Page 2 of 2 FirstFirst 12
Results 31 to 40 of 40

Thread: Atlas sucked

  1. #31
    Quote Originally Posted by Hazir View Post
    I am 100% convinced it is a payer issue, but that's about the last thing they want to admit.
    Hazir, please correct me if I'm wrong, but I think we're using the term 'payer issue' in a different context. My understanding from your post is that you think payers (which, in most of Europe, boil down to governments) are not interested in paying for the upfront cost of PrEP so they have intentionally slowed regulatory approval of Truvada for this indication. IMO that's vaguely conspiratorial and probably incorrect; most of the interactions I've had with regulatory agencies in this vein gives me very little reason to believe that there is that level of political interference in the process (though I know the US system much better than the EU system).

    When I say it's a payer problem, I mean to say that Truvada should in principle be available in the EU for PrEP as an off-label use - physicians can look at the data on its efficacy themselves and prescribe Truvada for PrEP even without official sanction of that indication (at least I assume this is so; US doctors routinely use/prescribe off-label therapies and are fully in their rights to do so). When I say it's a payer issue, I mean that I assume national payers (generally the government) won't subsidize the cost of PrEP without European approval of that specific indication, which is taking a bit longer than you'd like. But it's not a regulatory issue per se unless I was to believe that approval was being delayed for political reasons.

    But to be honest I haven't followed this story with any level of interest, so take what I say with a grain of salt.

    Quote Originally Posted by Loki View Post
    A real impediment to research: the Institutional Review Board.

    http://slatestarcodex.com/2017/08/29/my-irb-nightmare/
    Hmm. There's definitely a grain of truth here, but he's being intentionally obtuse (probably for comic effect). Most of the bureaucratic things he complains about have excellent reasons for existing, even for an ostensibly low risk study like the one he describes. And he bellyaches about this 22 page form he needs to fill out, when vast swaths of it could be filled out with a simple N/A - and, even better, his institution even HAD a shorter version for precisely this kind of purpose. It's not their fault that he didn't do the basic legwork to figure out what he needed to do - in my experience, it's reasonably well laid out on the various IRB websites/forms/etc. for any given institution.

    I have done a lot of work on animal studies, including getting protocols approved, and it's almost as laborious a process, involving lots of training, large numbers of forms, and a back-and-forth with the IACUC (which is essentially the animal version of an IRB). The considerations are different - obviously consent and protecting personal information are not issues with animals - but there's still a lot of bureaucracy even for ostensibly simple studies. I've found that in general, the process can take a bit of time but is relatively painless if two conditions are met: 1) you do your homework and 2) your institution's IACUC is doing their job, which is to help researchers craft protocols that meet statutory requirements. It sounds like in this case neither condition was met; maybe the institution had relatively little experience running human subjects research, or the clinician scientist was particularly obtuse (clinicians are notorious for being pretty awful at putting together a good research study and following through appropriately), but it clearly resulted in a suboptimal outcome. That is not the norm for a well-run institution.

    He claims that the system is rigged against the little guy who doesn't have an army of compliance people to jump through regulatory hoops. To an extent he's right (and big companies do tend to welcome regulations that increase the barrier to entry in their market), but IRBs and the associated legal requirements have not generally been driven by these concerns, but rather by pretty reasonable concerns about human subject welfare and the ethical challenges of experimenting on humans. And some of the concerns he pooh-poohs - such as privacy issues - are very much problems that would effect human subjects even for ostensibly low risk studies.

    There are a lot of great resources for IRBs for the 'little guy' - essentially contract IRB organizations that work with small companies and institutions that want to run human subjects research. They work very hard to streamline the process as much as possible in the framework of the law, and could easily whip up a protocol and get it rapidly approved for the little guy. Maybe it would have been better for the author's clearly small-scale IRB to outsource this function to someone who knows what they're doing.

    There's obviously scope to try to loosen some of the more onerous requirements for very low risk studies like those described by the author; no one really questions the value of an IRB for the kind of work Flixy or I might be doing. And I think it makes sense to make a 'fast track' IRB category specifically for this kind of work. But that doesn't mean that you shouldn't be trained in appropriate research practices, get informed consent, or worry about patient privacy, or go through some sort of review process.
    "When I meet God, I am going to ask him two questions: Why relativity? And why turbulence? I really believe he will have an answer for the first." - Werner Heisenberg (maybe)

  2. #32
    Wig, do you think it's reasonable to expect a grad student to hire an IRB organization to help them jump through all the hoops? I know smart and detail-oriented people who spent 2-3 years trying to get approval for field research, with most ultimately deciding to run some stats instead. Who does that benefit?
    Hope is the denial of reality

  3. #33
    Quote Originally Posted by Loki View Post
    Wig, do you think it's reasonable to expect a grad student to hire an IRB organization to help them jump through all the hoops? I know smart and detail-oriented people who spent 2-3 years trying to get approval for field research, with most ultimately deciding to run some stats instead. Who does that benefit?
    If an institution does a lot of that kind of research, they should have an in house IRB that doesn't suck, and then I'd definitely expect a grad student to be able to manage the process, with appropriate help from their advisor. If their institution doesn't do that kind of research much, they probably have a go-to external IRB they use and then I'd again expect the grad student to get with the program. I have lots of friends who did precisely this in graduate school as well as postdocs/fellowships - if your institution doesn't suck, there are adequate resources to make it no more painful than writing an animal protocol, which I have done (from scratch) probably half a dozen times in my academic career as a grad student and postdoc.

