonso-Garcia A, Blomström-Lundqvist C, de Backer G, Flather M, Hradec J, Oto A, Parkhomenko A, Silber S, Torbicki A. ACC/AHA/ESC guidelines for the management of patients with atrial fibrillation: Executive summary. Circulation. 2001; 104: 2118–2150.
Calkins H, Brugada J, Packer DL, Cappato R, Chen SA, Crijns HJ, Damiano RJ Jr, Davies DW, Haines DE, Haissaguerre M, Iesaka Y, Jackman W, Jais P, Kottkamp H, Kuck KH, Lindsay BD, Marchlinski FE, McCarthy PM, Mont JL, Morady F, Nademanee K, Natale A, Pappone C, Prystowsky E, Raviele A, Ruskin JN, Shemin RJ. HRS/EHRA/ECAS expert Consensus Statement on catheter and surgical ablation of atrial fibrillation: recommendations for personnel, policy, procedures and follow-up. A report of the Heart Rhythm Society (HRS) Task Force on catheter and surgical ablation of atrial fibrillation. Heart Rhythm. 2007; 4: 816–861.
Glotzer TV, Hellkamp AS, Zimmerman J, Sweeney MO, Yee R, Marinchak R, Cook J, Paraschos A, Love J, Radoslovich G, Lee KL, Lamas GA. Atrial high rate episodes detected by pacemaker diagnostics predict death and stroke: report of the Atrial Diagnostics Ancillary Study of the MOde Selection Trial (MOST). Circulation. 2003; 107: 1614–1619.