Download Arrhythmia/Electrophysiology

Arrhythmia/Electrophysiology
Dual-Chamber Versus Single-Chamber Detection
Enhancements for Implantable Defibrillator
Rhythm Diagnosis
The Detect Supraventricular Tachycardia Study
Paul A. Friedman, MD; Robyn L. McClelland, PhD; William R. Bamlet, MS; Helbert Acosta, MD;
David Kessler, MD; Thomas M. Munger, MD; Neal G. Kavesh, MD; Mark Wood, MD;
Emile Daoud, MD; Ali Massumi, MD; Claudio Schuger, MD; Stephen Shorofsky, MD;
Bruce Wilkoff, MD; Michael Glikson, MD
Downloaded from http://circ.ahajournals.org/ by guest on June 11, 2017
Background—Delivery of inappropriate shocks caused by misdetection of supraventricular tachycardia (SVT) remains a
substantial complication of implanted cardioverter/defibrillator (ICD) therapy. Whether use of optimally programmed
dual-chamber ICDs lowers this risk compared with that in single-chamber ICDs is not clear.
Methods and Results—Subjects with a clinical indication for ICD (n⫽400) at 27 participating centers received
dual-chamber ICDs and were randomly assigned to strictly defined optimal single- or dual-chamber detection in a
single-blind manner. Programming minimized ventricular pacing. The primary end point was the proportion of SVT
episodes inappropriately detected from the time of programming until crossover or end of study. On a per-episode basis,
42% of the episodes in the single-chamber arm and 69% of the episodes in the dual-chamber arm were due to SVT.
Mortality (3.5% in both groups) and early study withdrawal (14% single-chamber, 11% dual-chamber) were similar in
both groups. The rate of inappropriate detection of SVT was 39.5% in the single-chamber detection arm compared with
30.9% in the dual-chamber arm. The odds of inappropriate detection were decreased by almost half with the use of the
dual-chamber detection enhancements (odds ratio, 0.53; 95% confidence interval, 0.30 to 0.94; P⫽0.03).
Conclusions—Dual-chamber ICDs, programmed to optimize detection enhancements and to minimize ventricular pacing,
significantly decrease inappropriate detection. (Circulation. 2006;113:2871-2879.)
Key Words: arrhythmia 䡲 defibrillation 䡲 heart arrest 䡲 tachyarrhythmias
T
he implanted cardioverter/defibrillator (ICD) is highly
effective in decreasing mortality due to cardiac arrhythmias in high-risk patients.1,2 However, delivery of inappropriate shocks caused by misclassification of rapidly conducted supraventricular tachycardia (SVT) as ventricular
tachycardia (VT) remains a substantial complication of ICD
therapy, affecting 8% to 40% of patients.3– 8 Inappropriate
shocks can lead to pain, anxiety, depression, impaired quality
of life, proarrhythmia, and poor tolerance of life-saving ICD
therapy.9
Intuitively, dual-chamber detection enhancements, which use
atrial and ventricular intracardiac information to formulate a
diagnosis, should be superior to single-chamber, ventricularonly detection enhancements. However, early nonrandomized
studies and subsequent small randomized studies failed to
show any superiority of dual-chamber over single-chamber
diagnosis.3–5 Moreover, the finding that backup ventricular
pacing is not inferior to dual-chamber pacing in ICD recipients
who do not require pacing10 and recent studies showing the
effectiveness of single-chamber ICDs in prophylactic sudden
death prevention2 have led to a decrease in dual-chamber ICD
use. In the United States, the Center for Medicare and Medicaid
Services ruled that “providers must justify the medical necessity
of devices other than those with a single lead.”11
Editorial p 2862
Clinical Perspective p 2879
The most desirable method of limiting inappropriate therapies
is appropriate rhythm detection and discrimination by the ICD.
Received October 19, 2005; revision received February 17, 2006; accepted April 7, 2006.
From the Division of Cardiovascular Diseases (P.A.F., T.M.M.) and the Division of Biostatistics (R.L.M., W.R.B.), Mayo Clinic, Rochester, Minn;
Electrophysiology, Trinity Medical Center, Rock Island, Ill (H.A.); Electrophysiology and Cardiac Pacing, Austin Heart, Austin, Tex (D.K.);
Cardiology/Electrophysiology, Watson Clinic, Lakeland, Fla (N.G.K.); Internal Medicine/Cardiology, Virginia Commonwealth University, Richmond
(M.W.); Cardiovascular Disease, MidWest Research Foundation, Riverside Methodist Hospital, Columbus, Ohio (E.D.); Center for Cardiac Arrhythmias
and Electrophysiology, Texas Heart Institute, Houston (A.M.); Cardiology, Henry Ford Hospital, Detroit, Mich (C.S.); Maryland Heart Center, University
of Maryland Medical Center, Baltimore (S.S.); Cardiovascular Disease, The Cleveland Clinic Foundation, Cleveland, Ohio (B.W.); and Electrophysiology
and Pacing, Heart Institute, Sheba Medical Center, Tel Hashomer, Ramat Gan, Israel (M.G.).
Correspondence to Paul A. Friedman, MD, Division of Cardiovascular Diseases, Mayo Clinic, 200 First St SW, Rochester, MN 55905.
© 2006 American Heart Association, Inc.
Circulation is available at http://www.circulationaha.org
DOI: 10.1161/CIRCULATIONAHA.105.594531
2871
2872
Circulation
June 27, 2006
Figure 1. Default programming detection
enhancement parameters. A indicates atrial
rate; FIB, ventricular fibrillation; and V, ventricular rate.
