I had a fun conversation with Sarah Constantine at Manifest where I sought her advice on writing a technical blog. Because my areas of interest are deeply practical: pharmaceuticals and practical finance (specifically, stock picking and making decisions about how to express concrete beliefs about the future in ways that seem likely to be profitable.)
Mostly I asked her about the duty of care that a writer has for their audience. American culture has pretty negative views on popular writing about finance, health, or legal stuff: that they are topics best left to the credentialed experts, and there’s a (totally reasonable) horror at the idea that if you write about this topic then you might cause someone else to do something they regret forever, and it will be your fault.
I have mixed feelings about this taboo. On the one hand, yes. These three fields have the commonality where if you fuck up you can cause yourself extreme damage, either in terms of your bodily health or your finances or even your ability to exist as a free man not in prison. It’s probably reasonable to say that “you are required to have a certain bar of expertise before you dispense advice about this.”
On the other… these taboos result in an information environment that is utterly abysmal. If I want to know, for instance, about semaglutide in its oral form (aka Rybelsus), I have a few different information sources available to me and they all kinda suck in different ways:
Alas, WebMD
The “blessed” source for laymen’s information on such topics. It looks like this:
As a consumer who is considering taking semaglutide, I want to know a few things! Like: how does the oral version compare to the injectable version? Is one more effective? Should I take this or should I try out liraglutide instead? What alternatives exist?
WebMD is not interested in any of those questions. This website operates off the hardline stance that the doctor is making medical decisions on the patient’s behalf and that WebMD’s role is to ensure that the patient is obeying the doctor. That’s why the Side Effects section of oral semaglutide looks like this, with actual information interspersed with throat clearing about how your doctor probably had good reasons for prescribing this:
Sidenote: glaring omission in “side effects” section for WebMD’s article on Rybelsus
I am actually super annoyed here because I can’t help but notice WebMD neglects to mention anything in its writeup about the fact that Rybelsus causes massive, clinically significant weight loss in people who take it. I understand that “weight loss” isn’t an FDA-indicated use of oral semaglutide, which is why they don’t have it listed in the “main effects” writeup (remember, the FDA-approved use of oral semaglutide is glycemic control and not weight loss.)
But how did “significant weight loss“ avoid making it into the side-effects section? This is observed in literally every study ever performed on this substance! A lot of people actually cannot afford to lose weight; this is not universally a beneficial effect! The lowest dose of Rybelsus causes patients to lose 10 pounds on average! This is not a subtle effect.
My suspicion is that WebMD is just very afraid of being seen to endorse an off-label use of any medication, and the problem with “this may cause you to lose weight” is that the statement can readily be interpreted by readers as “this will cause you to lose weight” or even “you should take this to lose weight.” So, not knowing how to thread this needle, they instead decide to omit the most significant side effect of the medication from their “side effects” section and I guess the hope is that patients will discover this knowledge elsewhere?
Google Scholar
If I search up “semaglutide” on google scholar I get this:
So naturally this is much much more comprehensive, but it requires an exceptionally high bar of education to be able to make sense of journal articles. Plus, the information you want to find— “what can I expect to happen as a result of taking this medication“— is typically buried in the Results section. If you’re coming at this from the direction of wanting to achieve a particular goal— “lose weight”, for instance— then you’ll need to do a lot of digging if there’s not a review article specifically written for that.
It is also, unfortunately, all written in Scientific English, which is a lot like Regular English except it’s much harder for normal people to parse.
If you’re looking for the answer to a specific, simple question— “what can I expect to happen to me as a result of taking this medication, and how do we know that”— this is pretty much where you have to go. Then, if you’re like me, you DIY the world’s most slipshod review article in Notepad++ so you have some kind of record of what you found.
The advantage of Google Scholar, of course, is that it’s the ultimate repository of knowledge on all pharmaceuticals, full-stop. Whatever WebMD or whatever is telling you is coming from those same scientific journals. Even though WebMD doesn’t cite their sources.
Terrible Internet Listicles That SEO Their Way To The Top Of Google Search Results
Now, it’s much more common, I think, to start with a goal rather than being interested in a specific medication. “Weight loss” is a common one. Then you search google for that. And typically, you get some listicles. I’m not linking them here because I don’t want them to get the traffic.
They just say… stuff, I guess?
I’m not sure any of that was true, in the sense of “validated by a study of reasonable quality.”
Reddit
Reddit is a mixed bag. I appreciate that fundamentally Reddit does not pretend to be anything other than what it is: a scattershot collection of posts by online randos on any topic you care to name. Accordingly, posts range from “actively lying to you“ to “the literally-best source of information on X topic that exists online.”
I have an ex who, after consulting numerous specialists and general practitioners, found a lasting solution to [complex and painful medical issue] from a Reddit comment.
The Desired But Nonexistent Alternative: CrazyMeds But For [Topic]
There’s not, to my knowledge, a place online that serves the purpose of trying to consolidate all the known information about different goals one might attempt to pursue with pharmaceuticals.
Scott Alexander, a decade ago, wrote about a site called CrazyMeds (note: the CrazyMeds domain itself got bought out by some random internet scammer. Don’t go there.) The core idea of the site was that it just did a best-effort, well-cited review of “what can I expect of this medication and how does it compare to other medications I might take for the same purpose” for the wide world of psychiatric medications.
Why doesn’t that exist, but for non-psychiatric drugs? Come to think, why doesn’t CrazyMeds still exist? It was a good site, useful for many people!
I wrote an article several years ago that performed a comparison of several different weight loss drugs, in terms of side effect profiles and effects achieved. That was very satisfying, but it was also kind of surreal writing it because at the time I wasn’t aware of anywhere else online that attempted to do anything similar even though it seems so so obviously like a resource people need. So now that I have a lot more free time, I wonder what other things I can do in that vein.
In a bit I intend on putting an updated version of that article on this Substack. I don’t have any specific medical expertise, but frankly in that article I’m not doing anything other than “consolidating peer-reviewed studies on the drugs under discussion and pointing out different aspects of said articles”, which isn’t exactly rocket science; I figure that if I just make my layman status clear and do my best to do a balanced, well-cited review of the existing literature, I’ll have performed my duty of care.
Coming Soon Or Just When I Feel Like It
I’ve got a few articles in that vein I’d like to write:
As discussed, a revised version of my article where I give a literature review of weight loss drugs, with “measured quantity of weight lost” alongside “side effect profile” and “other important notes.”
A deep dive into the side effect profiles of the known GLP-1 agonists, especially semaglutide and tirzepatide, but (depending on how much good info is out there) also on cagrilintide and the other up-and-comers.
In particular, I’m interested in their psychiatric effects and apparent anti-addiction properties.
Some kind of treatment for the unknown-unknowns problem of being on a drug for an extended period of time. Is there a good way of thinking about this broadly?
A comparative review of dietary approaches to weight loss (warning: this one’s gonna be depressing).
A review of all the known common effects of testosterone supplementation.
There’s a lot!
great post! look forward to hearing more