I can tell you that we doctors prescribe things off label all day long. The problem is is that the doctors who chose to do this with other drugs were vilified and threatened and cajoled and are probably scared stiff to venture into these Waters without formal approval. You reap what you sow
I am reminded of Keynes' comment on bankers and banking, that failure is perfectly acceptable among bankers and results in no real consequences, as long as they fail in a strictly conventional way.
As a dude who might get covid at some point: is there any way to get this ten-day course of fluvoxamine for use just-in-case I get sick? Or do I need to wait until I'm already unambiguously sick, and then have the conversation with my doctor?
That drug still worries me. It is a psychotropic and I still think there are better alternatives.
Apart from the other therapeutics (monoclonals being a major one). I'm getting more interested in HCQ, which I was highly dismissive of under Trump, but as I look it more deeply (ignoring the several deeply flawed studies), may do some strong work if given in the first few days. Ivermectin? Who knows. I do think it also has promise.
We also keep ignoring zinc and D3, which are just amazing natural supplements in their ability to combat viruses.
This is a great article and I hope more doctors who treat Covid patients read it, but the FDA didn't approve fluvoxamine for depression. It's only approved for OCD, and used off-label for things like depression and anxiety.
Regarding fear of lawsuits, FDA approval isn't the only factor. Most doctors want to follow the "standard of care", which basically means doing what other doctors are doing. This leads to an obvious chicken-and-egg problem.
I am going to be an asshole and say I first asked you about fluvoxaminne in the comments March 31st after the first small trial came out and 60 minutes did their report. The lack of cost benefit analysis from the FDA and CDC is absurdly frustrating.
You brought up a good point which is they FDA processes require a sponsor. Without a sponsor nothing happens. Someone needs to file the paperwork. If any rich people are reading best case is probably create a foundation, hire some regulatory professionals and get them to do it. Better yet get a pharma company to do it. I might ask my employer.
It is on both the doctors and the FDA. We could live in a world where people could just get fluvoxamine without a prescription and the consequences would fall on them. Instead the blame falls on the doctors or the FDA or the politicians and they behave too cautiously.
You could do a cost-benefit analysis comparing lives saved v. lives lost but I think it is more egregious to kill someone by preventing them having access to life saving medication than to let someone kill themselves by taking medicine they shouldn't for whatever reason. It seems like there isn't even a willingness to compare lives v. lives but a standard of safety that is probably too high in many situations. For example, considering withdrawal from SSRIs on one side while death is on the other.
Here in my country (western Europe), generalists have been strangely quasi absent from covid crisis, it was all urgency plus retirement home at the beginning, at after mass vaccination in new temporary structures, with a big state reservation / tracking system. It's not only prescribing strange (old) drugs, it's the individualized and decentralized medecine that is sidestepped. I wonder if it will permanent.
Hello sorry to have been gone so long, I owe some responses in past threads.
Regarding:
> It’s not illegal to prescribe fluvoxamine for COVID. It’s not even going to get you in any trouble. It might not get covered by insurance, but it only costs about $10 anyway. The problem is just that it’s weird.
Wrong. You can get in trouble. And you will. Applying this same logic to Ivermectin and Hydroxychloroquine, these too have the same right to be prescribed by a doctor for off-label use. However doctors who have in the last year, are punished, vilified and attacked, reprimanded and fired. I have many many examples, Dr Paul Marik being one, but you can easily find hundreds and probably thousands of examples of this on your own.
We've opened up the flood gates and "totally not put doctors in trouble" for prescribing things for off label use, except we have. The number doctors who feel "weird" prescribing things the NIH and FDA don't explicitly approve, will only increase, as we apply higher selection pressures for the system to vilify or remove any who go against it. Good luck finding doctors who will prescribe this.
What about patients (such as some of yours, no doubt) who are already taking SSRIs for depression? Would there be risks of adding fluvoxamine on top of that? Serotonin syndrome?
The whole bloody system is set up for pharma, not for us - that's why. Why wouldn't there be widescale searching and trials for ALL possible helpful substances, not just pharmaceuticals, ala zinc and nigella sativa, from the very start?
One of the big knocks against ivermectin is the (lack of) obvious mechanism for it to work against covid at non-toxic doses.
But, like -- at least there *is* a mechanism! It has been shown to inhibit replication at (intolerably) high doses in vitro, and killing worms (which it definitely does) is *also* a potential mechanism.
What is the possible mechanism for a random SSRI to inhibit covid? It seems like many of the arguments that it's unlikely that a random dewormer would do anything apply in spades -- I am pretty ignorant on this, so maybe there's something obvious that I just don't know about -- but it seems like a weirdness that should be addressed. Otherwise given the mixed evidence from various ivermectin trials, the Insanity Wolf approach (just take everything) seems... kind of OK? (if a patient has covid and seems to be at high risk for bad outcomes, of course)
I strongly believe in what Scott calls the "meta-level heuristic that you need in order to practice good medicine".
Ivermectin seems to be a good argument for this. After being promoted by a bunch of people (the vast majority of whom meant well, I assume) and then becoming a political football, it's now so firmly lodged in people's minds that it takes massive institutional power to dislodge it. I don't understand how that's happened, but there are lots of other examples. And they will keep on emerging, and keep on distorting medicine. The only corrective that I can see is strong institutions. To give an analogy: they are signal-boosters that stop the signal being overwhelmed by massive noise. Even if institutions are imperfect signal boosters, they're still better than just the noise.
So I think doctors are right to follow the institutions as best they can most of the time. However, this post also seems to me to be pretty much the thing the internet was made for. This is what they promised us in the golden age of information: smart people pooling their knowledge for the greater good. I think these two intuitions are somewhat contradictory, and I have little hope of resolving this contradiction any time soon!
God that's depressing. It kind of says everything you need to know about the FDA that they're incapable of doing anything without a Pharma company with a suitcase full of money taking point.
I am going to be the Devil's advocate here and argue for why you should trust the Devil. It is because the consequences of not doing so are *illegible*. Like, what do you think will happen when you do this "slightly weird" thing? The worst case scenario isn't having to explain it to your colleagues, it's having it come up in conversation with someone a large amount of inferential distance away from where you are, but who you still respect the opinion of. (I don't know of a good example because I don't know what your social world, or the social world of a hypothetical person wanting but afraid to prescribe fluvoxamine, is like.) How are you going to justify yourself in that context, huh? (And you need to justify yourself, and not just defend yourself from threats, to protect your ego -- otherwise your mind will protect your ego with rationalizations, which are bad.)
The other worst case scenario is that you feel *yourself* the need to explain and justify your decision to someone, because that's how the human mind works. The decision of who to choose for this purpose can be made on many bases, such as previous trust, relevant expertise, and relevant domain knowledge. This post seems to point to Scott Alexander as an answer to the last question but doesn't say anything about the first two ("expertise" being thought of in a credentialed manner). In other words, talking to a representative of the Devil is easier and less scary than talking to Scott Alexander. Since fear usually represents some genuine danger that we should be worried about, this is an indication to trust the Devil.
I have said all of this as a Devil's advocate, or as a non-serious post. My serious reaction is that Scott is doing a good thing by providing his email for the purposes of having discussions on the psychology of providing fluvoxamine to patients. And I am defying the Devil myself somewhat by talking about His mechanisms. Hopefully I will not pay for it too badly :)
There is something wonderfully different about listening to American discussions about medicines. I think it is because in Britain all advertising of prescription medicine is illegal, so we never get into these debates where we fire off the names of what sound like summaries of the latest Star Wars movie.
Five years ago, I could definitely have tagged this onto the SW Wikipedia page and got away with it for a few days.... "At the battle of Covid 19, the forces of the Fluvoxamine Alliance face the combined might of Sith Lord Darth Luvox, and the Ivermectin insurgents..."
I assume that there is some cross over between pulp sci fi authors and drug marketeers.
This is a bit off-topic, but how come an antidepressant can treat COVID ? I'm no doctor nor a biologist, so it sounds to me like saying that "ice cream can be used as an effective shaped explosive", i.e. somewhat surprising.
Amazing, so you alone debunked the snake oil option and discovered the real thing, no one was talking about. You really are the smartest of them all, i knew it.
Is the evidence that Luvox helps purely empirical - of the "people threw lots of things at the wall and one stuck a little" variety? Is there a plausible mechanism by which Luvox should help with COVID?
If you want to do something even slightly unusual you get a lot of nonsensical opposition from senior clinicians. finding even a single example of a hospital anywhere in the country doing it already is like a magical key.