Downloaded from http://circ.ahajournals.org/ by guest on June 11, 2017
Detection enhancement programming is complex, and one
reason for the lack of superiority of dual-chamber over
single-chamber enhancements in previous studies may have
been the use of nonoptimal enhancement settings. Perhaps
more important, improvements in atrial sensing and
morphology-based algorithms, which may substantially improve dual-chamber detection, have recently become available.12–15 We hypothesized that the optimal use of the latest
generation dual-chamber detection enhancements would decrease inappropriate detection of SVT compared with optimal
single-chamber detection enhancements in a broad population
of ICD recipients. If true, this would indicate that dualchamber ICDs are preferable to single-chamber devices in
most ICD recipients because of a decrease in the morbidity
due to inappropriate therapy.
younger than 18 years, Mobitz II or greater heart block, previous
atrioventricular (AV) node ablation, permanent atrial fibrillation or
flutter, preexisting separate pacemaker pulse generator that was not
to be explanted, life expectancy of ⬍1 year or on a transplant waiting
list, intra-aortic balloon pump, inotropic drug (not digitalis) necessary for hemodynamic support, chronic serious bacterial infection,
inability to receive nonthoracotomy ICD, or inability to program
device as per protocol.
Subjects were randomly assigned to device programming for
ventricular-only or dual-chamber detection within 3 days of ICD
implantation in a single-blind manner (subject not aware). The
randomization was balanced within a site in blocks, with the use of
sequentially numbered, opaque sealed envelopes. Subjects were
followed up for 6 months, with a patient visit, device function
assessment, and data collection at baseline and at 3 follow-up visits
scheduled at 2 to 4 weeks, 3 months, and 6 months after programming. Data from unscheduled visits were also retrieved for analysis.
Devices and Programming
Methods
Study Conduct
The study was investigator initiated and managed. Data were
independently managed by Mayo Clinical Trial Services. A steering
committee of 2 electrophysiologists (P.A.F., M.G.), 2 statisticians
(R.L.M., W.R.B.), and a coordinating center representative met
approximately every 3 months to review safety and accrual data. The
members of the steering committee wrote the article and assume
overall responsibility for the data and analysis. The study was
supported by an unrestricted grant from St Jude Medical (St Paul,
Minn), which had no role in study design, analysis, or interpretation.
The institutional review boards of all participating institutions
approved the study.
Study Design
Subjects were enrolled between December 2002 and October 2004 at
27 participating centers. Eligible patients had a clinical indication for
ICD therapy and received a dual-chamber defibrillator. Patients were
excluded if any of the following conditions were present: age
All subjects received a dual-chamber ICD manufactured by St Jude
Medical. Systematic efforts were made during implantation to
eliminate far-field R wave oversensing, and the absence or presence
of far-field R wave oversensing was recorded.
The optimal detection enhancement programming used in this
study was defined by a previous analysis of a large episode
database14 and determined by the investigators, with manufacturer
input. Detection enhancement programming is summarized in Figure
1. Programming was specified to minimize ventricular pacing.
Defibrillator programmers were preloaded with study templates
for dual-chamber and ventricular-only detection enhancement to
facilitate uniform programming across all subjects. Dual-chamber
detection used rate branch classification based on the relative rates in
the atria (A) and ventricles (V; V⬎A, V⫽A, or V⬍A) before
application of detection enhancements to maximize enhancement
performance, as described in Figure 1. In all subjects, use of
extended timers that delivered therapy after a tachyarrhythmia
duration was met regardless of device diagnosis was disallowed to
prevent degradation of detection specificity. At least 1 detection zone
with rate enhancements was programmed. To ensure a large window
Friedman et al
Dual-Chamber vs Single-Chamber ICD Detection
2873
Figure 2. Patient enrollment, status, and
episodes of arrhythmia.
Downloaded from http://circ.ahajournals.org/ by guest on June 11, 2017
for rhythm discrimination, a VT detection rate no faster than 150
beats per minute (bpm) and a ventricular fibrillation (VF) detection
rate no slower than 200 bpm, with application of discriminators up to
the VF cutoff, were required. It was strongly suggested that devices
in all subjects be programmed to DDD or DDI with a lower rate of
50 to 60 bpm with an AV interval as long as possible to permit native
conduction when present. Use of antitachycardia pacing (ATP) in the
VT zone was strongly encouraged to minimize painful shocks.
However, to maximize detection programming compliance, all
therapy programming was at the discretion of the investigator.
Optimization of subparameters without changing detection type
(single- or dual-chamber) was permitted after a subject received an
inappropriate therapy. Optimization events were recorded.
End Points
The primary end point was the proportion of SVT episodes inappropriately detected (number of inappropriately detected SVT episodes
divided by total number of SVT episodes) from the time of
programming until crossover or end of study. Inappropriately detected SVT was defined as any SVT episode that met ventricular rate
detection criteria and was inappropriately classified as VT. Any
episode lasting long enough to trigger detection resulting in a stored
electrogram was reviewed and classified in an unblinded manner by
the site investigator and in a blinded manner by a core group
investigator; discrepancies in the overall type (SVT versus VT) were
adjudicated by an additional blinded review. Disagreements concerning arrhythmia subtypes (type of SVT or VT) were not adjudicated,
and results presented are based on the opinion of the blinded
reviewer. Secondary end points included time to first inappropriately
treated episode, VT/VF sensitivity, number of arrhythmia-related
hospitalizations or clinic visits, and the early termination rate. A post
hoc analysis was performed to determine how selection of a VT
detection rate ⬎150 bpm would alter detection and discrimination.
