Objective. Patient-activated transtelephonic electrocardiographic event monitors (TTMs) are often used for the evaluation of children and adolescents with suspected arrhythmias. Since their introduction 25 years ago, there has been little inquiry quantifying the usefulness of TTMs for pediatric patients. The objective of this study was to measure the utility of TTMs for children and adolescents with symptoms of a possible cardiac rhythm disturbance.
Methods. Medical records of all patients who received TTMs from C.S. Mott Children’s Hospital Electrocardiography Laboratory between February 1, 1993, and October 31, 2000, were reviewed. Patients with pacemakers, known arrhythmias, or age older than 18 years were excluded. Indications for monitoring included palpitations with or without other symptoms (N = 420), chest discomfort (N = 43), and presyncope or syncope (N = 32).
Results. A total of 495 studies (patient mean age: 10.2 ± 4.3 years; range: 0.1–17.9 years; 48% male) met inclusion criteria. Monitoring was performed for 1 to 1021 consecutive days (mean: 103 ± 97). Fifty-two percent (N = 257) of patients failed to transmit an electrocardiogram while experiencing symptoms. Fewer boys transmitted electrocardiograms (N = 100/238). Of 238 symptomatic patients, 15% (N = 35; mean age: 11.4 ± 4.7 years; range: 0.1–17.4 years; 51% male) had supraventricular tachycardia (SVT). No other significant arrhythmia that may warrant treatment was identified. All patients with SVT had palpitations. No patients with isolated chest discomfort, presyncope, or syncope had SVT (N = 75). SVT was documented more frequently in patients with postevent (N = 35/464) than loop recorders (N = 0/31). Of those with SVT, 71% (N = 25) and 91% (N = 33) transmitted events within 4 and 16 weeks, respectively. Follow-up for 1 to 108 months (mean: 32 ± 25; median: 26) in 53% (243 of 460) of patients without SVT uncovered a 3% (N = 7) rate of subsequent SVT detection. The overall sensitivity of the TTM test was 83% (35 of 42) for detection of SVT. The sensitivity of studies theoretically limited to 4 and 16 weeks would be 60% (25 of 42) and 79% (33 of 42), respectively. The negative predictive value of the TTM study was 99% in our patient population. The negative predictive value of tests theoretically limited to 4 and 16 weeks would be 96% and 98%, respectively. TTM studies of 2 weeks’ duration were most cost-effective in terms of total diagnostic yield. In contrast, studies of 4 weeks’ duration were most cost-efficient for SVT detection.
Conclusions. TTMs are useful for the evaluation of children and adolescents with palpitations but not with isolated chest pain, syncope, or presyncope. In this study, girls were more likely to transmit events. The sensitivity of TTMs for detection of SVT was 83%. The negative predictive value of the TTM test was 99%. Monitoring for longer than 16 weeks did not increase test sensitivity. Studies of 4 weeks’ duration proved most cost-effective for SVT detection.
Symptoms of cardiac rhythm disturbance in children and adolescents may include palpitations or the sensation of a rapid heartbeat, chest discomfort, syncope, or presyncope. Electrocardiographic documentation of cardiac rhythm during symptoms is imperative for accurate diagnosis and treatment. Because many patients’ symptoms are infrequent and episodic, transtelephonic electrocardiographic event monitors (TTMs) are often used for the evaluation of children with suspected arrhythmia.1–4
TTMs are patient-activated portable transmitters that record and translate electrocardiograms (ECGs) into audible output. Audible transmissions are transferred over telephone wires and retranslated into ECGs. With continuous-loop devices, electrodes are worn constantly. Hence, continuous-loop monitors can document cardiac electrical activity before activation (up to 2 minutes). In contrast, electrodes are placed after symptoms begin with postevent monitors, negating any ability to record previous ECG activity. When symptoms are intermittent, TTMs show better correlation with pediatric patients’ complaints than Holter monitors.5
TTMs have proved useful and cost-effective for the evaluation of adults with palpitations, syncope, or presyncope.6–9 However, there has been little inquiry measuring the utility of TTMs for young patients. The purpose of this study was to determine the diagnostic yield of TTMs used in the evaluation of children and adolescents with suspected cardiac arrhythmia. Data presented can be used to develop cost-effective strategies for TTM use in the pediatric population.
Approval for this study was obtained from the institutional review board at the University of Michigan. Medical records of all patients who were provided TTMs by C.S. Mott Children’s Hospital Electrocardiography Laboratory between February 1, 1993, and October 31, 2000, were reviewed. Every test was ordered by a staff pediatric cardiologist. Patients with pacemakers, known cardiac arrhythmias, incomplete records, or age >18 years were excluded.