Also it seems the UK's solution is NICE. Their job explicitly includes sorting out treatment protocols and recommending likely treatments to doctors
You mentioned before about doctors in the US being a bit blind to stuff in the same field abroad.
You might find the NICE page on covid drugs interesting.
Thank you for opening up about the moral dilemmas here which apply in all areas of life and especially where reputations and careers appear to be at stake. The price we pay for society is becoming clearer by the day and some of those age-old bargains are looking a little threadbare
This is really two posts. One about whether doctor should prescribe fluvoxamine for COVID (Yes)
The other is an aside but really much more important, it’s a problem with the basic set up of DFA that goes back to why it was founded. FDA is there to keep quack medicines and medical devices from being widely sold. There are some Libertarians who disagree with that goal but Scott does not and it’s not what he is writing about here.
There is also a critique of FDA that its procedures for keeping quack drugs off the market are flawed, that it pays too much attention to possible costs (including the cost of delay) and not enough to benefits (including the benefits beyond those to the recipient to others in the case of vaccines) of new drugs. I suspect that is true and it’s a long-standing problem, but that is not exactly what Scott is addressing, either.
Rather, it’s that the DFA is entirely reactive. If there is not a market player pushing DFA for a regulatory action – the issue here is changing the label of fluvoxamine to include COVID as a possible use – it does not happen. But this was the problem with screening test kits early on. There was no market “demand” for gazillions of kits whose use would be to screen lots of asymptomatic people so they would self-isolate. [FDA also dragged its feet on approving cheap fast for which there WAS demand but that was the long-standing problem of over-weighing risks relative to benefits.] Another example was the non-decision to use fractional doses of vaccines and to delay second doses to allow more first doses. There was no one pushing for that decision and it did not get consideration.
Is there any reason you don't ask the question "why do we require prescriptions for medicine that's already approved?"
Doesn't "prescriptions are optional for legal substances" just make this _entire problem_ go away if we let people make their own choice and suffer the consequences if they choose poorly?
On one hand, I spent my entire life until my mid-20s in a depressed fog, barely able to manage graduating at all (let alone fulfill all the wild dreams of researching this and that thing, writing that or this thing, tackling this or that ambition, etc etc).
The only reason I did manage to graduate, in fact — that I managed to *keep going* — was that I found fantastic reserves of joy and energy through hydrocodone, oxycodone, morphine, and the rest. When the choice was "become addicted or kill yourself", well...
On the other hand, this quickly became a nightmare too: super addicted, as predicted; could no longer afford enough; could no longer find suppliers; withdrawing all the time...
I had thought, before, that I'd rather be dependent upon opioids and alive than sober and dead, but I hadn't thought about what might happen if I couldn't get them any more. I was very bitter back then: "it's like $10 for a month's supply at the goddamn pharmacy; why can I not just be prescribed some weak opioid and *I'll* deal with the dependence?!"
I can see why the caution, of course. Once I started, I couldn't stop. My life is great now, and I have almost everything I've ever wanted — because I finally did get that prescription (for buprenorphine); even after friends dying from opioid abuse, even after all the heartache, I couldn't stop without the substitute of bupe.
But also... I don't think I regret it. I spent a decade in therapy and trying every possible drug, even ketamine; I even managed to charm a Desoxyn script (+ alprazolam!) out of my psychiatrist to trial it just for a month; and... nothing. I was ready to die. Opioids are the only thing that helped.
Do I have a point here? Uh... Yes, but the margins of this comment are too small to contain it.
This is something I wanted to ask as early as the IVM post, but the comments on it flew very fast and I thought it would be unlikely anyone would see it.
I work and have academic background in social sciences, an area strongly removed from medicine and/or biostatistics. In our field, no one would treat you with any degree of seriousness if you were to propose any kind of solution to any problem based purely on empirics without any sort of model that explains why what you are doing should work (not just a model in the rationalist sort of sense, you'd be expected to present a system of equations and how they relate to each other). In medicine, this seems very widespread - I opened 20 random Wikipedia pages on medicines that are on the WHO Essential list, and close to half of them have "no well understood" or "a subject of discussion" mechanisms of action, although their effects are empirically well studied.
I don't propose that our way of doing things is better than the medical way of doing things. But what's the reason there's such a big difference between fields in this?
Scott, I am not a doctor, but people who are doctors have complained to me about things called "carepaths" where there are relatively strongly entrenched algorithms for treating X, and if the doctor deviates from them, insurance won't cover it, their hospitals might get angry with them, and sometimes pharmacists (citing a relatively recent law perhaps designed to prevent overuse of opiods?) won't fill the prescriptions. Do you think this has any effect? (non-rhetorical, I am genuinely curious)
I noticed the dosages in the TOGETHER trial (100mg, 2x daily for 10 days after positive test) are in line with a typical adult dose of fluvoxatine for OCD/depression.
The on-label treatment assumes the patient will use the drug indefinitely, so tolerance for side effects is pretty low. Nobody wants nausea, emotional blunting, or sexual side effects long-term if they can be avoided.
It seems like side effect tolerance for a ten day treatment against a potentially deadly disease would be very different, so maybe higher dosages might make sense?
I think it's reasonable to be cautious given the unknowns, so I don't want to suggest people should start self-medicating with 10x doses or anything. Just wondering though if larger dosages are being studied and if we know anything about anti-inflammatory effects at higher dosages or significant escalation of risks?
Has anyone suggested a biological mechanism by which fluvoxamine helps treat COVID? I know almost nothing about pharmaceuticals, but what little I do know about SSRIs makes me think that they mostly affect brain chemistry. But COVID is a respiratory disease. Am I wrong about SSRIs?
If it really works, then it really works, but I would love to know more about why. Maybe this is the weird thing that Scott was alluding to in the article?
I'm not particularly left-leaning, or even very anal about word choice. But Scott using "man up" at the end made me uncomfortable, and not least because he is addressing female (and other non-binary) doctors too. I hope he's using it ironically in a way that I'm too stupid to understand.
I was hoping to see a short consideration of ongoing drug trials for fluvoxamine. They could help hesitant doctors thread the needle. They don't have to stake a claim for whether they think it works. They give their patients that have COVID and want something a "something." They help contribute to science!
I don't know how common it is for a doctor to recommend clinical drug trials to patients in most settings. Anecdotally, it's never happened to me. But there are a couple trials actively recruiting in the US that doctors could talk to patients about. (Note, I have no affiliation with either of these trials, but I saw them mentioned on Twitter.)
I was wondering about the potential for one of more of the large health insurance companies (United, Aetna, Cigna) to act as the sponsor for a drug like fluvoxamine. It seems like they would be have the money and proper incentives to play the role. If a cheap generic like this actually works, it could save them a fortune in hospital claims that they would otherwise be on the hook for. Seems like they could create a foundation, hire experienced people from pharma companies who have shepherded drugs through the process, and get to work.
I’m sure there’d be some pushback since everyone seems to hate on insurance companies but it’s hard for me to see how they would be any more conflicted than the pharma companies who make and profit off the drugs directly. I mean, at least they are on the hook for claims if there are side effects or situations where it doesn’t work. Unlike the drug companies, they only profit if it works and saves them claims down the line.
Your anecdote about ketamine is the reason I've never pursued a medical degree. I can give people who have treatment-resistant depression ketamine and shrooms. I can give MDMA to people with PTSD. I have done so many times with great success and it's a wonderful feeling. If I were an accredited doctor then doing all that stuff would cause me to lose my license at best.
If it's any consolation, the engineers' budget ventilator approach probably wouldn't have been very helpful; in the early days especially, we were short on fancy vents, the kind that could do advanced modes of ventilation to treat ARDS. Being on a vent is generally bad for COVID, but if you're going to be on the vent you should ideally be on the kind of computerized ventilation that can micromanage pressures, not the kind of simple machine that runs on gas pressure. Effective mechanical ventilation is an esoteric art that even many MDs don't understand, and there was some serious miscommunication going on in the early days of COVID as to what was needed. In purely consequentialist terms, I doubt you did any significant harm.
(I'm a respiratory therapist, though I was still in school when COVID hit)
Here is the thing, and my understanding of this is quite imperfect. There are multiple purposes the FDA serves. One purpose, which is pretty important and actually takes a lot of work but is viewed as boring and taken for granted by many, is to make sure the actual manufacturing of drugs is done correctly. That every pill of your generic aspirin constains exactly the same dose as the label says even though dozens of generic companies are making the low margin drug trying to keep costs as low as they can.