Sample Size and Power
We estimated that 30% of patients would have at least 1 SVT episode
during the 6-month follow-up period. We predicted that the singlechamber device would inappropriately detect 30% of the SVT
episodes.3 In contrast, we believed that if only 10% of all SVT
episodes for the dual-chamber group were inappropriately detected,
this would constitute a minimal clinically important decrease of 20
percentage points. Because only 30% of the subjects were anticipated to have any SVT episodes, the total number of patients was
increased to 400 (⬇200 subjects per arm) to maintain an 80% power.
Many subjects had ⬎1 episode, which increased the study power.
Statistical Analysis
Patient characteristics were compared between study arms with the
use of 2-sample t tests for continuous variables and ␹2 tests for
categorical variables. If the number of occurrences of a categorical
variable was very small, we verified the results using the Fisher exact
test.
The primary end point was the proportion of SVT episodes
inappropriately detected by the device within the first 6 months after
implantation. A logistic regression model was fit to investigate the
association between treatment group (dual- versus single-chamber)
and the odds of inappropriate detection. The correlation between
multiple episodes within the same subject was accounted for by
using a generalized estimating equations (GEE) approach, with a
compound symmetrical correlation structure.16,17
Note that the GEE model estimates an odds ratio, not a relative
risk. The odds ratio does not approximate the relative risk in this
study because the end point is not rare. We report the raw
per-episode rates of inappropriate detection and the GEE-adjusted
rates, which move the per-episode rate estimates toward the average
per-subject rate based on the within-subject correlation and the GEE
model. Time to first inappropriate detection was examined graphically with Kaplan-Meier plots and with Cox proportional hazards
regression models including a term for programming arm. Time was
indexed in 2 ways: by calendar time and by the number of SVT
episodes observed.
␹2 tests were used to compare rates of early study withdrawal and
crossovers between study arms. The total number of hospitalizations
and clinic visits related to arrhythmias or inappropriate therapies
delivered by the ICD was compared by Wilcoxon rank sum tests.
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.
Results
Patients and Events
During the study period, 400 subjects were prospectively
randomly assigned to ventricular-only (n⫽199) or dualchamber (n⫽201) detection. A total of 1253 arrhythmia
episodes occurred in 103 of the 199 subjects in the ventricular-only detection arm, and 1090 episodes occurred in 104 of
the 201 subjects in the dual-chamber detection arm (Figure
2). Sixty-two subjects (31%) in the ventricular-only arm and
75 subjects (37%) in the dual-chamber arm had SVTs; the
risk of SVT occurrence was similar in patients with and
without a history of SVT. On a per-episode basis, 42% of the
episodes in the single-chamber arm and 69% of the episodes
in the dual-chamber arm were due to SVT (P⫽0.06). Mortality (3.5% in both groups) and overall early study withdrawal (noncompletion of 6 months of follow-up due to any
2874
Circulation
June 27, 2006
Patient Characteristics and Medications by Study Group
Detection
Patient Characteristic/Medication
Age at implantation, y
Ventricular Only
(n⫽199)
Dual Chamber
(n⫽201)
P
0.46
65.1⫾11.3
64.3⫾11.3
Men
156 (78)
163 (81)
0.50
ICD indication, primary prevention
132 (66)
136 (68)
0.78
Left ventricular ejection fraction, %
32⫾13
32⫾13
0.95
Arrhythmia history
Atrial fibrillation
Atrial flutter
Other SVT
40 (20)
36 (18)
0.59
8 (4)
12 (6)
0.36
12 (6)
8 (4)
0.35
VT
128 (64)
119 (59)
0.32
VF
21 (11)
31 (15)
0.14
Catheter ablation (not AV node)
1 (0.5)
8 (4)
0.02
Surgical maze
2 (1)
1 (0.5)
0.56
Downloaded from http://circ.ahajournals.org/ by guest on June 11, 2017
Cardiac history
Hypertension
122 (61)
111 (55)
0.22
Coronary artery disease
161 (81)
163 (81)
0.96
Percutaneous coronary intervention
80 (40)
91 (45)
0.31
Coronary artery bypass graft
82 (41)
90 (45)
0.47
Congestive heart failure
98 (49)
109 (54)
0.32
Dilated cardiomyopathy
36 (18)
30 (15)
0.39
AV node agents
178 (89)
184 (92)
0.47
␤-Blocker
162 (81)
168 (84)
䡠䡠䡠
AV nodal slowing
12 (6)
15 (7)
䡠䡠䡠
Non–AV nodal slowing (dihydropyridines)
19 (10)
18 (9)
䡠䡠䡠
䡠䡠䡠
0.37
Calcium channel blocker
Digoxin/digitalis
53 (27)
70 (35)
Antiarrhythmic drugs
59 (30)
68 (34)
Sotalol
14 (7)
11 (5)
䡠䡠䡠
Amiodarone
44 (22)
55 (27)
䡠䡠䡠
Dofetilide
0 (0)
1 (0.5)
䡠䡠䡠
Other
9 (5)
9 (4)
䡠䡠䡠
Values are mean⫾SD or number of patients (percentage).
cause, including death; 14% ventricular-only, 11% dualchamber) were similar in both groups (P⫽0.46). More
patients crossed over from ventricular-only to dual-chamber
detection (n⫽17) than the reverse (n⫽2; P⬍0.001), but the
overall crossover rate was small.
