I worked in acute mental health settings for forty years, and one of the major issues was measuring patient compliance when they were out on a conditional discharge/outpatient commitment. Weekly bloodwork is expensive, and even more so for medications that every lab does not test for. Blood needs to be sent out to specialised labs. Someone suggested that simply putting a harmless dye in medications could be easily and inexpensively monitored. The researchers from Geisel Medical School at Dartmouth shook their heads. Each new compounding would have to be tested separately to be declared safe enough. Adding the dye made it a whole new ballgame.
I hope they were wrong and there were and are ways around this, but I fear they were all too accurate.
Maybe the move here is to talk to the terminal patients. They have a very sympathetic direct interest and nothing to lose. Find some who are themselves doctors or related professionals, and give them a platform
And I thought about how it was only possible specifically because it was a N=1 condition where the patient was going to be totally fucked if they didn’t get the experimental treatment, where a Hail Mary can get through the regulatory process. Were this to be something that affected more people, paradoxically this wouldn’t be attempted due to risk.
This is written spectacularly. Having worked at an earlier stage tech bio, there was often such an emphasis on finding the most likely to succeed programs that I never really took a step back to look into why the system is built the way it is. This is important work and I can’t wait to see what else this inspires.
I do sometimes wonder whether there are also principal agent problems underlying this. Besides the regulatory and compliance personnel who have no personal incentive for new therapies to make it to market, staff in biopharma have mixed incentives as well. The worst thing that can happen to someone in the industry is to have a clinical (or even preclinical) trial go badly. What follows such failures are recriminations, postmortems, and frequently layoffs. But so long as trials are delayed and results are pending, one can continue to draw a salary and sell hope to investors. The same misaligned incentives perhaps underlie all the me-too and overly-small-market drug programs. If the patient population who benefit from a drug is small or non-existent, that becomes a problem for investors and patients, but much less so for the people along the way who managed to hit all the desired endpoints. And the same thing perhaps underlies some of the recent interest in drug repurposing (not that this is necessarily a bad thing): a drug program without tox concerns will be around longer, even if it ultimately fails.
I work in software and recently was asked to make a small change to increase a limit for one of our customers. It's generally harmless and quick to do but it is a change in the live production environment. However, there has been a recent shift towards more safety and review in our company because of previous incidents caused by small changes. So these changes now require a slightly more formal review process. So when I was asked to make the change, I said this probably needs the formal review. They got back to us and said actually the review isn't required here, and so this small change took 1 day instead of a few minutes. I didn't bear the cost of the delay but I did bear the risk if something broke. I guess this is a small-scale vignette of what a lack of hobbitian courage and a system encouraging it looks like from the other side.
In the case you discuss, the culprit will be your local IRB, with its own oddly extra strict rules. But if you even gently suggest they could do it differently your research will be slow processed to nothing. None of your proposals will be approved - it’s a quasi subjective process anyway. Academics is littered with similar situations.
1) Incentive structures for decision makers within FDA reward risk aversion. There is no prize for greenlighting a transformative trial where someone gets hurt while the team is figuring things out.
2) Yes, these "safety" requirements are silly. Why on earth would we need to trace the complete origin of every single ingredient in a drug (gmp) rather than simply unambiguously validate the drug's final composition and performance? This is financially irresponsible / onerous expectation that throttles truth generating experiments in private companies and academics.
3) Executives at biotech companies should be able to dose themselves during an experiment in healthy normals and get rewarded with less draconian manufacturing regulation. Always struck me as immoral that we externalized that risk to cash strapped people anyway.
(3) is not tongue in cheek. I really think that everyone would respect biotech more if leadership quite literally led from the front / put their money where their mouth was.
I agree with your take, Ruxandra, and I actually just wrote a post about it. There's a tendency for regulations to be over-enforced by institutional bureaucracies. I call it the "regulatory cascade", and I blame a combination of vague regulations, severe punishments, and principal-agent problems. You can read much more here:
I also want to call out a specific flavor of policy that makes this problem worse: "risk-based" regulation or "risk-proportionate" regulation. It happens when an agency like FDA or some other agency issues guidance to industry saying "You don't have to follow all the rules for everything - just focus on the high-risk stuff." In the hands of corporate bureaucracies, everything is then interpreted as high risk. That's what happened with the Phase I GMP requirements. We also see it with IRBs and the new clinical trial GCP regulations. At best it accomplishes nothing but at worst it breeds complacency since agencies think they've granted flexibility even though nothing on the ground has changed.
I worked in acute mental health settings for forty years, and one of the major issues was measuring patient compliance when they were out on a conditional discharge/outpatient commitment. Weekly bloodwork is expensive, and even more so for medications that every lab does not test for. Blood needs to be sent out to specialised labs. Someone suggested that simply putting a harmless dye in medications could be easily and inexpensively monitored. The researchers from Geisel Medical School at Dartmouth shook their heads. Each new compounding would have to be tested separately to be declared safe enough. Adding the dye made it a whole new ballgame.