For a new drug, something doctors have no experiences with, the fear level can be high and the FDA I think does serve as providing a blessing. A doctor can say "I don't know anything about this drug for shingles but the FDA says this drug works for patients in this group so I'm not just taking the word of the 20-something drug rep". Since drug companies are only allowed to market drugs they have approval from the FDA for and they can only market for indications that are on the FDA approved label, there's an obvious profit incentive for them to spend the money to go out and do the studies to establish the bona fides of the new drug or to prove a drug that is already approved for one thing also works for something else.
But older drugs are a bit different. Doctors presumably have real life experience with patients on older drugs so they should feel more open to trying them if there's reason to believe they may work in an unexpected place. If some study said low dose aspirin helped certain types of depression, you should not feel a huge amount of fear trying it on your patients 'just to see'. Yes if some generic aspirin manufacturer spent $75M to run a clinical trial and submit it to the FDA for a label expansion you'd feel better, but is that the best use of resources? I'm not sure.
So I'm not sold on the idea that the FDA should go from passive entity here to active. By passive I mean companies come to the FDA asking if they can market a new drug or add something to the label of their existing drug. The FDA convenes the advisory boards with experts on the disease area and evaluates their evidence but they do the work of collecting the evidence, hoping for the best.
But should the FDA go active? Going out and trying to find new labels for old drugs? Picking up studies others did to try to add to the label rather than waiting for people to ask them to do it? The FDA does have limited resources and it isn't like you can just throw money to quickly expand it to infinity. It's not like disease experts grow on trees after all.
More importantly if the FDA did start doing this, I would fear that it could lead to the less formal knowledge generating institutions of the medical field starting to atrophy. Medicine form my impression is a combination of theretical and practical knowledge. Doctors combine a lot of information from less formal sources. These include the experiences of other doctors, articles in the literature (not just highly controlled RCT's but case reports, the chatter at conventions and meetings, etc. I suspect making the FDA pro-active in going out to find label expansions for existing drugs will cause doctors to pay less attention to these others areas of knowledge generation and they will atrophy.
Now that being said, doctors have seemed to be totally willing to write a lot of scripts to try to fight Covid. How many ivermectin and HCQ scripts were written? OK the FDA did give HCQ temporary approval but still, those scripts are still being written. I recall when the first cases happened news reports said doctors were trying tamiflu. Yet tamiflu isn't approved for Covid and even for the flu it only does good if you start it early but they were clearly trying anything that seemed like it may have a plausible path to working.
I think the systemic issue goes beyond the FDA. There are many other institutions that influence what doctors feel comfortable prescribing, and in principle that makes sense: individual doctors can't conduct large-scale medical research on their own, so they really need to be able to trust what's coming out of institutions. As well as the FDA, we should look at:
1. Popular medical resources like Uptodate, which provide information on common treatments including off-label prescribing. Lots of doctors consult this many times a day. Its article on fluvoxamine begins with a warning about Covid treatment that says there is "insufficient data to recommend for or against." I think persuading the editors of Uptodate to actually update their article in light of the TOGETHER trial could make a big difference.
2. Insurance companies, which put barriers in the way of unusual (or expensive) treatment. Refusing pointblank to cover something can get them into trouble so they're usually more devious, requiring a doctor to fill out a long form called a "prior authorization" that details all the other treatments a patient has already tried. The point of these forms seems to be to deter patients from filling their prescriptions, as getting the form and then getting the doctor to complete it can take a lot of time and effort. Patients usually don't know about this requirement until they get to the pharmacy (and doctors can't keep track of the constantly changing requirements of all the different insurance companies).
I'm not aware of any insurance companies requiring this for fluvoxamine, but the ~$10 cost Scott mentions is only at specific pharmacies using a coupon from somewhere like Goodrx. Thanks to our ridiculous drug pricing system, the walk-up cost is closer to $100.
3. State regulators, which have threatened to take away the medical licenses of doctors who write inappropriate mask or vaccine exemptions. I think the state boards are probably right to do this on balance, but it still makes doctors very reluctant to deviate from the consensus on Covid.
4. Large hospital systems and medical groups, which heavily influence the prescribing decisions of the doctors who work for them. The people in charge of these are usually far removed from clinical practice.
5. Supply chain issues, which resulted in shortages of HCQ last year. Do pharmacies have enough fluvoxamine to give it to everyone with Covid? There is a very obvious long-term solution if Covid is going to be endemic, but it could be a problem in the current surge.
Talking of early treatments, doctors going off protocol but researching themselves and treatments that don't get the funding to gather the large studies that are usually required in medicine:
Have anyone here already heard of Dr. Chetty (South Africa) and his findings? For me as a rational thinker without medical background this sounds really plausible and the way practical medicine should work.
Is there any obvious flaw i didn't get? I'm only speaking of medical or logical reasons. If he perhaps gave Interviews to the wrong people doesn't change the facts he has seen or done on his patients.
To be a little fairer to the FDA, they do seem to be aware of the drug trials but don't think there's enough evidence that fluvoxamine does make enough of a difference (EDIT not the FDA, it's the NIH):
"Anti-Inflammatory Effect of Fluvoxamine and Rationale for Use in COVID-19
In a murine sepsis model, fluvoxamine was found to bind to the sigma-1 receptor on immune cells, resulting in reduced production of inflammatory cytokines.1 In an in vitro study of human endothelial cells and macrophages, fluvoxamine reduced the expression of inflammatory genes.2 Ongoing studies are establishing whether the anti-inflammatory effects of fluvoxamine observed in nonclinical studies also occur in humans and are clinically relevant in the setting of COVID-19.
Recommendation
There is insufficient evidence for the COVID-19 Treatment Guidelines Panel (the Panel) to recommend either for or against the use of fluvoxamine for the treatment of COVID-19.
Rationale
Three randomized trials have studied the use of fluvoxamine for the treatment of nonhospitalized patients with COVID-19. In STOP COVID, a contactless, double-blind randomized placebo-controlled trial conducted in the United States among nonhospitalized adults with mild COVID-19 diagnosed within 7 days of symptom onset, fluvoxamine (100 mg up to 3 times daily for 15 days) reduced clinical deterioration at Day 15.3 Clinical deterioration was defined as shortness of breath plus oxygen saturation (SpO2) <92% or hospitalization plus SpO2 <92%. This was a small study (≤80 participants per arm) with limited cases of clinical deterioration and a short follow-up period. In addition, 24% of participants stopped responding to surveys prior to Day 15.
The subsequent STOP COVID 2, a Phase 3 randomized controlled trial (ClinicalTrials.gov Identifier NCT04668950) that enrolled >700 participants in the United States and Canada, was stopped for futility by a data safety monitoring board after lower than expected case rates and treatment effect were observed.4
TOGETHER is an adaptive platform, double-blind randomized placebo-controlled trial conducted in Brazil.5 Nonhospitalized adults with COVID-19 and a known risk factor for progression to severe disease were randomized to fluvoxamine 100 mg twice daily (n = 741) or placebo (n = 756) for 10 days. Fluvoxamine use was associated with a lower risk of the primary composite outcome of retention in the emergency department for >6 hours or admission to a tertiary hospital (79 of 741 participants [11%] in the fluvoxamine arm vs. 119 of 756 participants [16%] in the placebo arm [relative risk 0.68; 95% CrI, 0.52–0.88]). Of note, 87% of the primary outcome events were hospitalizations. There was no statistically significant difference between study arms for the secondary outcomes of need for hospitalization or time to symptom resolution. There was no significant difference in mortality between study arms in the intention-to-treat (ITT) population (17 of 741 participants [2%] in the fluvoxamine arm vs. 25 of 756 participants [3%] in the placebo arm [OR 0.69; 95% CI, 0.36–1.27]). In a secondary, per-protocol analysis of participants who received >80% of possible doses, death was the outcome for 1 of 548 participants (<1%) in the fluvoxamine arm versus 12 of 618 participants (2%) in the placebo arm (OR 0.09; 95% CI, 0.01–0.47). Participants in the fluvoxamine arm were less likely to present to an emergency setting for COVID-19 for any duration, although this analysis was not prespecified. Compared with those in the placebo arm, participants who received fluvoxamine were less adherent to therapy and discontinued therapy due to intolerance more often.