Demographic and clinical data for the 2 treatment groups
are shown in the Table. No clinically important differences
were seen between the groups at baseline. At baseline, the
mean left ventricular ejection fraction was 32⫾13%, the ICD
was implanted in 67% of subjects for primary sudden death
prevention, and 25% of subjects had a known history of SVT.
Use of AV node conduction slowing medications and antiarrhythmic medications was the same in both groups at baseline
(Table) and at each follow-up visit. Far-field R wave oversensing at device check occurred in 3% of patients in the
dual-chamber arm and 9% of subjects in the ventricular-only
arm.
Primary End Point: Inappropriate Detection
The rate of inappropriate detection of SVT was 39.5% (210 of
531 SVT episodes) (GEE rate estimate, 46.5%) in the
ventricular-only detection arm compared with 30.9% (232 of
750 episodes) (GEE rate estimate, 32.3%) in the dualchamber arm (Figure 3). After we accounted for the positive
correlation between multiple episodes occurring within the
same subject, the odds of inappropriate detection were
decreased by almost half with the use of the dual-chamber
detection enhancements (odds ratio, 0.53 [95% confidence
interval, 0.30 to 0.94]; P⫽0.03). With the exception of sinus
tachycardia, for which single- and dual-chamber detection
were similar, dual-chamber detection consistently had a lower
rate of inappropriate detection than ventricular-only detection
across all arrhythmia subtypes (Figure 3). However, differences in classification of arrhythmia subtype between the site
investigator and blinded reviewers were not adjudicated, and
Friedman et al
Downloaded from http://circ.ahajournals.org/ by guest on June 11, 2017
Figure 3. Rate of inappropriate detection of SVT, with breakdown by arrhythmia subtype (atrial fibrillation, atrial flutter/atrial
tachycardia [tach], sinus tachycardia, or other), for subjects with
single- or dual-chamber detection. “Other” arrhythmias include
atrial tachycardia, junctional tachycardia, AV nodal reentrant
tachycardia, and AV reentrant tachycardia. Subtype classification was based on blinded episode reviewer. The odds ratio
(OR) and probability value refer only to the overall comparison
of inappropriate detection of SVT.
these rates are presented only to examine consistency across
the subtypes. In general, rate branch sorting permitted differential application of the detection enhancements, resulting in
superior performance of dual-chamber algorithms, as shown
Dual-Chamber vs Single-Chamber ICD Detection
2875
in Figure 4. SVT occurred in the VF zone (in which detection
enhancements were not applied) in 17% of all inappropriately
detected episodes in the dual-chamber arm and 2% in the
single-chamber arm (P⬍0.001). Of all SVT episodes detected
as VF, the investigator classified 45% as atrial fibrillation,
10% as atrial flutter, and 45% as 1:1 SVT or “other.”
In terms of the statistical analysis, the within-subject
correlation was quite high (⬇0.50). This is likely because
subjects tend to experience the same types of arrhythmias
repeatedly. Thus, it is important to use a statistical model that
can handle correlated data. On a per-subject basis, among
those with some SVT episodes the median number of
episodes was 4 (range, 1 to 58). The average per-subject rate
of inappropriate detection was 51% for the single-chamber
group and 34% for the dual-chamber group.
Figure 5 depicts the percentage of subjects with at least 1
inappropriate detection based on the number of episodes
experienced. The cumulative number of recorded SVT episodes is used because detection enhancement effectiveness is
best assessed by the response to actual episodes, as opposed
to time to an episode. Subjects in the dual-chamber arm were
less likely to have an inappropriate detection for any given
number of recorded SVT episodes, although this difference
did not reach statistical significance (P⫽0.13). The median
number of episodes until the first inappropriate detection was
Figure 4. Examples of single-chamber and
dual-chamber detection. A, Stored electrogram of an atrial tachycardia episode (T)
inappropriately detected as VT in the
single-chamber arm, resulting in ATP therapy delivery. Episode of atrial tachycardia
begins with a premature atrial complex
(asterisk), which is followed by a regular
tachycardia. The arrhythmia has an abrupt
onset (indicating VT), is regular (favoring
VT), but has a morphology consistent with
SVT (check marks at arrow). Because this
patient had single-chamber detection, atrial
information is not used, and because 2 of
3 detection enhancements indicate VT,
therapy is delivered inappropriately. Note
that the arrhythmia was not terminated by
the ATP. This led to additional therapies
(not shown). B, Stored electrogram of an
appropriately diagnosed atrial tachycardia
episode resulting in withholding of therapy
in a patient in the dual-chamber arm.
Asterisk indicates initiating atrial premature
complex. The episode is detected in the
V⫽A rate branch indicated by the “I⫽” symbol (circled). In this rate branch, only morphology is programmed on, as per protocol.
Morphology indicates SVT (check marks at
arrows), so therapy is appropriately withheld.
Correct classification of SVTs with a 1:1 A to
V relationship is often problematic for ICDs
given their abrupt onset (mimicking VT) and
regular intervals (also mimicking VT). This
makes morphological discrimination potentially more useful in this group of arrhythmias. A indicates atrial electrogram; P,
sensed P wave; R, sensed R wave; S, sinus
beat; TB, tachycardia B; and V, ventricular
electrogram. Numbers indicate intervals
between P or R waves.
2876
Circulation
June 27, 2006
Figure 5. Number of SVT episodes until the first inappropriate
detection. Graph shows percentage of subjects with at least 1
inappropriate detection by total number of SVT episodes recorded (P⫽0.13 for difference between the 2 curves).
