'Routine' Use of Dual vs Single Chamber ICDs Again Questioned

August 23, 2011

August 22, 2011 (Washington, DC) — Most first-time implantable cardioverter-defibrillators (ICDs) that go in are dual-chamber devices, but dual-chamber devices also pose a significantly greater risk of serious procedural complications, including in-hospital death, than do single-chamber ICDs, according to a new analysis [1]. Moreover, it suggests, about 60% patients implanted with such dual-chamber ICDs don't have standard indications for cardiac pacing.

The two findings, based on 2006 and 2007 data from the ICD component of the National Cardiovascular Data Registry (NCDR-ICD), together raise risk-vs-benefit questions about the routine selection of dual-chamber devices, according to the authors, led by Dr Thomas A Dewland (University of California, San Francisco). The group's analysis was published online August 22, 2011 in the Journal of the American College of Cardiology.

In the US, a dual-chamber device, along with its extra atrial lead, will be chosen by default, on the belief that it "will better discriminate [supraventricular tachycardia] SVT from VT and thereby reduce the risk of inappropriate shocks," senior author Dr Paul D Varosy (University of Colorado, Denver) explained to heartwire . That belief, he said, is "certainly plausible but at this point unproven."

With that belief, he thinks, is the perception that the extra risk associated with dual-chamber devices "is so low that it's not even an issue." Varosy said his group's analysis--in which the risk of procedural complications, including in-hospital death, was at least 40% greater with dual-chamber than single-chamber ICDs (p<0.001)--challenges that assumption and points to a need for further research into the issue.

The guidelines don't provide much direction, he said, in part because the major ICD trials used primarily single-chamber devices. And observational data disagree. He did point to a recent analysis of a few thousand patients in Ontario, reported by heartwire on its publication last year, in which the risk of complications was directly related to the number of device leads. Compared with single-chamber ICDs, dual-chamber devices upped the complication risk by 82% (p=0.006) and biventricular devices more than doubled it (p<0.001).

The current analysis, write Dewland et al, "confirms and extends these findings in a large patient population receiving care across a wide spectrum of healthcare facilities in the US." Its indications for dual-chamber ICDs were limited to second- or third-degree heart block, bradycardic cardiac arrest, abnormal sinus-node function, and prior pacemaker implantation.

An accompanying editorial [2] agrees that "for implanting clinicians, the possible benefits of an additional lead must be counterbalanced by the possible risks to patients." But there are indications for cardiac pacing not considered by the NCDR-ICD registry analysis that could account for many of the dual-chamber-device implantations, write Drs Amin Al-Ahmad and James V Freeman (Stanford University School of Medicine, CA). Pacing can be appropriate in, for example, "patients with unexplained syncope when causes other than atrioventricular block or sinus-node dysfunction have been excluded or patients with alternating bundle branch block."

Moreover, they contend, there is increasing evidence supporting the use of dual-chamber devices to discriminate between SVT and VT in many patients. They point out that more dual-chamber patients than those getting single-chamber ICDs in the analysis had atrial fibrillation. So even if most dual-chamber-device recipients lacked standard pacing indications, "many of these patients may have had a need for enhanced arrhythmia detection, which can have significant clinical benefits."

Varosy agreed that for better arrhythmia discrimination, dual-chamber devices could be considered appropriate outside the indications for pacing they used in the analysis, including in patients with a known history of SVT, paroxysmal AF, atrial flutter, and long-QT syndrome. Limitations of the NCDR-ICD database didn't allow them to tease that out, he said. "But I think it's unlikely that it would account for the fact that 62% of devices going in are dual-chamber."

Their criteria for a pacing indication were met by only 40.4% of the >60 000 dual-chamber-device recipients. Their complication rate and in-hospital mortality was significantly increased compared with patients getting single-chamber devices.

Procedure-Related Complications of ICD Implantation by Type of Device in the NCDR-ICD Registry for 2006 and 2007

End point Dual-chamber ICDs (n=64 489) Single-chamber ICDs (n=39 560) p
Complications (%)* 3.17 2.11 <0.001
In-hospital mortality (%) 0.40 0.23 <0.001

*Including cardiac arrest, coronary perforation or dissection, valve injury, hematomas, lead dislodgements, peripheral nerve injury, phlebitis, pneumothorax, neurologic events, MI, tamponade, or infection

The odds ratio for any procedural complication was 1.40 (95% CI 1.28–1.52, p<0.001) and for in-hospital mortality was 1.45 (95% CI 1.20–1.74, p<0.001) for dual-chamber vs single-chamber ICDs after adjustment for demographics, comorbidities, diagnostic test data, and ICD indication, according to the group.

In their editorial, Al-Ahmad and Freeman say that despite the statistical adjustments, "the possibility of residual confounding biasing the results against dual-chamber devices is substantial."

Also, differences between the dual-chamber- and single-chamber-device recipients, such as the former's significantly wider average QRS interval (115 ms vs 109 ms, p<0.0001) "suggest that more of the dual-chamber ICD patients were aborted biventricular ICD recipients in whom left ventricular leads could not be placed." And that group would likely account for the excess in coronary venous dissections at implantation of dual-chamber devices (0.12% vs 0.01% for single-chamber devices, p<0.001), as those could have occurred only at attempted LV lead placement for a biventricular system.

"Their results, particularly with regard to in-hospital mortality, should be viewed in the context of important limitations and will need to be externally validated with future studies."

Varosy agreed that some of the patients recorded in the database as having received dual-chamber devices represented failed attempts at placing a biventricular ICD. But the rate of failed LV lead placement is "extraordinarily low," he said.

"I think it's unlikely to be the whole story. When we did the analysis excluding patients with coronary venous dissections, in both groups, the findings were essentially the same."

Varosy had no relevant disclosures; disclosures for the coauthors are listed in the paper. Al-Ahmad discloses honoraria from St Jude Medical, Boston Scientific, and Biotronik.