While fluvoxamine treatment significantly reduced the primary composite outcome in the TOGETHER trial (i.e., retention in the emergency department for >6 hours or admission to a tertiary hospital), the difference in hospitalizations between arms was not significant.5 Defining the clinical relevance of the >6 hour emergency department observation time endpoint is difficult, especially its applicability to practice settings in different countries. Moreover, the endpoint has not been used in other studies of interventions for nonhospitalized patients at high risk for hospitalization and death. While a per-protocol analysis found a significant treatment effect for mortality in patients taking >80% of possible doses (assessed by patient self-report), no such benefit was found in the primary ITT analysis. The 80% threshold has no clear justification, and only 74% of participants in the fluvoxamine arm reached this level of adherence. Since per-protocol analyses are not randomized comparisons, they can introduce bias when adherence is associated with factors that influence the outcome; this bias cannot be excluded in this study. Notably, mortality in the placebo arm was substantially higher in those with ≤80% adherence than in those with >80% adherence, suggesting that factors other than adherence differed in the per-protocol population. Finally, including only participants who could tolerate fluvoxamine does not reflect the actual effectiveness of the drug, since intolerance and adherence appeared to be related.
Additional studies are needed to provide more specific, evidence-based guidance on the role of fluvoxamine for the treatment of COVID-19. Further details of the studies discussed are provided in Table 4c."
That sounds less like "we can't find anyone willing to be responsible for this" and more "if some study comes out with yes yes yes it works!!!, then we'll approve it". Right now, they seem to think the studies are "meh, it's kinda okay".
Are there good studies on post-SSRI sexual dysfunction (including less visible symptoms like libido loss in women)? Wondering about the basis for Scott's characterization of the side effect as "very rare." I have some medical contacts who focus on those issues and believe that they're under-scrutinized and under-researched
Scott, I like to print out your essays to read with a cup of coffee as a break from my computer. Your writing is so good that I always enjoy them hugely, whether I'm interested in the subject matter or not.
But when I print them out from my web browser, the last paragraph of the first page is always obscured by a big 'Publish on Substack' advert, which only appears in the printed version, I don't see any such thing on the page in the web browser at all, it's only on the printout. And I can't find any way to work around this.
I realise that this is nothing to do with you, and indeed probably a bug in the Linux version of Firefox, but I wondered if you had the clout with Substack to get them to fix their button so it doesn't do this?
Let's say, purely hypothetically, that you have two octogenarian grandparents whom you love dearly. They're both vaccinated and boostered. One has recently finished bladder cancer treatment, another has a history of heart rate issues, beyond that and their age they have no major COVID comorbidities. What steps would or could you, their grandchild, take today to improve the odds that they can obtain a Fluxov prescription in the event that either one contracts COVID?
Just to point the 30% number in perspective, this is roughly equivalent to the estimated effect size for going from 0% masked -> 100% surgical masked from the Bangladesh study.
Unfortunately fluvoxamine only impacts your risk from Covid and has no real impact on psychological safety so it is utterly irrelevant to our societal response.
>The VA system took 35,000 high-risk older patients off of an unusually-likely-to-cause-QT-syndrome SSRI in 2011, and were unable to find any evidence that this prevented even a single case of the syndrome, let alone any negative outcome!
Is there a role for the physician organizations? Because I know for our kid, the recommendations about things like car seats and avoiding SIDS and stuff generally come from the American Association of Pediatrics, not the FDA but nor do individual pediatricians have to decide what to say about a lot of issues. And when my wife was pregnant we kept hearing about recommendations from the American College of Obstetricians and Gynecologists. So could one of the professional medical associations give physicians the okay to do this and know they aren’t being unprofessional?
"That is, FDA procedures usually assume there is a pharma company sponsoring a drug. But fluvoxamine is cheap and off-patent and no pharma company is involved in repurposing it for COVID. Nobody has a procedure for a drug without a sponsor, so they won’t do anything."
I'm mildly amused by the happenstance that this is pretty much the same thing Weinstein et al. were saying about getting Ivermectin re-labeled as a prophylactic for COVID.
No reflection on the virtues of fluxovamine, since I am not a doctor of any sort.
I find it very frustrating that a lot of the NPIs are justified in order to keep hospitals from being overwhelmed, and here is a treatment which would increase hospital capacity by 30% (or is it 43%?), which would probably have done more than any NPI to keep hospitals from being overwhelmed, but we aren't using it because there is no one to pay for marketing it.
So these doctors and nurses we see in the media who complain about how overloaded they are, they don't know about it?
I've been single for too long because asking people out would require me to do something potentially awkward.
"For some reason the same experts who don’t mind prescribing SSRIs when people have mild depression freak out about prescribing them when they’re the only evidence-based oral medication for a deadly global pandemic. "
While I agree wholeheartedly with the sentiment, correction: dexamethasone is oral.
While we're on the subjects of steroids and "flu-" drugs that 10 million Americans already use without issues, why not fluticasone inhalers to treat Covid? They're extremely safe (far fewer side effects than oral steroids) and for me they clear up all my respiratory inflammation completely within 12-48 hours depending on how bad it was before I started. A hypothetical downside is that it might stop working in a part of the lung that is filled with fluid, but as a prophylactic against pneumonia as soon as you test positive for covid it would work great (80% confidence). Maybe I'm just seeing it through rose-tinted glasses because i'm a super-responder to it. I have occasional inflammation in my lungs, not very bad. It usually stays under control with far less than the recommended 2 puffs of 110mcg twice a day. One puff once on days when I feel like I need it is usually sufficient. When I had covid in May 2020, I raised my dosage back up to what was prescribed (2 puffs twice a day) and got through it without letting my SPO2 dip below 95%. It lasted 10 days and was milder than some flus. Fluticasone inhaler was the only thing I was taking.
Scott, you shouldn't beat yourself up for not doing something which seems unlikely in retrospect to have mattered, or for hesitating to take steps which were probably legal until you had more Bayesian evidence thereof.
You should beat yourself up for not trying to convince the FDA to authorize vaccines for new variants under the precautionary principle, which seems like an iron-clad argument according to premises they might accept.
Also, stop using a technique of anti-epistemology which successfully stitches together anthropomorphism with dehumanization. FDA bureaucrats drag their feet on needed medicine because they don't understand the difference between their behavior and the archetypal young student standing up to the anti-controls expert in front of a crowd. It's almost like traditional rationality places too much emphasis on avoiding false claims, or being less wrong.
About that Bayes formula on the bottom in your subscribe thing... Bayes formula doesn't tend to give 60/40 odds.
Here's how it works: "prior probability", P(A) that a random drug is effective for COVID, is very very small, that should be understandable, right? P(A|B) is the posterior probability given the trial(B) .
The P(B) is probability of finding the drug to be effective whether it works or not. Which for this study is woefully high - there's a number of red flags (the stranger the end point, the more effective the drug was at it; non strange end points not even statistically significant). Note that P(B|A) is less than 1 (but presumably, close to 1), and to get a greater P(A|B) than P(A) you need P(B) to be small.
The end result is that your P(A) is small (fractions of percent) and your P(A|B) is higher, but still very small (fractions of percent).
To get odds close to 50/50 (like your 60/40), would require a rare numerical coincidence; the P(B) needs to be approximately equal to P(A) . That doesn't occur often, because people who could get P(B) to be this low, shoot to make it even lower.
If 3 is an issue, isn't the solution writing up some social permissions slips and handing them out to as many doctors as you can get them to?
It sounds like all you need to do is let doctors have the line "I know a bunch of other doctors that are prescribing patients with fluvoxamine for covid" and they'll start doing it themselves too.
I can tell you that we doctors prescribe things off label all day long. The problem is is that the doctors who chose to do this with other drugs were vilified and threatened and cajoled and are probably scared stiff to venture into these Waters without formal approval. You reap what you sow
I am reminded of Keynes' comment on bankers and banking, that failure is perfectly acceptable among bankers and results in no real consequences, as long as they fail in a strictly conventional way.
Remember to tell your doctor you're feeling depressed about having COVID, to clear his conscience.
As a dude who might get covid at some point: is there any way to get this ten-day course of fluvoxamine for use just-in-case I get sick? Or do I need to wait until I'm already unambiguously sick, and then have the conversation with my doctor?
It's the second thing, isn't it?
That drug still worries me. It is a psychotropic and I still think there are better alternatives.
Apart from the other therapeutics (monoclonals being a major one). I'm getting more interested in HCQ, which I was highly dismissive of under Trump, but as I look it more deeply (ignoring the several deeply flawed studies), may do some strong work if given in the first few days. Ivermectin? Who knows. I do think it also has promise.
We also keep ignoring zinc and D3, which are just amazing natural supplements in their ability to combat viruses.