Monitoring Devices and Protocols
Monitoring was performed with continuous-loop (King of Hearts; Instromedix, San Diego, CA) or postevent recorders. Most postevent devices had memory capability (Cardiobeeper, Telecare AG, Berlin, Germany; CardioCare, Cardiac Datacorp, Forest Hills, NY; CardioDiary, Instromedix; ER300, Braemer, Minneapolis, MN; Heart Aide/Heart Aide Plus, Ralin, Beaverton, OR; Wristrecorder, Ralin). One postevent device did not have memory capability (“nonmemory,” requiring simultaneous recording and transmission of the ECG; Teletrace, Medtronic, Minneapolis, MN). Cardiologists’ requests and device availability determined which type of monitor was dispensed. Continuous-loop recorders were reserved for patients with symptoms of brief duration (5–20 seconds), symptoms occurring 1 to 2 times per week, or those with syncope. Nonmemory postevent recorders were reserved for patients with symptoms of relatively long duration (>5 minutes). Length of the study was defined as days from device dispensation to return.
A baseline TTM ECG was obtained for nearly all patients. For the purpose of this study, an event was defined as a legible ECG tracing both recorded during symptoms and successfully transmitted. In cases in which TTM ECGs were illegible, all but 1 patient successfully transmitted at least 1 legible ECG during subsequent symptoms. Two investigators reviewed and agreed on interpretation of all ECGs.
Events were grouped into 1 of 2 categories for the purpose of data analysis: clinically significant arrhythmias that may warrant treatment (supraventricular tachycardia [SVT], atrial flutter, atrial fibrillation, junctional tachycardia, ventricular tachycardia, frequent atrial or ventricular ectopy, and bradycardia) or benign cardiac rhythms that do not warrant treatment (normal sinus rhythm, sinus arrhythmia, sinus tachycardia, low atrial rhythm, wandering atrial pacemaker, occasional atrial premature beats, and occasional ventricular premature beats). Both clinically significant and benign TTM event recordings were considered useful for the purpose of patient diagnosis.
Statistics were calculated using Excel 97 (Microsoft, Redmond, WA). All values are expressed as percentages of total or mean ± standard deviation. Comparisons between discrete variables were performed using the χ2 statistic. Comparisons between continuous variables were performed using Student’s t test. P < .05 was considered significant.
Cost analyses were performed assuming a fixed cost (x) per patient per week of monitoring. Cost per useful diagnosis calculation: .
Cost per SVT diagnosis calculation: .
Of 850 consecutive tests performed by C.S. Mott Children’s Hospital during the study period, 495 met criteria for inclusion. Excluded were 165 patients with pacemakers, 117 patients with known arrhythmias, 47 patients older than 18 years, and 26 patients with incomplete medical records. Patients ranged in age from 0.1 to 17.9 years, and 238 (48%) were male (Table 1). There was no significant difference between boys and girls for age, type of symptoms, frequency or type of congenital heart disease (CHD), medical diagnoses, medication use, type of monitor used, or length of study. Ninety-seven patients had CHD (mean age: 10.5 ± 4.9 years; range: 0.1–17.9 years; Table 2). There was no significant difference for age, type of symptoms, type of monitor prescribed, or length of study between those with and without CHD.
Patients were monitored for a median of 83 consecutive days (range: 1–1021; mean: 103 ± 97). Half of the patients never transmitted an event (N = 257). Fewer boys (N = 100) than girls (N = 138) recorded events (P = .02). Patients with continuous-loop devices transmitted events more frequently than those with postevent recorders (P = .005). Patients of all ages were equally likely to record events.
Thirty-five (15%) of the 238 patients with events had SVT (Table 3). No other significant arrhythmia was identified. The mean age of patients with SVT was similar to the total (Table 1). Boys and girls were equally likely to document SVT. There was no difference in frequency of CHD, medical diagnoses, or medication use between those with or without SVT.
All 35 patients with SVT complained of palpitations with or without other symptoms. Three patients with SVT had syncope or near-syncope associated with palpitations. No patient with isolated chest pain, syncope, or presyncope had a significant arrhythmia (P = .005). Despite a higher rate of event transmission, SVT detection was less frequent for patients with loop than postevent recorders (P = .03). The rate of SVT detection was higher for patients with nonmemory- than memory-capable postevent devices (P = .01).
Time to first event transmission ranged from 1 to 473 days, with a mean of 25 ± 46 days. Patients with benign events were more likely than those with SVT to transmit ECGs within 3 weeks (P < .001; Table 4). Of those with events, 75% transmitted within 4 weeks and 90% did so within 16 weeks. Cumulative diagnoses per week of monitoring and projected cost per TTM diagnosis are listed in Table 4. Monitoring for 4 weeks was most cost-effective for SVT detection. Tests limited to 2 weeks’ duration were most cost-effective for total useful diagnostic yield.