Downloaded from http://circ.ahajournals.org/ by guest on June 11, 2017
1.5 for the single-chamber group and 4 for the dual-chamber
group. The calendar time to first inappropriate detection of
SVT episodes was not significantly different between groups
(median, 103 days for ventricular-only versus 108 days for
dual-chamber; P⫽0.45).
Inappropriate Therapy Delivery
Inappropriate detection led to inappropriate therapy delivery
for the 2 groups, as depicted in Figure 6. The overall rate of
inappropriate therapy delivery (ATP or shock) of 33.0%
(GEE rate estimate, 38.3%) in the ventricular-only group was
significantly greater than that in the dual-chamber group
(24.8%; GEE rate estimate, 26.1%; P⫽0.02), for a 46%
decrease in the odds of inappropriate therapy. The odds of an
inappropriate shock were not significantly different (P⫽0.18)
because of the greater use of ATP in the ventricular-only arm.
VT Detection
Overall, of 937 VT episodes that occurred (615 in the
ventricular-only arm, 322 in the dual-chamber arm), 871
(93%) were detected. Of the 66 episodes that were not
detected, 16 were sustained (⬎30 seconds in duration): 9 in
the ventricular-only arm and 7 in the dual-chamber arm.
Thus, 98.2% of episodes were either detected or terminated
spontaneously in ⬍30 seconds. No VF episodes were missed
(of 12 in the ventricular-only arm and 10 in the dual-chamber
arm).
No difference was seen between the groups in the total
number of arrhythmia-related hospitalizations (6% of patients
in each group had at least 1 hospitalization) or additional
clinic visits (5% of ventricular-only subjects, 7% of dualchamber subjects). Optimization of detection enhancements
within each study group was also the same, occurring in 34
subjects (17%) in the ventricular-only group (19 due to
inappropriate detection of SVT, 6 due to underdetection of
VT/VF, and 9 not specified) and in 34 subjects (17%) in the
dual-chamber group (23 for inappropriate detection of SVT, 2
for underdetection of VT/VF, and 9 not specified).
Complications
In the entire study population, atrial lead fracture or failure
was reported in 1 patient (0.25%), atrial lead dislodgment in
4 patients (1%), and atrial sensing errors in 4 patients (1%).
There was no difference between groups in the rate of any
complication. Atrial sensing errors were uncommon, occurring in ⬍5% of patients in the dual-chamber group.
Discussion
Main Findings
Our study had 2 principal findings. First, the odds of
inappropriate detection of SVT as VT were decreased by half
with use of dual-chamber detection compared with ventricular-only detection with optimal programming of both. This
observation held true in a broad population of ICD recipients
and resulted from improved detection across all arrhythmia
subtypes, with the exception of sinus tachycardia, for which
detection was equivalent in both arms. Second, as expected,
the decrease in inappropriate detections led to a decrease in
inappropriate therapy, from 33% to 25% (including all
therapy types) at 6 months. Although the overall number of
therapies was decreased in the dual-chamber arm, the number
Figure 6. Therapy delivery for all SVT
episodes by study group. In the singlechamber group, 321 of the 531 SVT episodes (60.4%) were appropriately diagnosed as SVT, and therapy was withheld
(Rejected, no Rx). Thirty-five episodes
(6.6%) were misdiagnosed but received
no therapy (Detected, no Rx), predominantly because of termination after
detection but before therapy delivery;
154 (29.0%) received inappropriate ATP
only, and 21 (4.0%) received inappropriate shocks. In contrast, of the 750 SVT
episodes in the dual-chamber group, 518
(69.1%) were appropriately diagnosed as
SVT and therapy withheld, 46 (6.1%)
were misdiagnosed but received no therapy, 118 (15.7%) received inappropriate
ATP only, and 68 (9.1%) received inappropriate shocks. The rate of inappropriate therapy delivery was significantly
higher in the single-chamber group
(33.0%) than the dual-chamber group
(24.8%; P⫽0.03).
Friedman et al
of shocks delivered was slightly higher in the dual-chamber
arm, although this difference was not significant.
This discrepancy between therapies and shocks reflects the
greater use of ATP in the ventricular-only arm. In our study,
the type of therapy programmed was left up to the investigator to ensure compliance with the strict detection programming requirements. Treating physicians programmed more
ATP for patients in the single-chamber arm to minimize the
risk of inappropriate shock. Had identical therapies been
programmed, we predict that dual-chamber detection would
have resulted in fewer shocks. This result highlights the
importance of detection and therapy programming in minimizing ICD morbidity. The role of ATP in reducing shocks
has been described previously.18
Comparison With Other Studies
Downloaded from http://circ.ahajournals.org/ by guest on June 11, 2017
Most previous studies comparing dual-chamber with singlechamber ICDs have not found a superiority of dual-chamber
detection.3–5 The lack of superiority in most previous studies
relates to difficulties with atrial sensing, the predominant
cause of detection errors in dual-chamber ICDs.19 Accurately
determining that the ventricular rate is greater than the atrial
rate eliminates approximately half of all tachycardia episodes
from further analysis, which prevents subsequent detection
error14; conversely, underdetecting the atrial rate leads to
misclassification as “V⬎A” and to misdiagnosis of SVT as
VT with consequent inappropriate therapy. Rate branch
sorting also permits differential application of the detection
enhancements on the basis of relative atrial and ventricular
rates, improving the performance of the enhancements and
the ICDs (Figure 4).14
In the present study, careful attention to atrial sensing
resulted in ⬍3% far-field R wave oversensing in the dualchamber arm; atrial sensing errors accounted for only 5% of
the misclassifications in the dual-chamber arm. In contrast,
atrial sensing errors accounted for 41% to 75% of treated
misclassifications in previous comparative studies.3,4,8 In the
report by Kuhlkamp et al,3 an early-generation dual-chamber
defibrillator was used with a relatively long, fixed, atrial
blanking period (86 ms) after each sensed or paced ventricular event. This configuration resulted in the atrial undersensing that led to most misclassifications. In the trials by Theuns
et al20 and Deisenhofer et al,4 various defibrillators with
differing atrial sensing characteristics were used, but atrial
sensing errors were common.