This is a great article and I hope more doctors who treat Covid patients read it, but the FDA didn't approve fluvoxamine for depression. It's only approved for OCD, and used off-label for things like depression and anxiety.
Regarding fear of lawsuits, FDA approval isn't the only factor. Most doctors want to follow the "standard of care", which basically means doing what other doctors are doing. This leads to an obvious chicken-and-egg problem.
I am going to be an asshole and say I first asked you about fluvoxaminne in the comments March 31st after the first small trial came out and 60 minutes did their report. The lack of cost benefit analysis from the FDA and CDC is absurdly frustrating.
You brought up a good point which is they FDA processes require a sponsor. Without a sponsor nothing happens. Someone needs to file the paperwork. If any rich people are reading best case is probably create a foundation, hire some regulatory professionals and get them to do it. Better yet get a pharma company to do it. I might ask my employer.
It is on both the doctors and the FDA. We could live in a world where people could just get fluvoxamine without a prescription and the consequences would fall on them. Instead the blame falls on the doctors or the FDA or the politicians and they behave too cautiously.
You could do a cost-benefit analysis comparing lives saved v. lives lost but I think it is more egregious to kill someone by preventing them having access to life saving medication than to let someone kill themselves by taking medicine they shouldn't for whatever reason. It seems like there isn't even a willingness to compare lives v. lives but a standard of safety that is probably too high in many situations. For example, considering withdrawal from SSRIs on one side while death is on the other.
Here in my country (western Europe), generalists have been strangely quasi absent from covid crisis, it was all urgency plus retirement home at the beginning, at after mass vaccination in new temporary structures, with a big state reservation / tracking system. It's not only prescribing strange (old) drugs, it's the individualized and decentralized medecine that is sidestepped. I wonder if it will permanent.
Could a large part of it be the doctors don't even know? What percentage of doctors even know about this do you figure?
Hello sorry to have been gone so long, I owe some responses in past threads.
Regarding:
> It’s not illegal to prescribe fluvoxamine for COVID. It’s not even going to get you in any trouble. It might not get covered by insurance, but it only costs about $10 anyway. The problem is just that it’s weird.
Wrong. You can get in trouble. And you will. Applying this same logic to Ivermectin and Hydroxychloroquine, these too have the same right to be prescribed by a doctor for off-label use. However doctors who have in the last year, are punished, vilified and attacked, reprimanded and fired. I have many many examples, Dr Paul Marik being one, but you can easily find hundreds and probably thousands of examples of this on your own.
We've opened up the flood gates and "totally not put doctors in trouble" for prescribing things for off label use, except we have. The number doctors who feel "weird" prescribing things the NIH and FDA don't explicitly approve, will only increase, as we apply higher selection pressures for the system to vilify or remove any who go against it. Good luck finding doctors who will prescribe this.
What about patients (such as some of yours, no doubt) who are already taking SSRIs for depression? Would there be risks of adding fluvoxamine on top of that? Serotonin syndrome?
The whole bloody system is set up for pharma, not for us - that's why. Why wouldn't there be widescale searching and trials for ALL possible helpful substances, not just pharmaceuticals, ala zinc and nigella sativa, from the very start?
One of the big knocks against ivermectin is the (lack of) obvious mechanism for it to work against covid at non-toxic doses.
But, like -- at least there *is* a mechanism! It has been shown to inhibit replication at (intolerably) high doses in vitro, and killing worms (which it definitely does) is *also* a potential mechanism.
What is the possible mechanism for a random SSRI to inhibit covid? It seems like many of the arguments that it's unlikely that a random dewormer would do anything apply in spades -- I am pretty ignorant on this, so maybe there's something obvious that I just don't know about -- but it seems like a weirdness that should be addressed. Otherwise given the mixed evidence from various ivermectin trials, the Insanity Wolf approach (just take everything) seems... kind of OK? (if a patient has covid and seems to be at high risk for bad outcomes, of course)
I strongly believe in what Scott calls the "meta-level heuristic that you need in order to practice good medicine".
Ivermectin seems to be a good argument for this. After being promoted by a bunch of people (the vast majority of whom meant well, I assume) and then becoming a political football, it's now so firmly lodged in people's minds that it takes massive institutional power to dislodge it. I don't understand how that's happened, but there are lots of other examples. And they will keep on emerging, and keep on distorting medicine. The only corrective that I can see is strong institutions. To give an analogy: they are signal-boosters that stop the signal being overwhelmed by massive noise. Even if institutions are imperfect signal boosters, they're still better than just the noise.
So I think doctors are right to follow the institutions as best they can most of the time. However, this post also seems to me to be pretty much the thing the internet was made for. This is what they promised us in the golden age of information: smart people pooling their knowledge for the greater good. I think these two intuitions are somewhat contradictory, and I have little hope of resolving this contradiction any time soon!
God that's depressing. It kind of says everything you need to know about the FDA that they're incapable of doing anything without a Pharma company with a suitcase full of money taking point.
I am going to be the Devil's advocate here and argue for why you should trust the Devil. It is because the consequences of not doing so are *illegible*. Like, what do you think will happen when you do this "slightly weird" thing? The worst case scenario isn't having to explain it to your colleagues, it's having it come up in conversation with someone a large amount of inferential distance away from where you are, but who you still respect the opinion of. (I don't know of a good example because I don't know what your social world, or the social world of a hypothetical person wanting but afraid to prescribe fluvoxamine, is like.) How are you going to justify yourself in that context, huh? (And you need to justify yourself, and not just defend yourself from threats, to protect your ego -- otherwise your mind will protect your ego with rationalizations, which are bad.)
The other worst case scenario is that you feel *yourself* the need to explain and justify your decision to someone, because that's how the human mind works. The decision of who to choose for this purpose can be made on many bases, such as previous trust, relevant expertise, and relevant domain knowledge. This post seems to point to Scott Alexander as an answer to the last question but doesn't say anything about the first two ("expertise" being thought of in a credentialed manner). In other words, talking to a representative of the Devil is easier and less scary than talking to Scott Alexander. Since fear usually represents some genuine danger that we should be worried about, this is an indication to trust the Devil.
I have said all of this as a Devil's advocate, or as a non-serious post. My serious reaction is that Scott is doing a good thing by providing his email for the purposes of having discussions on the psychology of providing fluvoxamine to patients. And I am defying the Devil myself somewhat by talking about His mechanisms. Hopefully I will not pay for it too badly :)
There is something wonderfully different about listening to American discussions about medicines. I think it is because in Britain all advertising of prescription medicine is illegal, so we never get into these debates where we fire off the names of what sound like summaries of the latest Star Wars movie.
Five years ago, I could definitely have tagged this onto the SW Wikipedia page and got away with it for a few days.... "At the battle of Covid 19, the forces of the Fluvoxamine Alliance face the combined might of Sith Lord Darth Luvox, and the Ivermectin insurgents..."
I assume that there is some cross over between pulp sci fi authors and drug marketeers.
This is a bit off-topic, but how come an antidepressant can treat COVID ? I'm no doctor nor a biologist, so it sounds to me like saying that "ice cream can be used as an effective shaped explosive", i.e. somewhat surprising.
Amazing, so you alone debunked the snake oil option and discovered the real thing, no one was talking about. You really are the smartest of them all, i knew it.
Is the evidence that Luvox helps purely empirical - of the "people threw lots of things at the wall and one stuck a little" variety? Is there a plausible mechanism by which Luvox should help with COVID?
I work with a lot of clinicians.
If you want to do something even slightly unusual you get a lot of nonsensical opposition from senior clinicians. finding even a single example of a hospital anywhere in the country doing it already is like a magical key.
Also it seems the UK's solution is NICE. Their job explicitly includes sorting out treatment protocols and recommending likely treatments to doctors
You mentioned before about doctors in the US being a bit blind to stuff in the same field abroad.
You might find the NICE page on covid drugs interesting.
https://www.nice.org.uk/covid-19/rapid-c-19-treatments-currently-monitored
Thank you for opening up about the moral dilemmas here which apply in all areas of life and especially where reputations and careers appear to be at stake. The price we pay for society is becoming clearer by the day and some of those age-old bargains are looking a little threadbare
I think Lewis's "Inner Ring" is one of the best things ever written.
This is really two posts. One about whether doctor should prescribe fluvoxamine for COVID (Yes)
The other is an aside but really much more important, it’s a problem with the basic set up of DFA that goes back to why it was founded. FDA is there to keep quack medicines and medical devices from being widely sold. There are some Libertarians who disagree with that goal but Scott does not and it’s not what he is writing about here.