    For less clinical stuff that still counts as human subjects research, I have friends who have managed fieldwork among disadvantaged populations in sub-Saharan Africa, looking at controlled studies with different public health-y interventions to see how that affects HIV transmission. This is wildly sensitive work that touches on all sorts of political and ethical issues, but they were able (as graduate students) to successfully design their studies and get them approved through their IRB, working in concert with local authorities. It's hard, but with adequate institutional support, legwork, and a knowledgeable PI, it's eminently doable. Not to mention all of the clinical psych and econ grad students I know who got their own data rather than relying on existing datasets.

    I don't deny that it's hard and that it can and should be made easier for certain classes of research. And I also acknowledge that most of my personal knowledge comes from people working at institutions that will obviously have a very well developed IRB apparatus (generally affiliated with world-class hospitals and deep-pocketed universities associated with them). But that's precisely my point: either your institution should do a good job streamlining the process as much as legally and ethically possible, or they should turn to an external organization that is good at this sort of thing. Small companies routinely use external contract IRBs for their human subjects work and it works just fine.
    "When I meet God, I am going to ask him two questions: Why relativity? And why turbulence? I really believe he will have an answer for the first." - Werner Heisenberg (maybe)

  4. #34
    Quote Originally Posted by wiggin View Post
    If an institution does a lot of that kind of research, they should have an in house IRB that doesn't suck, and then I'd definitely expect a grad student to be able to manage the process, with appropriate help from their advisor. If their institution doesn't do that kind of research much, they probably have a go-to external IRB they use and then I'd again expect the grad student to get with the program. I have lots of friends who did precisely this in graduate school as well as postdocs/fellowships - if your institution doesn't suck, there are adequate resources to make it no more painful than writing an animal protocol, which I have done (from scratch) probably half a dozen times in my academic career as a grad student and postdoc.

    For less clinical stuff that still counts as human subjects research, I have friends who have managed fieldwork among disadvantaged populations in sub-Saharan Africa, looking at controlled studies with different public health-y interventions to see how that affects HIV transmission. This is wildly sensitive work that touches on all sorts of political and ethical issues, but they were able (as graduate students) to successfully design their studies and get them approved through their IRB, working in concert with local authorities. It's hard, but with adequate institutional support, legwork, and a knowledgeable PI, it's eminently doable. Not to mention all of the clinical psych and econ grad students I know who got their own data rather than relying on existing datasets.

    I don't deny that it's hard and that it can and should be made easier for certain classes of research. And I also acknowledge that most of my personal knowledge comes from people working at institutions that will obviously have a very well developed IRB apparatus (generally affiliated with world-class hospitals and deep-pocketed universities associated with them). But that's precisely my point: either your institution should do a good job streamlining the process as much as legally and ethically possible, or they should turn to an external organization that is good at this sort of thing. Small companies routinely use external contract IRBs for their human subjects work and it works just fine.
    Disciplines have different norms about working with advisors. In the social sciences, you're not spending most of your time in your advisor's lab. While you might team up for some research, there's only so much assistance you can expect to receive. That means when you expect that grad student to be "informed", you really mean they should have experience they've had no opportunity to obtain or work with an advisor who has little incentive to dedicate a substantial amount of their time to getting a project through the IRB. In cases where there's serious risk to the subjects of an experiment, you can say "too bad." But that's not what most social science experiments look like. This means the institution's IRB might very well have a lot of experience dealing with hard science requests, but then shows little sympathy for the nuances of other disciplines.
    Hope is the denial of reality

  5. #35
    Quote Originally Posted by Loki View Post
    Disciplines have different norms about working with advisors. In the social sciences, you're not spending most of your time in your advisor's lab. While you might team up for some research, there's only so much assistance you can expect to receive. That means when you expect that grad student to be "informed", you really mean they should have experience they've had no opportunity to obtain or work with an advisor who has little incentive to dedicate a substantial amount of their time to getting a project through the IRB. In cases where there's serious risk to the subjects of an experiment, you can say "too bad." But that's not what most social science experiments look like. This means the institution's IRB might very well have a lot of experience dealing with hard science requests, but then shows little sympathy for the nuances of other disciplines.
    In a big STEM research lab, you're lucky if you meet your advisor once a week for half an hour. They're not going to write your protocol for you (or even read it), but they will point you in the right direction for resources on how to do it right. And typically those resources are enough to write a decent protocol on your own, with appropriate revisions from the first pass of the IRB. It's bureaucracy, not rocket science.

