Thursday, December 11, 2008

What Price Safety?

In February, 1996, the FDA issued this directive:
"In keeping with its commitment to speed up patient access to new medical devices, FDA is exempting 122 categories of low-risk devices from premarket review, adding to the 450 categories already exempted from such review. Since the exempted devices will no longer have to wait for premarket review, they can reach patients sooner and FDA can shift resources to more critical needs."
With the increasing complexity of implantable medical devices and a growing desire to innovate ways to manage patient's chronic medical conditions remotely, there is a need to downsize connections of leads and sensors to the computer electronics of these sophistocated devices. Until now, a standard two-electrode connection system called an "IS-1" standard has been used to connect leads and sensors to the housing of implantable cardiac devices.

For at least the past three years, the large medical device manufacturers have been developing a new 4-electrode connection system called "IS-4" that promises to supplant the IS-1 connection system. Because the medical device industry has argued that the changes to the IS-1 standard used to implement the IS-4 standard were minor, they applied for pre-market approval through the FDA's 510K exemption that does not require human clinical testing prior to market release of devices, but rather relies on historical information from existing technology supplemented by animal and specialized bench testing. While this approach has worked well for some new defibrillator leads, others have not fared so well.

Earlier, the Medtronic Sprint Fidelis defibrillator lead family, a thinner cousin of prior defibrillator leads, was released under the same 510K exemption ruling of the FDA, but subsequently recalled in October, 2007 after post-market surveillance demonstrated an increased failure rate of this lead family compared to its predecessors. Some 235,000 patients were affected.

Now, just before the release of new defibrillator leads containing the new IS-4 standard connections, Drs. Robert G. Hauser and Adrian K. Almquist of the Minneapolis Heart Institute are sounding their concerns over upcoming release, arguing that human testing should be performed before the lead is released to the public at large. They argue, in part, four concerns:
  • "In our opinion, the decision by the FDA and industry to forgo premarket clinical testing of the connector is not in the best interest of patients. For starters, the four-pole connector is a complex electromechanical design that requires major changes to the ICD lead and pulse generator; it should therefore be considered an investigational technology."

  • "Second, no published study has shown that the results of bench testing or testing in animals accurately predict the clinical reliability of ICD systems. If they did, recalls would be uncommon."

  • "Third, there is no medically necessary reason for rapidly deploying the four-pole connector. Current connector systems have been in use for more than a decade and are performing reliably."

  • "Fourth, FDA approval would allow the four-pole connection system to be used in patients without the level of informed consent that is usually required for investigational devices. After the recent Supreme Court decision in Riegel v. Medtronic, FDA approval would also leave patients without the ability to sue the manufacturer for compensation for injuries caused by a four-pole connector."
Hauser's and Almquist's concerns are being voiced at a particularly sensitive time as batteries of the devices connected to the recalled Sprint Fidelis leads are expiring. Doctors are sensitive to the particular challenges that lead revisions impart in the management of their patients since lead revisions or replacements during battery changes greatly complicates an otherise simple battery change procedure.

But the FDA, in their accompanying editorial, countered:
"When preclinical studies and existing clinical data support approval and expected long-term failure rates are low, postapproval studies may provide an appropriate means for verifying long-term device performance. Ensuring device safety does not necessarily require that every device be clinically evaluated before marketing. Approval of some implantable leads with IS-4 connectors may be appropriate without new clinical data. The FDA decides what information is needed to support an approval on the basis of its review of an individual application. This is a scientifically sound approach to protecting public health.
So which side is right in this debate?

Well, in some respects, they both are. No one can argue that patient safety should take priority in this debate, but at what cost? When we review the failure rates of even a known faulty lead like the Sprint Fidelis, its failure is only slightly higher than prior leads. For instance, the failure rate of a "good" Medtronic Model 6947 Sprint Quatro lead has a 1.9%% failure rate at 3 years, versus the 5% 3-year failure rate of the "bad" Medtronic Sprint Fidelis lead. Detecting a signficant difference of 3 percentage-point failure rate would take two fairly large patient comparison groups randomized and followed for many years to detect such a small difference in failure rate with sufficient statitical power. This would add signficant cost to devices, certainly, and would delay market release of newer innovations. The chilling effect on device development would cause every company to think long and hard before bringing any new technology to market.

