More on Biotronik's Exploding ('Venting'?) ICD

According to MedPageToday, it appears an earlier case report that was mysteriously withdrawn from the peer-reviewed journal Europace will soon be republished:

A controversial article about problems with an implanted cardiac device -- published by and then withdrawn from the journal Europace -- has been resubmitted and is under review, according to the journal's editor.

"I expect that a decision on publication will be made very shortly," John Camm, MD, of St. George's University of London, told MedPage Today in an e-mail.

Hopefully, the journal will explain why they failed to notify their readers about the withdrawl as well. To withdraw an article of such signficance to their readership without explanation should not be tolerated by the scientific community.

Several other issues:

1. Perhaps even more concerning this whole ordeal has been the FDA's management of the device report made to them in May. It seems public reporting of Biotronik's filing did not appear on the MAUDE database until after my blog post was published in October.
   
   If this is their policy to withhold reports in patients that are injured for this length of time irrespective of "cause," there are bigger concerns with the government's policies that should be immediately addressed.
2. I should also explain my rationale for my "defensive blogging" earlier, too.
   
   The Fair Use Act of US Copyright law has been a favorite place for malicious lawyers to attack bloggers who republish content in their blogs. Irrespective of whether or not one could defend their actions in the court of law on the consitutional basis of "free speech."
   
   But when challenged, as soon as a lawyer gets involved, thanks to the large costs involved, you've "lost" your case even before going to trial. I did not need that expense at the time, so I caved and withdrew the pictures I had published (note: they have since been republished on the blog Cardiobrief.org, courtesy of Google cache, but have included a pdf of my copy of the entire case report in this blog post).

-Wes

Biotronik Responds to Europace Exploding ICD Case Report Article Withdrawl

Yesterday evening, I met with Rex Richmond (Vice President of Marketing) and Dan Schlewitz (Executive Vice President U.S.A. Sales) from Biotronik, Inc. to hear their side of the controversy surrounding the withdrawal of a case report previously published in Europace pertaining to an "exploding" Biotronik implantable cardiac defibrillator (ICD). They claimed they have been transparent regarding the reporting of this incident to the FDA, filing their incident report with the FDA within seven days of the event (within thirty days is required). Mr. Richmond granted me permission to release their internal company memo (pdf) circulated to their employees regarding the case report.

I must supply several comments.

First, regarding why the case report was withdrawn, they state in their memo:

The author reported that after submitting the case report to Europace in June, further analysis was conducted but not included in the original report. As such, there are inaccuracies that need to be corrected. Specifically, the author stated that the term "explosion" was not accurate given that the device was distorted, but had not exploded as previously described. The author also observed that while this is the first such incident with a BIOTRONIK device, it is not in fact the first experience of a battery overheating in the industry.

Note that the updated MDA was submitted to the FDA 6 June 2010. The published article was submitted to Europace 29 June 2010 and accepted after revision 16 August 2010. Wouldn't these "inaccuracies" have been corrected with the revisions submitted back to Europace following their review well before the 27 Sep 2010 publishing date?

Second is the issue of FDA reporting mentioned in the memo:

Where do reports like this end up? Aren’t they public?

Yes, they do become public. The FDA posts MDR’s to their online Manufacturer and User Facility Device Experience (MAUDE) database at the following address:

http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfMAUDE/TextSearch.cfm

(ed: Emphasis mine) Although the FDA has our MDR, it has yet to be posted to the database.

I do not know why the FDA would not have published the MDA's received in May with revisions in June on their website by 31 August 2010 (August data appear to have the latest updates available online) if the original MDA was submitted in May as Biotronik suggests. Does it really take this long to review MDA's before they are published? It is one thing to verify an MDA from the public before publishing on the FDA MAUDE database, but a manufacturer's MDA regarding their own device should be published without delay.

-Wes

Reference: pdf of my personal copy of withdrawn Europace case report obtained online 5 Oct 2010 before its withdrawl with timeline of the article's submission/revision dates.