I hope they were wrong and there were and are ways around this, but I fear they were all too accurate.
Maybe the move here is to talk to the terminal patients. They have a very sympathetic direct interest and nothing to lose. Find some who are themselves doctors or related professionals, and give them a platform
That’s a good idea. Yes, they do not have anything to lose.
Have you seen Bess Stillman’s substack? She’s an ER doctor whose husband died of a terminal disease, and wrote movingly about the urgency of clinical trials https://open.substack.com/pub/bessstillman/p/just-because-you-wont-see-it-doesnt
And from him directly:
https://jakeseliger.com/2023/07/22/i-am-dying-of-squamous-cell-carcinoma-and-the-treatments-that-might-save-me-are-just-out-of-reach/
And this from her:
https://bessstillman.substack.com/p/please-be-dying-but-not-too-quickly
I thought this was cool: https://www.research.chop.edu/cornerstone-blog/in-landmark-study-chop-penn-team-treats-newborn-with-base-editing-therapy
And I thought about how it was only possible specifically because it was a N=1 condition where the patient was going to be totally fucked if they didn’t get the experimental treatment, where a Hail Mary can get through the regulatory process. Were this to be something that affected more people, paradoxically this wouldn’t be attempted due to risk.
This is written spectacularly. Having worked at an earlier stage tech bio, there was often such an emphasis on finding the most likely to succeed programs that I never really took a step back to look into why the system is built the way it is. This is important work and I can’t wait to see what else this inspires.
I do sometimes wonder whether there are also principal agent problems underlying this. Besides the regulatory and compliance personnel who have no personal incentive for new therapies to make it to market, staff in biopharma have mixed incentives as well. The worst thing that can happen to someone in the industry is to have a clinical (or even preclinical) trial go badly. What follows such failures are recriminations, postmortems, and frequently layoffs. But so long as trials are delayed and results are pending, one can continue to draw a salary and sell hope to investors. The same misaligned incentives perhaps underlie all the me-too and overly-small-market drug programs. If the patient population who benefit from a drug is small or non-existent, that becomes a problem for investors and patients, but much less so for the people along the way who managed to hit all the desired endpoints. And the same thing perhaps underlies some of the recent interest in drug repurposing (not that this is necessarily a bad thing): a drug program without tox concerns will be around longer, even if it ultimately fails.
I work in software and recently was asked to make a small change to increase a limit for one of our customers. It's generally harmless and quick to do but it is a change in the live production environment. However, there has been a recent shift towards more safety and review in our company because of previous incidents caused by small changes. So these changes now require a slightly more formal review process. So when I was asked to make the change, I said this probably needs the formal review. They got back to us and said actually the review isn't required here, and so this small change took 1 day instead of a few minutes. I didn't bear the cost of the delay but I did bear the risk if something broke. I guess this is a small-scale vignette of what a lack of hobbitian courage and a system encouraging it looks like from the other side.
In the case you discuss, the culprit will be your local IRB, with its own oddly extra strict rules. But if you even gently suggest they could do it differently your research will be slow processed to nothing. None of your proposals will be approved - it’s a quasi subjective process anyway. Academics is littered with similar situations.
1) Incentive structures for decision makers within FDA reward risk aversion. There is no prize for greenlighting a transformative trial where someone gets hurt while the team is figuring things out.
2) Yes, these "safety" requirements are silly. Why on earth would we need to trace the complete origin of every single ingredient in a drug (gmp) rather than simply unambiguously validate the drug's final composition and performance? This is financially irresponsible / onerous expectation that throttles truth generating experiments in private companies and academics.
3) Executives at biotech companies should be able to dose themselves during an experiment in healthy normals and get rewarded with less draconian manufacturing regulation. Always struck me as immoral that we externalized that risk to cash strapped people anyway.
(3) is not tongue in cheek. I really think that everyone would respect biotech more if leadership quite literally led from the front / put their money where their mouth was.
Can retirees speak up? Or is this for research that they largely haven't kept up with.
I agree with your take, Ruxandra, and I actually just wrote a post about it. There's a tendency for regulations to be over-enforced by institutional bureaucracies. I call it the "regulatory cascade", and I blame a combination of vague regulations, severe punishments, and principal-agent problems. You can read much more here:
https://learninghealthadam.substack.com/p/the-badness-police
I also want to call out a specific flavor of policy that makes this problem worse: "risk-based" regulation or "risk-proportionate" regulation. It happens when an agency like FDA or some other agency issues guidance to industry saying "You don't have to follow all the rules for everything - just focus on the high-risk stuff." In the hands of corporate bureaucracies, everything is then interpreted as high risk. That's what happened with the Phase I GMP requirements. We also see it with IRBs and the new clinical trial GCP regulations. At best it accomplishes nothing but at worst it breeds complacency since agencies think they've granted flexibility even though nothing on the ground has changed.