But the public has grown wary of lack of regulation in many sectors of our economy, especially that portion that has been ridden roughshod by the financial sector. They are aware that the FDA needs the device manufacturers for their review fees, and hence have become skeptical of the FDA's ability to remain unbiased in their review process. They are no longer naïve about the physical and psychological toll such failures have upon patients, thanks, in part, to the unrelenting press coverage that these recalls generate ("Bad news sells best, 'cause good news is no news.") Regretably, we rarely hear about the lives saved by these devices every day.

The technology boom in medicine is under tremendous pressure to contract as health care costs continue to skyrocket and Congress looks for new ways to cut costs. The days of unfettered capitalism and quick profits in medicine are quickly coming to an end. Hitching the bandwagon to "patient safety" will be a powerful driving force for policy advocates to stifle innovation in the health care sector as we cling to the security of the known during this economic contraction.

What are the implications of the loss of post-market surveillance as a means of determining patient safety? What are the implications for costs of devices going forward? How much are we willing to spend to assure relative safety perfection? How safe is safe enough? Have not the existing standards for device approval resulted in some very beneficial and reliable systems? Must we assume there is too little value or economy to post-market surveillance to permit it's exclusive use for introducing new technology?

There's no question that Drs. Hauser and Almquist make important points. The bigger question might ultimately be: how much do we want to spend?



Hauser RG, Almquist AK. "Learning from our mistakes? Testing new ICD technology." N Engl J Med 2008;359:2517-2519.

Shein MJ, Schultz DJ. "Testing New ICD Technology," New Engl J Med 2008; 359: 2610.


Anonymous said...

Dr. Wes
I enjoy reading your blog and always appreciate your perspective. The following need not be posted because I am honestly not trying to detract from the point of your story or that of Dr. Hauser. As someone who has been involved in the past in the development of standards, we are sticklers for detail and I just had to point out few items:
+ the FDA approval process is case is called a PMA supplement or PMA-S, not a 510(k). They are two very different reviews. A 510(k) is a statement by the manufacturer that the product is similar to a device that was in use prior to 1976 when the device laws went into effect. Often little test data is provided to the FDA for a 510(k) approval. A PMA-S is submitted for a modification to a product that was previously approved by the FDA via the PMA process. Depending on the change being made it can require mechanical testing, animal testing and/or human testing or even more. A PMA-S is required if you change the outside vendor for any material or part used in the existing product. Typically this requires very little testing, only an assurance that the original specifications for the part are being met by the new vendor. A PMA-S can also be used when adding a single new therapy to an existing device. In this case, the FDA may require human clinincal data but not a full scale IDE clinical trial. An evaluation plan that assesses the need for mehanical, clinical, biological, etc. testing is often approved, modified or totally rejected by the FDA during an FDA submission strategy review done before any testing is even started.
+ IS-4 is a new connector and does not replace IS-1. IS-1 is a standard for the low voltage, pace/sense terminal and will still be used for pacemaker leads and the low voltage atrial & left heart lead connectors on ICDs/CRT-D devices. DF-1 is the standard for a single terminal, high voltage (HV), lead connector and will likely still be used for a spare HV port in case there is a need for a Sub-q lead. IS-4 is a new, additional, connector standard that has been being designed by representative from all device manufactures willing to participate for at least the last 6 years. It is meant to combine the 2 HV & 1 IS-1 terminal of a typical single pass defib lead into one inline connector.
+ FDA approval. The IS-1 connector standard was the first recognized standard in the device industry and took a very long time to be designed. FDA approval by each manufacturer was done in the same manner as is being proposed for the IS-4 connector. The DF-1 took considerably less time to develop (a much less challenging task) but again was approved in the same basic way.
As inferred above, I am from a device company and was part of the original group of people who struggled for years to standardize the IS-1 connector. I have not been part of the IS-4 effort but know that this has been an even bigger effort given the safety & performance requirements of the design. I just wanted to set the record straight that no one has rushed to get this on the market and that there has been significant FDA, HRS, and other agency involvement to the process, as was the case for the two previous standards.

DrWes said...

Anony 4:38PM-
Thanks so much for your clarification and your contributions to this blog. Your insights are appreciated and add to our insights of industry's efforts.

So much for my knowledge of the specifics of the regulatory process within the FDA, eh?

Guess I better get back to the patients...