The 1⫹1 trial21 randomly assigned patients to a crossover
comparison of single-chamber and dual-chamber detection.
Only patients with known slow VT were enrolled. A “more
than moderate superiority” of dual-chamber detection was
found for the study end point, which was a combination of
the number of inappropriate therapies and VTs slower than
the VT detection interval. Our study confirms the results
of the 1⫹1 trial and extends them beyond patients with
known slow VTs to the general ICD population. Indeed, two
thirds of the patients in the present study received their device
for primary sudden death prevention.
Like the 1⫹1 trial and in contrast to other previous studies,
all patients in our study received dual-chamber ICDs that
recorded atrial and ventricular intracardiac electrograms. This
Dual-Chamber vs Single-Chamber ICD Detection
2877
ensured the accuracy of human-reviewer rhythm diagnosis in
both study arms. In single-chamber devices, the absence of
stored atrial electrograms limits interpretation of some arrhythmia episodes. Additionally, because the presence of
atrial and ventricular electrograms in both groups gives their
episode printouts an identical appearance, blinding for core
group review was permitted, thus eliminating bias.
It must be emphasized that the high frequency of end point
occurrence in the present study (40% in the ventricular-only
arm, 31% in the dual-chamber arm) was because we used
inappropriate detection rather than inappropriate therapy as
our study end point. This end point was chosen because the
selection of dual- or single-chamber detection directly affects
detection, not ventricular therapy. The rate of inappropriate
detection in our single-chamber arm was similar to that
reported as a secondary end point by Theuns et al5 (42%).
Moreover, the rate of inappropriate shock in our population
(4% to 9%) compares favorably with the 10% to 41%
described in the other comparative studies.3–5 Because detection enhancements are not applied during redetection in
current ICDs and because many SVTs are not terminated by
ventricular therapies, a single misclassified episode may
result in many therapies.
Clinical Implications
Although the implantable defibrillator is the most effective
therapy available for sudden death prevention, substantial
morbidity results from inappropriate therapy. We found that
with optimally programmed current-generation defibrillators,
use of dual-chamber devices decreases the odds of inappropriate detection by approximately half compared with optimal
single-chamber detection. Although this decrease in detection
resulted in fewer inappropriate therapies, number of shocks
was not decreased because of greater use of ATP in the
single-chamber arm. The benefit of the atrial lead may be
modest in that aggressive use of ATP may prove as effective
as dual-chamber detection for preventing shocks; however,
we believe appropriate diagnosis is superior to inappropriate
ATP for shock prevention. Inappropriate ATP carries a small
risk of proarrhythmia.
Other studies have found that a history of SVT predicts
subsequent SVT.8 In our study, we found no such association.
The proportion of subjects having at least 1 episode of SVT
during the 6 months after implantation was high (32%),
similar to that reported in other studies3–5 and almost identical
for those with and without a history of SVT. This result
suggests that dual-chamber devices should be considered in
all ICD patients. Our findings also support the previous
observation that therapy programming with use of ATP
further minimizes patient shocks.18,22 Shock delivery is the
final step in a cascade of events beginning with detection.
Attention to each step of the cascade will minimize ICD
morbidity.
It is important to note that detection enhancements are
nominally off in most currently available ICDs. In clinical
use, the great majority of ICD parameters are left at nominal
settings; therefore, if manufacturers were to set the detection
enhancements nominally on, use of the enhancements might
increase.
2878
Circulation
June 27, 2006
Finally, devices were programmed to minimize ventricular
pacing. Frequent ventricular pacing might have prevented
template updates, affecting morphology function. Additionally, ventricular pacing introduces blanking and refractory
periods that may have affected arrhythmia detection and
classification, and it promotes heart failure and atrial fibrillation,23 which may also have affected the results.
Limitations
Downloaded from http://circ.ahajournals.org/ by guest on June 11, 2017
ICD detection involves numerous complex functions, and
unique approaches have been developed by each manufacturer. The results with dual-chamber detection found in the
present study may not extend to other manufacturers’ devices.
A blinded core group prevented physician bias, and GEE
analysis mitigated biased variability estimates due to excessive episodes from a single subject. The study was designed
to be powered for the primary end point. The Kaplan-Meier
approach does not use as much information as the GEE
approach (only the first inappropriately detected episode per
subject is used) and hence has somewhat less power. However, the point estimates using this approach were in the same
direction, and the magnitude of the effect appeared (qualitatively) similar. We chose to explore and present this analysis
because it has a useful graphical component.
Conclusion
SVTs were common, occurring in 34% of ICD recipients
within 6 months of implantation. SVT was inappropriately
detected in ⬇66% of subjects in whom it occurred. More than
three fourths of inappropriately detected episodes resulted in
inappropriate therapy. Current-generation dual-chamber detection enhancements significantly lowered the rate of inappropriate detection from 40% to 31% (P⬍0.05).
