Monday, February 05, 2007

The Jet Stream

When I was a young boy, I would lay in our pasture and stare up at the clouds, fascinated that they took different shapes as they moved across the sky and could grow to become a serious storm, or wither to just obscure the sun. I later learned that the growth of clouds was caused in part by ambient heating of the earth, moisture flows from the North or South, or surface wind currents. Still later I learned that above it all was the jet stream – the high-altitude fast moving winds created by warming of the earth from the sun coupled with the Earth’s rotation – that really created the change in weather patterns. Predicting the weather involves knowing these jet stream patterns and interpreting them – getting these predictions wrong could put lives at stake.

The practice of medicine is like studying ever-transforming clouds – you never really know what shape a patient’s illness will bring, nor their response to their illness. But by using deductive reasoning or listening to trusted colleagues or opinion experts, we usually can hazard a guess at where a case will go and react accordingly.

But sometimes we have to look beyond the clouds, beyond what the experts and textbooks tell us. Sometimes we have to collect our own observations, gather data, and delve deeper to get the real gist of a case to save a life.

Beginning in February 2005, a series of defibrillator recalls rocked our cardiac electrophysiology community. First, was an extensive defibrillator battery recall from Medtronic. Soon after in May 2005, Guidant Corporation issued a recall of its Ventak Prism II defibrillator line. What was concerning about this recall was that it occurred, we learned, only after two physicians notified Guidant that the New York Times was to publish an article about a young 21-year-old patient that died when their defibrillator failed to perform properly nearly three years after the company knew of the root problem. The episode highlighted an important issue:
New York Times, 5/24/05: “Doctors and patients are not always told when a medical device maker has data indicating that its product has a flaw that, while rare, poses potential dangers. Also, companies are not required to report immediately all safety modifications to the Food and Drug Administration.”
The cases regarding this recall and others are still winding through the courts, and will be for some time to come. St. Jude, too, had their own recall that year as well. In all, over 100,000 patients were affected by a recall or advisory condition about their device that year. Companies realized that there could be significant risks with keeping the public uninformed about known flaws in the design of their devices. They hurried to inform physicians and patients about their concerns, leaving doctors to notify their patients and try to determine the best course of action with very little real information. These concerns lead the Heart Rhythm Society, in conjunction with physician, patient, government and industry input, to issue new device performance policies and guidelines regarding notification of physicians and patients. A new defibrillator registry was also implemented to track currently implanted devices to determine the incidence data of defects with defibrillators.

But most important with the former fiasco, implanting physicians like me were sensitized, ever-cautious to take what industry and the FDA tells us with a grain of salt.

So on January 29, 2007, when I learned that Medtronic was being investigated by the FDA for concerns raised by a former engineer, Christopher Fuller, regarding how their new wireless Concerto defibrillator was tested, I listened. Carefully. According to the Wall Street Journal, he resigned over concerns the the device "exhibited instabilities in testing" and that Medtronic "has chosen to ignore the problem." Press releases like this one and this one were released on 30 January with FDA spokesperson Heidi Rebello reassuring us that “We are investigating.” Funny, because Medtronic didn’t seem to know anything about an investigation: 30 Jan 07 3:22PM: The Minneapolis-based device maker put 11,000 Concerto devices on the U.S. market after the FDA approved it in May of 2006, said Medtronic spokesman Robert Clark. But Clark said the FDA has not discussed the investigation with Medtronic.
"Medtronic has not been contacted by the FDA in regards to an investigation in this matter," said Clark.
I called David Steinhaus, MD, Director of Cardiac Rhythm Management at Medtronic in an attempt to better understand the situation and see if they felt there issues I should know about with their device. He returned my call the next day and reassured me that there were no issues with any of the 11,000 devices implanted to date. He referred me to this news release from Ms. Rebello, issued later the same day as the earlier press releases, that “all is well.” Everything’s been checked. We’ve looked into it, really. Medtronic tells us that over 11,000 devices have been implanted and no patient safety issues have occurred. Really. It’s fine. Nothing to worry about. They’ve checked the manufacturing plant – its fine. They’ve reviewed their FDA submission and their data – it all looks good:
"The agency has completed a thorough investigation of the charges. To date, FDA has found nothing that would have influenced its decision to approve this device." Rebello added the investigation included a scientific review of the employee's allegations, an inspection of the company's manufacturing facility and a review of various available data. "All of this information was analyzed with respect to whether or not the allegations identified a safety concern for patients that substantively altered the risk/benefit profile for this device," she added.
We can only hope for such governmental efficiency in all of our agencies: the ability to achieve complete clarity in a matter of hours. What, exactly, can you clarify in a few hours – the data or the spin? If everything was wrapped up, why didn’t they tell us this with the first press releases? And what does the term “substantively altered the risk/benefit profile for this device” mean, really? Did they find something that looked “risky?” If so, what was it?
Reuters Jan 30,2007 @ 6:49PM: Medtronic said it has had no reports of patient safety issues related to the wireless features of the device.
Then I checked the FDA database for Concerto-related injuries myself and found this. Since I cannot determine the nature of this device’s defect, I am left to wonder: could this have been caused by one of the ex-Medtronic engineer’s concerns? And regarding Medtronic’s knowledge of an FDA inspection: did they know about the FDA inspection of their manufacturing facility or didn’t they? Why not level with me, one of their biggest advocates?