There is also a critique of FDA that its procedures for keeping quack drugs off the market are flawed, that it pays too much attention to possible costs (including the cost of delay) and not enough to benefits (including the benefits beyond those to the recipient to others in the case of vaccines) of new drugs. I suspect that is true and it’s a long-standing problem, but that is not exactly what Scott is addressing, either.
Rather, it’s that the DFA is entirely reactive. If there is not a market player pushing DFA for a regulatory action – the issue here is changing the label of fluvoxamine to include COVID as a possible use – it does not happen. But this was the problem with screening test kits early on. There was no market “demand” for gazillions of kits whose use would be to screen lots of asymptomatic people so they would self-isolate. [FDA also dragged its feet on approving cheap fast for which there WAS demand but that was the long-standing problem of over-weighing risks relative to benefits.] Another example was the non-decision to use fractional doses of vaccines and to delay second doses to allow more first doses. There was no one pushing for that decision and it did not get consideration.
Is there any reason you don't ask the question "why do we require prescriptions for medicine that's already approved?"
Doesn't "prescriptions are optional for legal substances" just make this _entire problem_ go away if we let people make their own choice and suffer the consequences if they choose poorly?
Opioids are hard.
On one hand, I spent my entire life until my mid-20s in a depressed fog, barely able to manage graduating at all (let alone fulfill all the wild dreams of researching this and that thing, writing that or this thing, tackling this or that ambition, etc etc).
The only reason I did manage to graduate, in fact — that I managed to *keep going* — was that I found fantastic reserves of joy and energy through hydrocodone, oxycodone, morphine, and the rest. When the choice was "become addicted or kill yourself", well...
On the other hand, this quickly became a nightmare too: super addicted, as predicted; could no longer afford enough; could no longer find suppliers; withdrawing all the time...
I had thought, before, that I'd rather be dependent upon opioids and alive than sober and dead, but I hadn't thought about what might happen if I couldn't get them any more. I was very bitter back then: "it's like $10 for a month's supply at the goddamn pharmacy; why can I not just be prescribed some weak opioid and *I'll* deal with the dependence?!"
I can see why the caution, of course. Once I started, I couldn't stop. My life is great now, and I have almost everything I've ever wanted — because I finally did get that prescription (for buprenorphine); even after friends dying from opioid abuse, even after all the heartache, I couldn't stop without the substitute of bupe.
But also... I don't think I regret it. I spent a decade in therapy and trying every possible drug, even ketamine; I even managed to charm a Desoxyn script (+ alprazolam!) out of my psychiatrist to trial it just for a month; and... nothing. I was ready to die. Opioids are the only thing that helped.
Do I have a point here? Uh... Yes, but the margins of this comment are too small to contain it.
This is something I wanted to ask as early as the IVM post, but the comments on it flew very fast and I thought it would be unlikely anyone would see it.
I work and have academic background in social sciences, an area strongly removed from medicine and/or biostatistics. In our field, no one would treat you with any degree of seriousness if you were to propose any kind of solution to any problem based purely on empirics without any sort of model that explains why what you are doing should work (not just a model in the rationalist sort of sense, you'd be expected to present a system of equations and how they relate to each other). In medicine, this seems very widespread - I opened 20 random Wikipedia pages on medicines that are on the WHO Essential list, and close to half of them have "no well understood" or "a subject of discussion" mechanisms of action, although their effects are empirically well studied.
I don't propose that our way of doing things is better than the medical way of doing things. But what's the reason there's such a big difference between fields in this?
Scott, I am not a doctor, but people who are doctors have complained to me about things called "carepaths" where there are relatively strongly entrenched algorithms for treating X, and if the doctor deviates from them, insurance won't cover it, their hospitals might get angry with them, and sometimes pharmacists (citing a relatively recent law perhaps designed to prevent overuse of opiods?) won't fill the prescriptions. Do you think this has any effect? (non-rhetorical, I am genuinely curious)
I noticed the dosages in the TOGETHER trial (100mg, 2x daily for 10 days after positive test) are in line with a typical adult dose of fluvoxatine for OCD/depression.
The on-label treatment assumes the patient will use the drug indefinitely, so tolerance for side effects is pretty low. Nobody wants nausea, emotional blunting, or sexual side effects long-term if they can be avoided.
It seems like side effect tolerance for a ten day treatment against a potentially deadly disease would be very different, so maybe higher dosages might make sense?
I think it's reasonable to be cautious given the unknowns, so I don't want to suggest people should start self-medicating with 10x doses or anything. Just wondering though if larger dosages are being studied and if we know anything about anti-inflammatory effects at higher dosages or significant escalation of risks?
Has anyone suggested a biological mechanism by which fluvoxamine helps treat COVID? I know almost nothing about pharmaceuticals, but what little I do know about SSRIs makes me think that they mostly affect brain chemistry. But COVID is a respiratory disease. Am I wrong about SSRIs?
If it really works, then it really works, but I would love to know more about why. Maybe this is the weird thing that Scott was alluding to in the article?
I'm not particularly left-leaning, or even very anal about word choice. But Scott using "man up" at the end made me uncomfortable, and not least because he is addressing female (and other non-binary) doctors too. I hope he's using it ironically in a way that I'm too stupid to understand.
I was hoping to see a short consideration of ongoing drug trials for fluvoxamine. They could help hesitant doctors thread the needle. They don't have to stake a claim for whether they think it works. They give their patients that have COVID and want something a "something." They help contribute to science!
I don't know how common it is for a doctor to recommend clinical drug trials to patients in most settings. Anecdotally, it's never happened to me. But there are a couple trials actively recruiting in the US that doctors could talk to patients about. (Note, I have no affiliation with either of these trials, but I saw them mentioned on Twitter.)
https://covidout.umn.edu/
https://activ6study.org/
I was wondering about the potential for one of more of the large health insurance companies (United, Aetna, Cigna) to act as the sponsor for a drug like fluvoxamine. It seems like they would be have the money and proper incentives to play the role. If a cheap generic like this actually works, it could save them a fortune in hospital claims that they would otherwise be on the hook for. Seems like they could create a foundation, hire experienced people from pharma companies who have shepherded drugs through the process, and get to work.
I’m sure there’d be some pushback since everyone seems to hate on insurance companies but it’s hard for me to see how they would be any more conflicted than the pharma companies who make and profit off the drugs directly. I mean, at least they are on the hook for claims if there are side effects or situations where it doesn’t work. Unlike the drug companies, they only profit if it works and saves them claims down the line.
Why isn’t this happening?
Vox is one of the shittiest sources you could possibly listen to. They're just awful. I wouldn't trust Kelsey Piper further than I could chuck her.
Your anecdote about ketamine is the reason I've never pursued a medical degree. I can give people who have treatment-resistant depression ketamine and shrooms. I can give MDMA to people with PTSD. I have done so many times with great success and it's a wonderful feeling. If I were an accredited doctor then doing all that stuff would cause me to lose my license at best.
How much of this is that the doctors know Paxlovid is right around the corner?
If it's any consolation, the engineers' budget ventilator approach probably wouldn't have been very helpful; in the early days especially, we were short on fancy vents, the kind that could do advanced modes of ventilation to treat ARDS. Being on a vent is generally bad for COVID, but if you're going to be on the vent you should ideally be on the kind of computerized ventilation that can micromanage pressures, not the kind of simple machine that runs on gas pressure. Effective mechanical ventilation is an esoteric art that even many MDs don't understand, and there was some serious miscommunication going on in the early days of COVID as to what was needed. In purely consequentialist terms, I doubt you did any significant harm.
(I'm a respiratory therapist, though I was still in school when COVID hit)
Social acceptability is a helluva drug. Even compared to ketamine.
Here is the thing, and my understanding of this is quite imperfect. There are multiple purposes the FDA serves. One purpose, which is pretty important and actually takes a lot of work but is viewed as boring and taken for granted by many, is to make sure the actual manufacturing of drugs is done correctly. That every pill of your generic aspirin constains exactly the same dose as the label says even though dozens of generic companies are making the low margin drug trying to keep costs as low as they can.
For a new drug, something doctors have no experiences with, the fear level can be high and the FDA I think does serve as providing a blessing. A doctor can say "I don't know anything about this drug for shingles but the FDA says this drug works for patients in this group so I'm not just taking the word of the 20-something drug rep". Since drug companies are only allowed to market drugs they have approval from the FDA for and they can only market for indications that are on the FDA approved label, there's an obvious profit incentive for them to spend the money to go out and do the studies to establish the bona fides of the new drug or to prove a drug that is already approved for one thing also works for something else.
