Bullshit Job Notes

Scott talks about a bit of medical theater, in which companies demand doctors notes to allow people to bring in their own chairs to work; he is often the person in best position to provide that note, despite being a psychiatrist who usually works remotely. He’s knows little about back pain and nothing about chairs. His compromise has been to write out a note saying “[Patient] reports back pain that would be improved by a chair they have chosen.” This shouldn’t work because he’s not adding any additional information to the patient’s complaint, but it does.

I had my own experience with this in my first real-world programming jobs. The overhead lights in my office gave me a headache, so I turned them off. Someone noticed and didn’t think it was fair my roommate be forced to work in the dark (although he told me it was fine). I indeed had to go to the doctor, who dutifully wrote out “Elizabeth says the lights give her headaches…”. Luckily I had fabulous health care insurance and flexible hours at my job, so this was a minor annoyance.

I really disliked the way the company handled it though. Their solution was a lamp, which was fine, but there was a lot of pressure to say either “yes, the problem is entirely solved” or “no, it is definitely not solved.” “Can I take a week to see if I develop a headache?” was not an option. I don’t think I even got to pick out my own lamp.

Several years later I had a similar problem at a different large tech company you have definitely heard of, one known for being an amazing place to work. I hated working in an open office and tried to leverage my post-dental-surgery fragility into a private office or the right to work from home. In retrospect, this was not a reasonable accommodation: despite having offices in 50 countries and running much of the world’s telework infrastructure, that company’s workflow was set up for in person communication. But they couldn’t just say that, so they tried to find ways to meet the letter of my doctor’s note without actually giving anything up. This was a problem because what I really wanted was control over my environment, and that was the exact thing they didn’t want to give me.

I think a thing that’s going on with the notes is that gatekeeping makes sense in some circumstances (substantial changes in job duties), and it was no skin off the decision-maker’s nose to make that a universal rule. “Protection from lawsuits” factors in, but if companies were paying the cost of the note they would draw the line in a different place. As it stands, it’s worth infinite amounts of your time and money to avoid risk of a lawsuit that might cost them either.

Today in Horrifying Systems

Recently I listened to Econtalk’s episode on the doctor-patient relationship, which was primarily about a study showing that hospitalized patients did better when their care was supervised by their primary care physician. The study shows what it shows (although pilot studies are prone to not replicating), but implicit in the discussion was the idea that continuity of care was important because doctors remembering things is better than writing them down. This may be true, and the podcast in fact cover some of the problems with medical records*, but if so that is a profound failure of medical record keeping. Human memory is the last thing you want to trust your life to. I pay extra for the kind of doctors that spend actual time talking to me, and I still have to remind the one that that drug she loves has unacceptable side effects for me, every time I see her. I in no way trust her to remember something as dense and detailed as test results. That’s why she writes them down.

My understanding is that many people do understand this, but fixing medical record keeping involves a huge government entanglement that stifles all progress. So we’re stuck with human memory as the thing keeping us alive.

Note that there are good reasons to have your PCP supervising your hospital stay even if everything is properly written down, which is hospitals are full of specialists who aren’t even trying to coordinate care. You benefit immensely from someone thinking “how are these things going to interact?” In theory this should be done by a hospitalist, but they only have so much time and not doing it is much faster. But let’s not pretend we’ve proved human memory is in any way a good solution to any problem ever.

*EHRs simultaneously have too much information, taking a long time to complete and read, and don’t have space for information off the beaten path.

The Longevity Research Institute

A theme on this blog is “there isn’t enough good information.” Anti-aging research is worse than most topics at this, perhaps even worse than nutrition. There are several reasons for this: time/money requirements of research, the FDA’s focus on pathology rather than health, and science as a whole disincentives the kind of replication you need to truly trust a result. But there’s a new organization, the Longevity Research Institute, aiming to fill this gap. (Full disclosure: I’m on the board).

To run a longevity experiment, you have three choices: research something with a short life cycle (limited applicability to humans), research something with a long life cycle but track biomarkers instead of actual lifespan (risks goodharting the biomarker while not affecting or even hurting lifespan), or wait. Even a mouse experiment can take four years for your subjects to die. And the more successful your treatment, the longer (and thus more expensive) the experiment.