Everything now is in question. The bloom is off the rose.

To be fair, I still believe that all of the device manufacturers in this business are reputable companies. Medtronic, Guidant (now Boston Scientific), and St. Jude have been in the business for years and have remarkably reliable devices that have saved many, many lives that otherwise would have been lost. They constantly strive to improve their products, tirelessly check devices for me in the field, and help develop policies to promote the well-being of patients. Many individuals from these companies have devoted their lives to improving the well being of their fellow man. The FDA, too, has a very arduous approval process in the interest of public safety. Every side effect of every device tested is scrutinized by Data Safety Monitoring Boards made up of many physician-reviewers.

But in regards to the handling of this piece from the Wall Street Journal – I feel it would have been far better to have stated the data when responding to physicians or the press. An example might have been something like this (this is my example only):
Review of our worldwide experience to date has identified only one (actual number) of 11,012 implanted devices (give the actual number here) that has demonstrated device malfunction with compromised therapy. This failure rate falls well below the anticipated incident rate of annual failures of 20.7 per 1000 devices implanted annually for such devices outlined by the Heart Rhythm Society Task Force on Device Performance Policies and Guidelines. We are investigating/have investigated the root cause of this malfunction. (Explain what was/was not found here). We do/do not feel this device malfunction is related to the concerns raised by our former employee.
But this is not what was said. Instead we were obfuscated with PR-speak. “We are investigating” became “we found nothing” in less than 24-hours. Well if I could find one injury on the FDA’s own database, how hard did the FDA look?

Which returns to my real question: how can a physician verify if the press releases from industry and the media are valid?

Has the FDA’s responded to the engineer’s concerns? Wouldn’t it be nice to see their response to him on the MAUDE database? Instead, to investigate the current situation, I called the Office of the Inspector General (OIG) over at the Department of Health and Human Services (to whom the FDA Director reports) and was politely told by HHS OIG spokesperson Donald White that they can’t and won’t release if or when there was an investigation. “I’m sorry, we cannot help you. We do not release information regarding investigations. This information is not covered under the Freedom of Information Act.” I was told.

The lines are blurred. By not giving physicians and the public the truth, the whole truth, and nothing but the truth in press releases or through company spokespersons, public trust is eroded. Worse still, could I be liable to my patient for not telling knowing the facts that exist on the publicly searchable FDA database because I listened to others rather than checked myself? My patients trust that I know the data.

But there is nothing verifiable in this PR world, it seems. It’s as though my clouds of concern are being manipulated by a bigger jet stream high above our work-a-day world. Doctors are often blown about by economic and political forces about which they know little. But what was clear about the handling of the current allegations, especially in light of our prior experience in 2005 with the earlier defibrillator device recalls, is that device manufacturers and the FDA have learned little about earning public trust.

When it comes to public safety, there is no room for PR spin.

So where, my friends, is the weather satellite above the jet stream?



Anonymous said...

...checked the FDA database....found "this" link does not work.

DrWes said...

Anony - Thanks, the link has been fixed.

The link can also be found here.


Anonymous said...

try this

DrWes said...

Anony 2:22PM - In the FDA MAUDE database entry that you highlighted, the event can be explained by a number of different programming errors by the operator (such as not having "mode switch" programmed "ON", for atrial rates of 150 for instance) and is not a problem with the device itself. While unfortunate for the patient, I really think it had nothing to do with a device malfunction in this case.

Anonymous said...


I wonder why they used a low frequency (non wireless) laptop to reprogram? I wonder whether Conexus wireless communication was not working.

Do you think the report would indicate a malfunctioning ICD and indicate that the report was a product problem report if it was merely a mode switch programming issue?

Some have suggested the patient wasn't injured. It's hard for me to imagine that the patient didn't suffer even if it was just from 10:30 P.M. to 2:30 P.M. What do you think?