Dual-chamber ICDs, programmed to optimize detection
enhancements and to minimize ventricular pacing, significantly decrease inappropriate detection.
Acknowledgments
John McGuire and Tamara Shipman of St Jude Medical reviewed the
manuscript for technical accuracy. Editing, proofreading, and reference verification were provided by the Section of Scientific Publications, Mayo Clinic.
Source of Funding
This study was sponsored by an unrestricted grant from St Jude
Medical (St Paul, Minn).
Disclosures
Dr Daoud reports having served as a consultant for St Jude and
having served on the Speakers Bureau for St Jude. Dr Wilkoff
reports having served as a consultant for St Jude Medical. The
remaining authors report no conflicts.
References
1. Moss AJ, Zareba W, Hall WJ, Klein H, Wilber DJ, Cannom DS, Daubert
JP, Higgins SL, Brown MW, Andrews ML, for the Multicenter Automatic
Defibrillator Implantation Trial II Investigators. Prophylactic implantation of a defibrillator in patients with myocardial infarction and reduced
ejection fraction. N Engl J Med. 2002;346:877– 883.
2. Bardy GH, Lee KL, Mark DB, Poole JE, Packer DL, Boineau R,
Domanski M, Troutman C, Anderson J, Johnson G, McNulty SE, ClappChanning N, Davidson-Ray LD, Fraulo ES, Fishbein DP, Luceri RM, Ip
JH, for the Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT)
Investigators. Amiodarone or an implantable cardioverter-defibrillator for
congestive heart failure [published correction appears in N Engl J Med.
2005;352:2146]. N Engl J Med. 2005;352:225–237.
3. Kuhlkamp V, Dornberger V, Mewis C, Suchalla R, Bosch RJ, Seipel L.
Clinical experience with the new detection algorithms for atrial fibrillation of a defibrillator with dual chamber sensing and pacing. J Cardiovasc Electrophysiol. 1999;10:905–915.
4. Deisenhofer I, Kolb C, Ndrepepa G, Schreieck J, Karch M, Schmeider S,
Zrenner B, Schmitt C. Do current dual chamber cardioverter defibrillators
have advantages over conventional single chamber cardioverter defibrillators in reducing inappropriate therapies? A randomized, prospective
study. J Cardiovasc Electrophysiol. 2001;12:134 –142.
5. Theuns DA, Klootwijk AP, Goedhart DM, Jordaens LJ. Prevention of
inappropriate therapy in implantable cardioverter-defibrillators: results of
a prospective, randomized study of tachyarrhythmia detection algorithms.
J Am Coll Cardiol. 2004;44:2362–2367.
6. Grimm W, Flores BF, Marchlinski FE. Electrocardiographically documented unnecessary, spontaneous shocks in 241 patients with implantable
cardioverter defibrillators. Pacing Clin Electrophysiol. 1992;15:
1667–1673.
7. Schmitt C, Montero M, Melichercik J. Significance of supraventricular
tachyarrhythmias in patients with implanted pacing cardioverter defibrillators. Pacing Clin Electrophysiol. 1994;17:295–302.
8. Theuns DA, Klootwijk AP, Simoons ML, Jordaens LJ. Clinical variables
predicting inappropriate use of implantable cardioverter-defibrillator in
patients with coronary heart disease or nonischemic dilated cardiomyopathy. Am J Cardiol. 2005;95:271–274.
9. Schron EB, Exner DV, Yao Q, Jenkins LS, Steinberg JS, Cook JR,
Kutalek SP, Friedman PL, Bubien RS, Page RL, Powell J. Quality of life
in the antiarrhythmics versus implantable defibrillators trial: impact of
therapy and influence of adverse symptoms and defibrillator shocks.
Circulation. 2002;105:589 –594.
10. Wilkoff BL, Cook JR, Epstein AE, Greene HL, Hallstrom AP, Hsia H,
Kutalek SP, Sharma A, for the Dual Chamber and VVI Implantable
Defibrillator Trial Investigators. Dual-chamber pacing or ventricular
backup pacing in patients with an implantable defibrillator: the Dual
Chamber and VVI Implantable Defibrillator (DAVID) Trial. JAMA.
2002;288:3115–3123.
11. Centers for Medicare & Medicaid Services. Coverage determinations:
implantable automatic defibrillators. In: Medicare National Coverage
Determinations Manual. Baltimore, Md. CMS-Pub 100-3; Pt 1, section
20.4 [cited 2005 Jul 20]. Available at: http://www.cms.hhs.gov/mcd/
viewncd.asp?ncd_id⫽20.4&ncd_version⫽3&basket⫽ncd%3A20%2E4%
3A3%3AImplantable⫹Automatic⫹Defibrillators. Accessed May 24,
2006.
12. Swerdlow CD, Brown ML, Lurie K, Zhang J, Wood NM, Olson WH,
Gillberg JM. Discrimination of ventricular tachycardia from supraventricular tachycardia by a downloaded wavelet-transform morphology
algorithm: a paradigm for development of implantable cardioverter defibrillator detection algorithms. J Cardiovasc Electrophysiol. 2002;13:
432– 441.
13. Gold MR, Hsu W, Marcovecchio AF, Olsovsky MR, Lang DJ, Shorofsky
SR. A new defibrillator discrimination algorithm utilizing electrogram
morphology analysis. Pacing Clin Electrophysiol. 1999;22:179 –182.
