3 ms·

"... if a device was sold before 1976, an updated version can be released without new clinical trials, so long as the manufacturer deems it to be 'substantiall

"... if a device was sold before 1976, an updated version can be released without new clinical trials, so long as the manufacturer deems it to be 'substantially equivalent' to the existing device. Thirty years after the 1976 legislation was passed, the F.D.A. had approved more than a thousand high-risk devices, only sixteen per cent of which had gone through rigorous clinical testing before sale." Shades of Boeing here. Interesting.

ska 7y ago
It's not really shades of boeing, though I can see how it may look like that from the outside. The problem is that a pre-market assessment (PMA) is a huge and expensive beast, if you require every device/implant/etc. to go through that it will definitely slow things down a lot and make them more expensive... So the pre-market notification and de-novo paths were created to keep rigor in the system without adding years and millions to development costs of everything. For the most part it works fairly reasonably to balance these concerns. There is already a classification system (I, II, III) which captures this and can force your device through a PMA if it is deemed high enough risk. One could argue that class II is being used to much I suppose, but that is a very different issue than pointed at by boeing.
- ken 7y ago
  I must be missing something. That sounds almost exactly like the Boeing situation to me.
  - ska 7y ago
    It's a rigorous process, that is designed to remove or reduce the need for a different rigorous process. It's been working pretty well for decades. How is that parallel to the Boeing situation? We know there was an engineering failure followed by a process failure wrt 737 max. If we have similar evidence in devices (i.e. the process isn't working, or the developers are cutting corners, or both) that should certainly be addressed. But drawing that line with information such as this article seems a stretch, at best.
    - kansface 7y ago
      Forgive me, but I also had the same thought as OP and honestly can't tell the difference between what you are describing and Boeing. > It's a rigorous process, that is designed to remove or reduce the need for a different rigorous process. It's been working pretty well for decades. Just like the FAA and Boeing until it didn't? > We know there was an engineering failure followed by a process failure wrt 737 max. My reading of the situation was a process failure, a couple of engineering screw ups, and possibly user error.
      - ska 7y ago
        How are the both of you not included all regulatory oversight in the same bucket? Could the FDA process become weaker to the point of a problem, sure. Is there any evidence that it currently is? I haven't seen any. Is this something they (the FDA and similar) should keep an eye on, absolutely Honestly, I don't see how you are drawing the parallel. In what wad do you think the FDA is making the "same" mistake as the FAA did?
        at_a_remove 7y ago
        Ah, well, the FAA did not make Boeing go through recertification for the new "version" of the plane. Close enough. Same thing. Did it back then, should be alright. And so the safety process was not seen through on this new model. Similarly, the FDA grandfathered in a bunch of devices and because they are "substantially equivalent," the FDA did not ask for recertification and the safety process was not gone through for these devices.
        ska 7y ago
        What I am saying is I think that your last paragraph is a misunderstanding of what the 510(k) and de-novo pathways are, and that they are a safety process, which was and is followed. If you argued that the V&V necessary for a 510(k) filing is insufficient, I would understand where you are coming from, and it could be interesting. If you argued that a PMA is needed for all medical devices, I would understand the argument but probably disagree. Your 'similarly' just isn't that similar, unless I've vastly misunderstood the FAA/Boeing case. Again it's clear that the FDA could fall into a state where it's review and audit processes had the same sort of problems that the FAA exhibited with the 737max. But I don't see evidence at all that this has actually happened. What do you think the substantively equivalent bar for a 510(k) allows a company to do? Regulatory capture is a real risk for any regulatory oversight, but it is not synonymous with that regulatory oversight.

ska 7y ago

It's not really shades of boeing, though I can see how it may look like that from the outside. The problem is that a pre-market assessment (PMA) is a huge and expensive beast, if you require every device/implant/etc. to go through that it will definitely slow things down a lot and make them more expensive... So the pre-market notification and de-novo paths were created to keep rigor in the system without adding years and millions to development costs of everything. For the most part it works fairly reasonably to balance these concerns. There is already a classification system (I, II, III) which captures this and can force your device through a PMA if it is deemed high enough risk. One could argue that class II is being used to much I suppose, but that is a very different issue than pointed at by boeing.

ken 7y ago
I must be missing something. That sounds almost exactly like the Boeing situation to me.
- ska 7y ago
  It's a rigorous process, that is designed to remove or reduce the need for a different rigorous process. It's been working pretty well for decades. How is that parallel to the Boeing situation? We know there was an engineering failure followed by a process failure wrt 737 max. If we have similar evidence in devices (i.e. the process isn't working, or the developers are cutting corners, or both) that should certainly be addressed. But drawing that line with information such as this article seems a stretch, at best.
  - kansface 7y ago
    Forgive me, but I also had the same thought as OP and honestly can't tell the difference between what you are describing and Boeing. > It's a rigorous process, that is designed to remove or reduce the need for a different rigorous process. It's been working pretty well for decades. Just like the FAA and Boeing until it didn't? > We know there was an engineering failure followed by a process failure wrt 737 max. My reading of the situation was a process failure, a couple of engineering screw ups, and possibly user error.
    - ska 7y ago
      How are the both of you not included all regulatory oversight in the same bucket? Could the FDA process become weaker to the point of a problem, sure. Is there any evidence that it currently is? I haven't seen any. Is this something they (the FDA and similar) should keep an eye on, absolutely Honestly, I don't see how you are drawing the parallel. In what wad do you think the FDA is making the "same" mistake as the FAA did?
      - at_a_remove 7y ago
        Ah, well, the FAA did not make Boeing go through recertification for the new "version" of the plane. Close enough. Same thing. Did it back then, should be alright. And so the safety process was not seen through on this new model. Similarly, the FDA grandfathered in a bunch of devices and because they are "substantially equivalent," the FDA did not ask for recertification and the safety process was not gone through for these devices.
        ska 7y ago
        What I am saying is I think that your last paragraph is a misunderstanding of what the 510(k) and de-novo pathways are, and that they are a safety process, which was and is followed. If you argued that the V&V necessary for a 510(k) filing is insufficient, I would understand where you are coming from, and it could be interesting. If you argued that a PMA is needed for all medical devices, I would understand the argument but probably disagree. Your 'similarly' just isn't that similar, unless I've vastly misunderstood the FAA/Boeing case. Again it's clear that the FDA could fall into a state where it's review and audit processes had the same sort of problems that the FAA exhibited with the 737max. But I don't see evidence at all that this has actually happened. What do you think the substantively equivalent bar for a 510(k) allows a company to do? Regulatory capture is a real risk for any regulatory oversight, but it is not synonymous with that regulatory oversight.