But older drugs are a bit different. Doctors presumably have real life experience with patients on older drugs so they should feel more open to trying them if there's reason to believe they may work in an unexpected place. If some study said low dose aspirin helped certain types of depression, you should not feel a huge amount of fear trying it on your patients 'just to see'. Yes if some generic aspirin manufacturer spent $75M to run a clinical trial and submit it to the FDA for a label expansion you'd feel better, but is that the best use of resources? I'm not sure.
So I'm not sold on the idea that the FDA should go from passive entity here to active. By passive I mean companies come to the FDA asking if they can market a new drug or add something to the label of their existing drug. The FDA convenes the advisory boards with experts on the disease area and evaluates their evidence but they do the work of collecting the evidence, hoping for the best.
But should the FDA go active? Going out and trying to find new labels for old drugs? Picking up studies others did to try to add to the label rather than waiting for people to ask them to do it? The FDA does have limited resources and it isn't like you can just throw money to quickly expand it to infinity. It's not like disease experts grow on trees after all.
More importantly if the FDA did start doing this, I would fear that it could lead to the less formal knowledge generating institutions of the medical field starting to atrophy. Medicine form my impression is a combination of theretical and practical knowledge. Doctors combine a lot of information from less formal sources. These include the experiences of other doctors, articles in the literature (not just highly controlled RCT's but case reports, the chatter at conventions and meetings, etc. I suspect making the FDA pro-active in going out to find label expansions for existing drugs will cause doctors to pay less attention to these others areas of knowledge generation and they will atrophy.
Now that being said, doctors have seemed to be totally willing to write a lot of scripts to try to fight Covid. How many ivermectin and HCQ scripts were written? OK the FDA did give HCQ temporary approval but still, those scripts are still being written. I recall when the first cases happened news reports said doctors were trying tamiflu. Yet tamiflu isn't approved for Covid and even for the flu it only does good if you start it early but they were clearly trying anything that seemed like it may have a plausible path to working.
I think the systemic issue goes beyond the FDA. There are many other institutions that influence what doctors feel comfortable prescribing, and in principle that makes sense: individual doctors can't conduct large-scale medical research on their own, so they really need to be able to trust what's coming out of institutions. As well as the FDA, we should look at:
1. Popular medical resources like Uptodate, which provide information on common treatments including off-label prescribing. Lots of doctors consult this many times a day. Its article on fluvoxamine begins with a warning about Covid treatment that says there is "insufficient data to recommend for or against." I think persuading the editors of Uptodate to actually update their article in light of the TOGETHER trial could make a big difference.
2. Insurance companies, which put barriers in the way of unusual (or expensive) treatment. Refusing pointblank to cover something can get them into trouble so they're usually more devious, requiring a doctor to fill out a long form called a "prior authorization" that details all the other treatments a patient has already tried. The point of these forms seems to be to deter patients from filling their prescriptions, as getting the form and then getting the doctor to complete it can take a lot of time and effort. Patients usually don't know about this requirement until they get to the pharmacy (and doctors can't keep track of the constantly changing requirements of all the different insurance companies).
I'm not aware of any insurance companies requiring this for fluvoxamine, but the ~$10 cost Scott mentions is only at specific pharmacies using a coupon from somewhere like Goodrx. Thanks to our ridiculous drug pricing system, the walk-up cost is closer to $100.
3. State regulators, which have threatened to take away the medical licenses of doctors who write inappropriate mask or vaccine exemptions. I think the state boards are probably right to do this on balance, but it still makes doctors very reluctant to deviate from the consensus on Covid.
4. Large hospital systems and medical groups, which heavily influence the prescribing decisions of the doctors who work for them. The people in charge of these are usually far removed from clinical practice.
5. Supply chain issues, which resulted in shortages of HCQ last year. Do pharmacies have enough fluvoxamine to give it to everyone with Covid? There is a very obvious long-term solution if Covid is going to be endemic, but it could be a problem in the current surge.
Talking of early treatments, doctors going off protocol but researching themselves and treatments that don't get the funding to gather the large studies that are usually required in medicine:
Have anyone here already heard of Dr. Chetty (South Africa) and his findings? For me as a rational thinker without medical background this sounds really plausible and the way practical medicine should work.
https://youtu.be/2T7KlrzPAYY
Is there any obvious flaw i didn't get? I'm only speaking of medical or logical reasons. If he perhaps gave Interviews to the wrong people doesn't change the facts he has seen or done on his patients.
Just as an aside, I hate that that CS Lewis link forces javascript in order to read a piece of text.
To be a little fairer to the FDA, they do seem to be aware of the drug trials but don't think there's enough evidence that fluvoxamine does make enough of a difference (EDIT not the FDA, it's the NIH):
https://www.covid19treatmentguidelines.nih.gov/therapies/immunomodulators/fluvoxamine/
"Anti-Inflammatory Effect of Fluvoxamine and Rationale for Use in COVID-19
In a murine sepsis model, fluvoxamine was found to bind to the sigma-1 receptor on immune cells, resulting in reduced production of inflammatory cytokines.1 In an in vitro study of human endothelial cells and macrophages, fluvoxamine reduced the expression of inflammatory genes.2 Ongoing studies are establishing whether the anti-inflammatory effects of fluvoxamine observed in nonclinical studies also occur in humans and are clinically relevant in the setting of COVID-19.
Recommendation
There is insufficient evidence for the COVID-19 Treatment Guidelines Panel (the Panel) to recommend either for or against the use of fluvoxamine for the treatment of COVID-19.
Rationale
Three randomized trials have studied the use of fluvoxamine for the treatment of nonhospitalized patients with COVID-19. In STOP COVID, a contactless, double-blind randomized placebo-controlled trial conducted in the United States among nonhospitalized adults with mild COVID-19 diagnosed within 7 days of symptom onset, fluvoxamine (100 mg up to 3 times daily for 15 days) reduced clinical deterioration at Day 15.3 Clinical deterioration was defined as shortness of breath plus oxygen saturation (SpO2) <92% or hospitalization plus SpO2 <92%. This was a small study (≤80 participants per arm) with limited cases of clinical deterioration and a short follow-up period. In addition, 24% of participants stopped responding to surveys prior to Day 15.
The subsequent STOP COVID 2, a Phase 3 randomized controlled trial (ClinicalTrials.gov Identifier NCT04668950) that enrolled >700 participants in the United States and Canada, was stopped for futility by a data safety monitoring board after lower than expected case rates and treatment effect were observed.4
TOGETHER is an adaptive platform, double-blind randomized placebo-controlled trial conducted in Brazil.5 Nonhospitalized adults with COVID-19 and a known risk factor for progression to severe disease were randomized to fluvoxamine 100 mg twice daily (n = 741) or placebo (n = 756) for 10 days. Fluvoxamine use was associated with a lower risk of the primary composite outcome of retention in the emergency department for >6 hours or admission to a tertiary hospital (79 of 741 participants [11%] in the fluvoxamine arm vs. 119 of 756 participants [16%] in the placebo arm [relative risk 0.68; 95% CrI, 0.52–0.88]). Of note, 87% of the primary outcome events were hospitalizations. There was no statistically significant difference between study arms for the secondary outcomes of need for hospitalization or time to symptom resolution. There was no significant difference in mortality between study arms in the intention-to-treat (ITT) population (17 of 741 participants [2%] in the fluvoxamine arm vs. 25 of 756 participants [3%] in the placebo arm [OR 0.69; 95% CI, 0.36–1.27]). In a secondary, per-protocol analysis of participants who received >80% of possible doses, death was the outcome for 1 of 548 participants (<1%) in the fluvoxamine arm versus 12 of 618 participants (2%) in the placebo arm (OR 0.09; 95% CI, 0.01–0.47). Participants in the fluvoxamine arm were less likely to present to an emergency setting for COVID-19 for any duration, although this analysis was not prespecified. Compared with those in the placebo arm, participants who received fluvoxamine were less adherent to therapy and discontinued therapy due to intolerance more often.