14. Glikson M, Swerdlow CD, Gurevitz OT, Daoud E, Shivkumar K,
Wilkoff B, Shipman T, Friedman PA. Optimal combination of discriminators for differentiating ventricular from supraventricular
tachycardia by dual-chamber defibrillators. J Cardiovasc Electrophysiol. 2005;16:732–739.
15. Swerdlow CD, Friedman PA. Advanced ICD troubleshooting: part II.
Pacing Clin Electrophysiol. 2006;29:70 –96.
16. Wilkoff BL, Kuhlkamp V, Volosin K, Ellenbogen K, Waldecker B, Kacet
S, Gillberg JM, DeSouza CM. Critical analysis of dual-chamber
implantable cardioverter-defibrillator arrhythmia detection: results and
technical considerations. Circulation. 2001;103:381–386.
17. Liang K-L, Zewger SL. Longitudinal data analysis using generalized
linear models. Biometrika. 1986;73:13–22.
18. Wathen MS, DeGroot PJ, Sweeney MO, Stark AJ, Otterness MF,
Adkisson WO, Canby RC, Khalighi K, Machado C, Rubenstein DS,
Volosin KJ, for the PainFREE Rx II Investigators. Pacing Fast Ventric-
Friedman et al
ular Tachycardia Reduces Shock Therapies (PainFREE Rx II) trial
results. Circulation. 2004;110:2591–2596.
19. Israel CW, Gronefeld G, Iscolo N, Stoppler C, Hohnloser SH. Discrimination between ventricular and supraventricular tachycardia by dual
chamber cardioverter defibrillators: importance of the atrial sensing
function. Pacing Clin Electrophysiol. 2001;24:183–190.
20. Theuns D, Klootwijk AP, Kimman GP, Szili-Torok T, Roelandt JR,
Jordaens L. Initial clinical experience with a new arrhythmia detection
algorithm in dual chamber implantable cardioverter defibrillators.
Europace. 2001;3:181–186.
21. Bansch D, Steffgen F, Gronefeld G, Wolpert C, Bocker D, Mletzko RU,
Schols W, Seidl K, Piel M, Ouyang F, Hohnloser SH, Kuck KH. The 1⫹1
Dual-Chamber vs Single-Chamber ICD Detection
2879
trial: a prospective trial of a dual- versus a single-chamber implantable
defibrillator in patients with slow ventricular tachycardias. Circulation.
2004;110:1022–1029.
22. Wilkoff B, Al E. EMPIRIC study. Paper presented at: Heart Rhythm 2005
Scientific Sessions; May 4 –7, 2005; New Orleans, La.
23. Sweeney MO, Hellkamp AS, Ellenbogen KA, Greenspon AJ, Freedman
RA, Lee KL, Lamas GA, for the MOde Selection Trial Investigators.
Adverse effect of ventricular pacing on heart failure and atrial fibrillation
among patients with normal baseline QRS duration in a clinical trial of
pacemaker therapy for sinus node dysfunction. Circulation. 2003;107:
2932–2937.
CLINICAL PERSPECTIVE
Downloaded from http://circ.ahajournals.org/ by guest on June 11, 2017
The implanted cardioverter/defibrillator (ICD) has become the gold standard for sudden cardiac death prevention in
high-risk patients. Delivery of inappropriate shocks caused by misclassification of rapid supraventricular tachycardia
(SVT) is a substantial limitation of ICDs, affecting up to 40% of patients. Dual-chamber ICDs, which use atrial and
ventricular electrograms for arrhythmia diagnosis, were anticipated to improve discrimination between SVT and ventricular
arrhythmias. Early studies did not support this hope, suggesting that single-ventricular-lead ICDs should be favored if atrial
pacing is not required. The present multicenter prospective study sought to determine whether dual-chamber arrhythmiasensing algorithms reduce ICD misclassification of SVT. Four hundred patients receiving dual-chamber ICDs were
randomly assigned to dual- or single-chamber detection enhancements and followed up for 6 months. All ICDs were
programmed to minimize ventricular pacing. SVTs occurred frequently, regardless of whether a history of atrial
tachyarrhythmias was present. The odds of misclassification of SVT were decreased by almost half with the use of
dual-chamber detection enhancements (odds ratio, 0.53; P⫽0.03). This improved detection translated to a reduction in
inappropriate ICD therapies. It did not, however, reduce inappropriate shocks, because antitachycardia pacing, rather than
shocks, was used more frequently in the single-lead group. These data show that use of a dual-chamber ICD that provides
atrial sensing is a reasonable consideration for most patients. They also demonstrate the importance of both appropriate
detection and programming of antitachycardia pacing for minimizing shocks.
Dual-Chamber Versus Single-Chamber Detection Enhancements for Implantable
Defibrillator Rhythm Diagnosis: The Detect Supraventricular Tachycardia Study
Paul A. Friedman, Robyn L. McClelland, William R. Bamlet, Helbert Acosta, David Kessler,
Thomas M. Munger, Neal G. Kavesh, Mark Wood, Emile Daoud, Ali Massumi, Claudio
Schuger, Stephen Shorofsky, Bruce Wilkoff and Michael Glikson
Downloaded from http://circ.ahajournals.org/ by guest on June 11, 2017
Circulation. 2006;113:2871-2879; originally published online June 12, 2006;
doi: 10.1161/CIRCULATIONAHA.105.594531
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2006 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circ.ahajournals.org/content/113/25/2871
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial
Office. Once the online version of the published article for which permission is being requested is located,
click Request Permissions in the middle column of the Web page under Services. Further information about
this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation is online at:
http://circ.ahajournals.org//subscriptions/