CORRECTION - 12:55pm CST 12 Oct 2010: Here's a link to the FDA report regarding this incident filed on the FDA's MAUDE database. I apologize for the inaccuracy.

Addendum 09:45 AM CST 12 Oct 2010 - Reactions from around the web:

Cardiobrief: "The Plot Thickens in the Case of the Exploding ICD"

Happy Hospitalist: "Exploding ICD Gets Biotronik Response"

A Rare Case of a Spontaneously Exploding Defibrillator

Dr. Wes removed this post on 7 October 2010.

Why?

He was practicing a little Defensive Blogging.

Remote ICD Interrogation and the Challenges of Telemedicine

This week in Circulation an article was published (before print) about the safety and efficacy of remote monitoring for implantable cardioverter-defibrillator (ICD) follow-up in the industry-sponsored TRUST (Lumos-T Safely RedUuceS RouTine Office Device Follow-up) Trial. This trial randomized 1339 patients in a 2:1 fashion between internet-based “home-monitoring" of patient’s ICD’s and conventional office-based follow-up of ICD’s. While the results of the trial were interesting in regards to ICD follow-up management using the two venues, the trial is even more interesting if one considers the implications of this trial to telemedicine in general.

By way of background, patients with ICD’s are burdened with frequent follow-up to their doctors with guidelines requiring visits very three to six months. With more patients receiving implanted devices that contain more capabilities beyond just pacing and defibrillation (like arrhythmia and fluid-status monitoring), the volume of patients risks outstripping the number of available appointments. Further, it is no secret that device companies would love to reduce their manpower burden for follow-up to reduce their costs since many small doctors offices turn to the device companies for the technical aspects of follow-up support.

Enter home monitoring via the Internet. Using this technology, patients can upload their device data from home, have it reviewed online, and since 90% of device checks require no “actionable” follow-up, forgo a trip to the device clinic if no problems are identified. In fact, the TRUST trial concluded that “Home monitoring is safe and allows more rapid detection of actionable events compared with conventional monitoring in patients with implantable electronic devices.” The authors clearly achieved their stated goals, demonstrating a significant reduction in hospital encounters from 3.8 visits per patient/yr in the control group to 2.1 visits per patient/yr in the intervention group, with no increase in death, stroke, or surgical interventions in patients treated with home monitoring.

What’s not to like, right?

But there are some potential challenges ahead.

For one, in the TRUST trial the decrease in scheduled hospital encounters was partially offset by a significant increase in unscheduled encounters from 0.5 per patient/yr to 0.78 per patient/yr, mostly due to visits following event notifications from the home monitoring device.

Secondly, in his editorial on the study, Dan Matlock, MD from the University of Colorado cautioned that the quality of the TRUST trial fell short in it’s “patient-centeredness” and suggests that “the trial was designed with the belief that multiple follow-up visits are a burden that primarily falls on the providers.” He offers the following cautionary analogy:

A good basis for interpreting the TRUST trial is to draw comparisons with the prostate-specific antigen (PSA) controversy. There was never a debate that the PSA test detects prostate cancer earlier, and the test was widely adopted by the medical community without any evidence of improved outcomes. Finally, after two decades of use, two large, randomized trials studying the effects of PSA screening were published. Only one of the two trials actually demonstrated a benefit, showing a reduction in death from prostate cancer of seven per 10 000 men screened over nine years. This benefit was coupled with an additional 340 diagnoses of prostate cancer, 177 prostatectomies, and 97 radiation therapies. The second trial did not demonstrate any benefit of PSA testing at all. Rather than being a story of saving lives, the story of PSA testing has been one of false positives, frequent biopsies, indolent cancer, patient anxiety, impotence, and incontinence.

Considering the lessons learned from PSA, the TRUST trial leaves some important questions unanswered. Namely, what are the benefits and harms of detecting “actionable events,” and what do patients think of home monitoring?”