Follow-up (review of all available subsequent medical records from office visits and tests performed by cardiologists, primary care physicians, and emergency department physicians) was obtained for 243 (53%) of the 460 patients without documented SVT. Length of follow-up ranged from 1 to 108 months (mean: 32 ± 25; median: 26). None of the patients died. Seven (3%) subsequently received a diagnosis of a significant arrhythmia, all of whom had SVT. Three of the 7 patients had transmitted TTM events demonstrating sinus rhythm during the study time period. The remaining 4 patients did not transmit an event during the study. Four of the 7 documented SVT using a TTM subsequent to the study. Two of the patients documented SVT during subsequent emergency department visits. A single patient later underwent an invasive electrophysiology study that demonstrated SVT.
The sensitivity of the TTM study for SVT detection was 83% (Table 4). The sensitivity of studies limited to 6 or fewer weeks would be 64% or less. The negative predictive value of the TTM study was 99% in our patient population. The negative predictive value of studies limited to 4 and 16 weeks would be 96% and 98%, respectively.
Our data indicate that TTMs provide useful diagnoses for 48% of children and adolescents with suspected arrhythmia. The remainder of patients failed to transmit a single legible ECG during symptoms. Two previous, smaller pediatric studies reported comparable results.2,4 This experience is in contrast with several investigations in adult populations that report higher diagnostic yields of 57% to 84% for patients with suspected arrhythmia.6,7,9,10
Fifteen percent of the patients with recorded events had SVT. There are several plausible reasons for why no other significant arrhythmia was identified. First, other cardiac rhythm disturbances are less prevalent in the pediatric population. Second, other arrhythmias are often diagnosed by alternative methods, including Holter monitor, surface ECG, or an exercise test, because symptoms are more persistent or are reliably induced by physical activity.11
Patients with postevent recorders documented SVT more frequently than patients with continuous-loop devices. This finding is relevant because loop monitors cost more than postevent recorders. Moreover, patients with nonmemory postevent recorders documented SVT more often than those with memory-capable devices. Both of these observations can be explained by distribution protocols at C.S. Mott Children’s Hospital, rather than inherent differences in the monitors or monitor–-patient interactions. Patients with symptoms of brief duration, occurring 1 or 2 times per week, or syncope preferentially receive continuous-loop monitors; these patients are less likely to have SVT. In a similar manner, patients with palpitations lasting >5 minutes, who preferentially received nonmemory recorders, are more likely to have SVT.
We also found that girls were more likely than boys to transmit events. Because the rate of SVT detection was similar among the sexes, this means that girls transmitted benign events more frequently than boys. Given that male and female patients did not significantly differ for age, types of symptoms, frequency of CHD, medical diagnoses, medication use, type of monitor used, or length of study, the causes of this male-female discrepancy are unclear.
Our data show that all patients with chest pain, syncope, or presyncope without palpitations transmitted benign events. This observation correlates with data from previous pediatric studies.2,5 When limited to the evaluation of syncope and presyncope, rather than palpitations, adult TTM studies also report a lower diagnostic yield of 25% to 55%.6,7,12
Limitations of this study include the retrospective design and a lack of medical follow-up for nearly half of patients without SVT. Incomplete follow-up could lead to underestimation of the true rate of subsequent arrhythmia diagnosis and hence error in calculation of sensitivity and negative predictive value of the TTM studies. Future investigations, with a prospective design and more complete long-term follow-up, may provide more reliable estimates of TTM sensitivity and predictive value for pediatric patients with suspected arrhythmia.
Cost-effectiveness of time-limited TTM studies has been evaluated in the adult population.7,8,10 The majority of adult reports recommend 2-week test duration for optimal cost-efficiency. However, a minority of adult studies found 1-week test duration most cost-efficient. This contrasts with our data, which demonstrate optimal cost-efficiency of 4-week TTM tests for arrhythmia detection in children and adolescents.
When used for evaluation of children and adolescents with palpitations, TTMs yielded a useful diagnosis in half of patients. In this study, girls were more likely to transmit events. The only clinically significant arrhythmia found was SVT. TTMs are not useful for pediatric patients with isolated chest pain, syncope, or presyncope. Indeed, none of these patients documented a significant arrhythmia. The sensitivity of the TTM test for the detection of SVT was 83%. The sensitivity of tests theoretically limited to 6 or fewer weeks was significantly less for SVT detection. The negative predictive value of the TTM was 99% in our patient population. Tests theoretically limited to 4 weeks had a negative predictive value of 96% and were most cost-effective for SVT detection.
- Received April 8, 2003.
- Accepted July 15, 2003.
- Reprint requests to (E.V.S.) Pediatric Cardiology, University of Michigan, L1242 Women’s Box 0204, 1500 East Medical Center Dr, Ann Arbor, MI 48109-0204. E-mail:
- Williams RL. Use of transtelephonic electrocardiography in patients with symptoms suggesting cardiac arrhythmia. Pediatrics.1978;61 :493– 495
- ↵Fyfe DA, Holmes DR Jr, Neubauer SA, Feldt RH. Transtelephonic monitoring in pediatric patients with clinically suspected arrhythmias. Clin Pediatr (Phila).1984;23 :139– 143
- Copyright © 2004 by the American Academy of Pediatrics