While fluvoxamine treatment significantly reduced the primary composite outcome in the TOGETHER trial (i.e., retention in the emergency department for >6 hours or admission to a tertiary hospital), the difference in hospitalizations between arms was not significant.5 Defining the clinical relevance of the >6 hour emergency department observation time endpoint is difficult, especially its applicability to practice settings in different countries. Moreover, the endpoint has not been used in other studies of interventions for nonhospitalized patients at high risk for hospitalization and death. While a per-protocol analysis found a significant treatment effect for mortality in patients taking >80% of possible doses (assessed by patient self-report), no such benefit was found in the primary ITT analysis. The 80% threshold has no clear justification, and only 74% of participants in the fluvoxamine arm reached this level of adherence. Since per-protocol analyses are not randomized comparisons, they can introduce bias when adherence is associated with factors that influence the outcome; this bias cannot be excluded in this study. Notably, mortality in the placebo arm was substantially higher in those with ≤80% adherence than in those with >80% adherence, suggesting that factors other than adherence differed in the per-protocol population. Finally, including only participants who could tolerate fluvoxamine does not reflect the actual effectiveness of the drug, since intolerance and adherence appeared to be related.
Additional studies are needed to provide more specific, evidence-based guidance on the role of fluvoxamine for the treatment of COVID-19. Further details of the studies discussed are provided in Table 4c."
That sounds less like "we can't find anyone willing to be responsible for this" and more "if some study comes out with yes yes yes it works!!!, then we'll approve it". Right now, they seem to think the studies are "meh, it's kinda okay".
Are there good studies on post-SSRI sexual dysfunction (including less visible symptoms like libido loss in women)? Wondering about the basis for Scott's characterization of the side effect as "very rare." I have some medical contacts who focus on those issues and believe that they're under-scrutinized and under-researched
Scott, I like to print out your essays to read with a cup of coffee as a break from my computer. Your writing is so good that I always enjoy them hugely, whether I'm interested in the subject matter or not.
But when I print them out from my web browser, the last paragraph of the first page is always obscured by a big 'Publish on Substack' advert, which only appears in the printed version, I don't see any such thing on the page in the web browser at all, it's only on the printout. And I can't find any way to work around this.
I realise that this is nothing to do with you, and indeed probably a bug in the Linux version of Firefox, but I wondered if you had the clout with Substack to get them to fix their button so it doesn't do this?
Typo: But But CS Lewis
Let's say, purely hypothetically, that you have two octogenarian grandparents whom you love dearly. They're both vaccinated and boostered. One has recently finished bladder cancer treatment, another has a history of heart rate issues, beyond that and their age they have no major COVID comorbidities. What steps would or could you, their grandchild, take today to improve the odds that they can obtain a Fluxov prescription in the event that either one contracts COVID?
So, speaking practically rather than legally, is there a way one can get one's hands on fluvoxamine?
Just to point the 30% number in perspective, this is roughly equivalent to the estimated effect size for going from 0% masked -> 100% surgical masked from the Bangladesh study.
Unfortunately fluvoxamine only impacts your risk from Covid and has no real impact on psychological safety so it is utterly irrelevant to our societal response.
>The VA system took 35,000 high-risk older patients off of an unusually-likely-to-cause-QT-syndrome SSRI in 2011, and were unable to find any evidence that this prevented even a single case of the syndrome, let alone any negative outcome!
Is it just me or is this phrasing quite like the "no evidence" phrasing from https://astralcodexten.substack.com/p/the-phrase-no-evidence-is-a-red-flag?
4. If people believe there is a safe and relatively effective treatment for COVID, they might be less likely to bother with a weird GMO vaccine.
Is there a role for the physician organizations? Because I know for our kid, the recommendations about things like car seats and avoiding SIDS and stuff generally come from the American Association of Pediatrics, not the FDA but nor do individual pediatricians have to decide what to say about a lot of issues. And when my wife was pregnant we kept hearing about recommendations from the American College of Obstetricians and Gynecologists. So could one of the professional medical associations give physicians the okay to do this and know they aren’t being unprofessional?
"That is, FDA procedures usually assume there is a pharma company sponsoring a drug. But fluvoxamine is cheap and off-patent and no pharma company is involved in repurposing it for COVID. Nobody has a procedure for a drug without a sponsor, so they won’t do anything."
I'm mildly amused by the happenstance that this is pretty much the same thing Weinstein et al. were saying about getting Ivermectin re-labeled as a prophylactic for COVID.
No reflection on the virtues of fluxovamine, since I am not a doctor of any sort.
I find it very frustrating that a lot of the NPIs are justified in order to keep hospitals from being overwhelmed, and here is a treatment which would increase hospital capacity by 30% (or is it 43%?), which would probably have done more than any NPI to keep hospitals from being overwhelmed, but we aren't using it because there is no one to pay for marketing it.
So these doctors and nurses we see in the media who complain about how overloaded they are, they don't know about it?
I love this post!
I've been single for too long because asking people out would require me to do something potentially awkward.
"For some reason the same experts who don’t mind prescribing SSRIs when people have mild depression freak out about prescribing them when they’re the only evidence-based oral medication for a deadly global pandemic. "
While I agree wholeheartedly with the sentiment, correction: dexamethasone is oral.
While we're on the subjects of steroids and "flu-" drugs that 10 million Americans already use without issues, why not fluticasone inhalers to treat Covid? They're extremely safe (far fewer side effects than oral steroids) and for me they clear up all my respiratory inflammation completely within 12-48 hours depending on how bad it was before I started. A hypothetical downside is that it might stop working in a part of the lung that is filled with fluid, but as a prophylactic against pneumonia as soon as you test positive for covid it would work great (80% confidence). Maybe I'm just seeing it through rose-tinted glasses because i'm a super-responder to it. I have occasional inflammation in my lungs, not very bad. It usually stays under control with far less than the recommended 2 puffs of 110mcg twice a day. One puff once on days when I feel like I need it is usually sufficient. When I had covid in May 2020, I raised my dosage back up to what was prescribed (2 puffs twice a day) and got through it without letting my SPO2 dip below 95%. It lasted 10 days and was milder than some flus. Fluticasone inhaler was the only thing I was taking.
There's a preregistered study in progress, still recruiting: https://clinicaltrials.gov/ct2/show/NCT05054322
10% of Americans being on SSRIs is my personal “I don’t know anybody who voted for Nixon” moment.
I don’t even know anyone who sees a therapist except for the people in our drug treatment court which I work with.
Are people’s primary doctors handing these out?
Dr. Faust is a name I can believe in.
12/23/21 Update:
"Ontario becomes the first province to list fluvoxamine as a COVID-19 treatment to consider"
https://www.ctvnews.ca/health/coronavirus/ontario-becomes-the-first-province-to-list-fluvoxamine-as-a-covid-19-treatment-to-consider-1.5717489
Scott, you shouldn't beat yourself up for not doing something which seems unlikely in retrospect to have mattered, or for hesitating to take steps which were probably legal until you had more Bayesian evidence thereof.
You should beat yourself up for not trying to convince the FDA to authorize vaccines for new variants under the precautionary principle, which seems like an iron-clad argument according to premises they might accept.
Also, stop using a technique of anti-epistemology which successfully stitches together anthropomorphism with dehumanization. FDA bureaucrats drag their feet on needed medicine because they don't understand the difference between their behavior and the archetypal young student standing up to the anti-controls expert in front of a crowd. It's almost like traditional rationality places too much emphasis on avoiding false claims, or being less wrong.
Which of the three is your reason to not prescribe [Desoxyn](https://astralcodexten.substack.com/p/know-your-amphetamines)?
About that Bayes formula on the bottom in your subscribe thing... Bayes formula doesn't tend to give 60/40 odds.
Here's how it works: "prior probability", P(A) that a random drug is effective for COVID, is very very small, that should be understandable, right? P(A|B) is the posterior probability given the trial(B) .
The P(B) is probability of finding the drug to be effective whether it works or not. Which for this study is woefully high - there's a number of red flags (the stranger the end point, the more effective the drug was at it; non strange end points not even statistically significant). Note that P(B|A) is less than 1 (but presumably, close to 1), and to get a greater P(A|B) than P(A) you need P(B) to be small.
The end result is that your P(A) is small (fractions of percent) and your P(A|B) is higher, but still very small (fractions of percent).
To get odds close to 50/50 (like your 60/40), would require a rare numerical coincidence; the P(B) needs to be approximately equal to P(A) . That doesn't occur often, because people who could get P(B) to be this low, shoot to make it even lower.
Read RFK Jr.’s book and all becomes painfully obvious.
If 3 is an issue, isn't the solution writing up some social permissions slips and handing them out to as many doctors as you can get them to?
It sounds like all you need to do is let doctors have the line "I know a bunch of other doctors that are prescribing patients with fluvoxamine for covid" and they'll start doing it themselves too.