It is easy for those of use who are computer-savvy to assume that patients will all leap to home monitoring or telemedicine as their preferred means to address many of their medical concerns. But as the sensitivity for “actionable” items increases thanks to the capabilities of this technology, there is the potential that treatments that ensue adds other unforeseen risks for our patients. For instance, take the example of asymptomatic atrial fibrillation detected by a heart failure patient’s ICD who is already on warfarin: do they really need an immediate cardioversion (and the costs, risks of sedation, and anxiety surrounding the finding) or might the patient have be just as well served by less aggressive conventional follow-up?

Certainly, there appears to be great potential for improved access for care with remote monitoring and telemedicine in our increasingly cost-constrained health care system. Still, doctors and telemedicine developers must assure our recommended treatment strategies that result from improved detection do not conflict with our patients' perspectives, needs and values for their care.

-Wes

References:

Varma N, Epstein AE, Irimpen A, Schweikert R, Love C, for the TRUST Investigators. "Efficacy and Safety of Automatic Remote Monitoring for Implantable Cardioverter-Defibrillator Follow-Up: the Lumos-T Safely Reduces Routine Office Device Follow-Up (TRUST) Trial." Circulation. 2010;122:325–332.

Matlock DD. "Big Brother Is Watching You - What Do Patients Think About ICD Home Monitoring? Circulation 2010;122:319-321.

Defibrillator Lead Report Misleads

It didn't take long for the media to chew up this week's report on the high failure rate of old defibrillator leads published in Circulation from a single center registry from Germany today. What's better than to scare the heck out of patients? I mean, fear sells, doesn't it?

For those of you who haven't seen it yet, doctors from Germany decided to look back at their last 990 consecutive defibrillator lead implants between 1992 and 2005. They found 148 (15%) of defibrillator leads failed during follow-up and required replacement. They found that annual failure rates increased over time, peaking at 20% in 10-year-old leads.

But before you go out an insist that your defibrillator be removed, let's look at a few facts from the very same article:

• 15% of the leads had problems, 85% of them did just fine.


• No patent died as a result of ICD lead failure


• Older lead materials that have since been show to have high failure rates are no longer used, but included in this report


• The authors performed 95% of their implants using a subclavian approach, an approach they freely admit is prone to an unusually high level of lead "crush" as the lead passes between the first rib and clavicle. Like a coat-hanger repetitively bent, leads implanted by this approach are subject to a higher failure rate.

But the authors do make several valuable points: (1) defibrillators and their leads need constant monitoring, even after they are implanted and (2) in patients implanted for "primary prevention" of a cardiac arrhythmia (that is, they've never had an arrhythmia before the device is implanted), consideration of device failure over the lifetime of the patient should be considered when recommending a device for implantation.

It is important to note that the Heart Rhythm Society has been aware that these devices can be subject to malfunction on occasion. As such they have recommended that implantable cardiac defibrillators (ICDs) be evaluated every three months by a qualified physician. Further, most of the ICD manufacturers are developing wireless telemetry systems that can check these devices even as often as daily to assure their reliability. Furthermore, companies have migrated away from the form of polyurethane that can degrade when exposed to metal oxides (as occurred in leads manufactured before 1997). The fact that these older leads were included in the analysis serves to make the data more sensational and press-ready.

I would suggest that people evaluate the performance data maintained by each of the companies and published on-line. Boston Scientific's (formerly Guidant) performance data on their devices can be found here, Medtronic's here, St. Jude's here and Biotronik's can be found here. These data suggest current failure rates for most ICD's is approximately 10% after 10 years, fully one-half of that described by the authors in the above article. Certainly, this number is not perfect, but helps keep a reasonable perspective on defibrillator lead reliability, and reflects the world-wide experience of patient's with a particular lead or device system, and may be a more accurate reflection of lead trends across multiple centers and operators.

-Wes