    Also, big institutions typically have different IRBs for their clinical research through the med school (which has a much higher probability of risks) and their other human subjects research in the rest of the university - in fact, you'll often also have a separate IRB for schools of public health or other institutions that do a lot of human subjects research with specific needs. So you're not exactly getting a bunch of MDs who couldn't care less about your obscure social science question.

    I readily recognize that this is not a perfect system, but for major institutions I think you exaggerate the problems. And for minor institutions, I question whether they shouldn't rely on outside sources to streamline their process. My wife's company routinely does human subject research of the low risk variety, and they typically get turnaround on IRBs on the order of days. For someone doing it for the first time, I have no doubt it would take a lot longer to draft a protocol, but once you've done it once it's orders of magnitude easier to do it again.
    "When I meet God, I am going to ask him two questions: Why relativity? And why turbulence? I really believe he will have an answer for the first." - Werner Heisenberg (maybe)

  6. #36
    Loki, it just sounds like an remarkably adversarial relationship. No doubt legal aspects are different in the US but I have never experienced or heard of anyone else experiencing that kind of attitude from our equivalent. Objections have always been reasonable and clear, often with suggestions that improve the protocol both wrt the ethical aspects and wrt the science. Designing a protocol and applying to our equivalent of an IRB is a part of the basic training of students in medicine, biomedicine and some other subjects, and I don't think I've ever heard of anyone encountering that kind of soul-crushing resistance, although obviously you have more help from senior researchers at that stage.

    Perhaps it is a cultural thing or perhaps they have difficulties dealing with non-medical research. It's possible that Swedish IRBs are more permissive--even dangerously and scandalously so --but it seems unlikely given that papers still have to pass peer review. Incidentally, I have seen much more obtuse bullshit objections from reviewers who have very obviously not read a manuscript properly than I have seen silly objections from IRBs.

    I can't draw any conclusions from the blogger's description of his experiences. As much as I enjoy his blog, I think it's fairly clear that many of the problems he perceives and the challenges that frustrate him have more to do with his idiosyncrasies than with the malevolent or braindead obstructionism of people around him. 80% of the issues he described could have been avoided or easily resolved, unless there is something truly toxic about the institutional culture.

    On a side note, while I recognize that things may be different over there than they are here, I have yet to see a single doctor at any level use one of the screening tests he describes as a diagnostic test. At most I've encountered counselors fresh out of uni accidentally telling someone they have eg. ADHD or some form of bipolar disorder on the basis of screening tests. I always imagined American doctors would be even more rigorous than wishy-washy lazy Swedish doctors. Then again, this is just anecdotal, and my experiences may be less representative than his. But my latest psych rotation featured 3 months working primarily with patients with affective disorders, and much of my primary care rotation had a similar focus. Not saying it's not an interesting research question, just that his description of his reality struck me as very different from what I've experienced.
    "One day, we shall die. All the other days, we shall live."

  7. #37
    Getting back to the original thread topic for a moment: do you guys believe the scientists and organizations involved should be sanctioned in any way? Eg. by research institutions, journals, conference organizers etc?
    "One day, we shall die. All the other days, we shall live."

  8. #38
    Wig, this was at the U of Illinois, not exactly a tiny program. It seems a frustration shared by many political scientists and economists both there and elsewhere. I can't speak for other disciplines.

    Seems like an issue for their professional associations to consider. I don't know what kind of norms/rules exist in those.
    Hope is the denial of reality

  9. #39
    I can't speak to UIUC but I do have friends who run studies out of UIC and didn't have too much trouble. Maybe this is just a matter of perspective, or the specifics of a university/department/whatever. In my experience, though, IRBs are occasionally an annoyance but rarely so difficult to work with that they actually prevent research from happening.

    Interesting thought on the professional associations; they can't change the law directly (though certainly there might be scope to carve out some better exceptions), but perhaps they could advocate for better tools to help those performing low-risk studies to rapidly clear regulatory hurdles.

    A friend of mine is the chair of one of the IRBs at Hopkins Medicine; if I get a chance I'll try to get his perspective on this sort of issue. Obviously his IRB doesn't cover econ or poli sci, though.
    "When I meet God, I am going to ask him two questions: Why relativity? And why turbulence? I really believe he will have an answer for the first." - Werner Heisenberg (maybe)

  10. #40
    Ironically, political science associations have a less than stellar record lobbying for changes that would help political scientists.
    Hope is the denial of reality

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •