Quality Requirements and Standards
A Dark Secret about the Safety of Medical Devices
The public assumes that the FDA is the gatekeeper for safety, and that everything with an FDA clearance is therefore safe and effective. Even some laboratories and other healthcare professionals make this assumption. But a recent report from the Institute of Medicine and an editorial in the NY Times are revealing a cold truth about the 510(k) process...
A Dark Secret about the Safety of Medical Devices
James O. Westgard, PhD
August 2011
- How Safe Are Medical Devices?
- No proof of quality or quality control
- Predicate chain of evidence does not provide traceability
- What should FDA do?
- Links
The American public is now being informed about something that we in the laboratory have long known – that the FDA process for approval of new devices for clinical testing (known as the 510k submission) does not assure the safety and efficacy of those medical devices. Here’s what the NY Times editorial of August 5, 2011, had to say:
How Safe Are Medical Devices? For all but the most risky, there appears to be no way to know for certain.
“A prestigious medical group is concerned that most medical devices are cleared for sale without evidence that they are safe and effective. That should give everyone pause. A panel of experts assembled by the Institute of Medicine found the current regulatory approach so flawed that it called for it to be scrapped and replaced with a system better designed to protect the public’s health.
“The committee found no reason to believe that any devices on the market are unsafe or ineffective; it said their use in clinical practice provides at least some level of confidence. But it threw up its hands and said it had no way to determine which, if any, devices might be risky because there is virtually no data to rely on.
“That’s because only the highest-risk devices, such as implantable defibrillators and replacement heart values, undergo clinical trials in patients to show that they are safe and effective. Devices deemed only a ‘moderate risk’ are cleared if the manufacturer can show that they are ‘substantially equivalent’ to earlier devices that may no longer be in use.”
No proof of quality or quality control
We have long recommended that laboratories validate the performance of new devices to assure that the necessary clinical and/or analytical quality can be achieved, not just verify that the precision and bias are consistent with the manufacturer’s claims (verification of claims, as allowed under CLIA). Current claims mainly are evaluated by FDA under the principle of “truth in labeling,” meaning that as long as the manufacturer’s data supports the claim, the claim is reliable. However, a manufacturer can claim performance that is totally inadequate for the intended medical use but still satisfy the 510k clearance process. Furthermore, there are NO claims for quality or for quality control that have to be demonstrated in the data submitted for clearance of new analytic systems. Rather, manufacturers need only show the performance of the new system relative to an older, already-approved system (predicate device). Such performance could be better or worse and still be adequately documented as truth in labeling.
Predicate chain of evidence does not provide traceability
“Problems can arise when the devices are not all that similar or when none of the devices in a chain of supposed equivalents have ever been shown safe and effective…
“The Food and Drug Administration was taken aback when the institute’s panel concluded, unanimously, that the current system could not be fixed because ‘substantial equivalence’ is based on comparisons reaching back for more than 35 years to devices that were never systematically assessed for safety or effectiveness…”
There are two really serious issues. (1) There is no evidence of safety and effectiveness for the predicate devices used to establish “equivalence.” (2) The chain of evidence is not equivalent to the “traceability chain” that is necessary to assure the comparability of test results from method to method. Such a traceability chain requires the inclusion of reference methods and reference materials to demonstrate comparability and achieve harmonization of test results.
What should FDA do?
"The F.D.A. promptly rejected the notion of junking the system but said it would hold hearings on the report. Industry leaders, who had been making scurrilous attacks on the panel’s objectivity even before the report was issued, voiced opposition to such a radical change.”
Actually, FDA should seize this opportunity to put the issue of traceability on the table and substitute the traceability chain for the current predicate chain. Unless this is done, the comparability of test results from method to method will not be achievable. FDA knows that its current regulatory structure cannot provide comparability of results and has admitted as much in current scientific discussions about traceability. Here’s the opportunity to step forward and improve the regulatory process and support the harmonization of test results!