This would make lifespan-increasing treatments more expensive, but that’s hardly the end of the world. What could be more worth paying for than life? Unfortunately, the FDA takes another view. To get FDA approval, a drug must show effectiveness treating a specific disease, and aging is not considered a disease. This has been challenged recently, with the TAME trial, but has been a fact of life for a very long time and there’s no guarantee TAME will win the fight.

Finally, science in general is bad at replication. Repeated studies are necessary to have any confidence in a treatment, but glory goes to the original discoverers of a treatment and their funders, not to replicator three of ten. So we end up with a list of promising treatments that no one can trust because no one has any incentive to test them again.

All together, this leads to a lot of low-hanging fruit in longevity research; there’s a ton of promising molecules out there that only need money to be tested. The Longevity Research Institute aims to pick that fruit. It has a very targeted mission: take compounds already shown promising in at least one trial and fund replication trials where you actually wait to see how long the mouse lives. That’s it. Yes, it should worry you that this isn’t already being done.

The LRI is run by Sarah Constantin, who writes the blog this blog wants to be when it grows up. You can see her past work in e.g. treatment-resistant depression, STD transmission, and chronic fatigue syndrome. Because there are very few fixed costs (Sarah’s salary, some legal consulting) and studies are so expensive, there’s lots of room for more funding. You can see the LRI’s roadmap here and an initial list of promising compounds here. LRI’s first trial, of the synthetic pineal peptide epitalon, launched just last month.

For more information, you can follow the blog or Facebook group. If you feel so moved, you can donate to LRI here. I’m happy to talk to anyone considering donating (elizabeth-at-this-domain-name), and for larger amounts Sarah herself is available.

In Robert Kegan’s In Over Our Heads, he says “it is a poor school whose favorite students are the ones it does not have to teach.” (page 171). I get what he’s going for, but having lived it, I can also say it’s a poor school whose least favorite students are the ones it does not have to teach. It tends to lead to things like resentfully insisting you do have something to teach the student and forcing them to go through the motions of learning it, or standing in their way when they try to do something they’ll actually learn from. Or an intrusive focus on their personal life.

I feel like the whole concept of favorites is leading people astray here, which is a very Kegan thing to say.

Self Help Epistemic Spot Check Results

In a word: ¯\_(ツ)_/¯

I reviewed several self-help books with a wide range of scientific backing. For posterity:

Polyvagal Theory
The Tapping Solution/EFT
Full Catastrophe Living/Mindfulness
Exercise for Mood and Anxiety
A Guide to Better Movement
There were a few others I never published because I didn’t get very far into.

As a reminder, epistemic spot checks are checking a book’s early claims for truth/scientific validity/coherent modeling, to determine whether it’s worth continuing. After a few books I concluded that scientific backing didn’t seem that predictive of a book’s helpfulness, and started focusing on modeling. But that wasn’t predictive either.

I never officially decided to quit this project, but I can no longer get excited about checking out a new book, because nothing short of trying it seems to have any predictive ability of whether or not it is helpful. This leads me to believe that most of the effects are placebo effect, not in the sense of  “imagined” as people usually use the word, but in the sense that it’s your own brain doing most of the work, and people just have to try things until something clicks for them, starting with the cheapest. I find this answer deeply unsatisfying, but what are you gonna do?


Review: SOMA

I said I wanted to get this done while the game was still in a Humble Bundle, and technically I succeeded, but it turns out it doesn’t matter because I’m not going to recommend it.

SOMA was sold as “Like Amnesia but sci-fi and underwater” and I love all of those things, so I had very high hopes. Hopes that were dashed. I don’t know if my tastes have changed since I played Amnesia or SOMA just wasn’t as good, but I stopped playing two hours in.

If we’re assuming the former, the problem is I’ve played a lot of puzzle games since Amnesia and my standards have gone up. Both Amnesia and Portal have puzzles so you have something to do while running from monsters and…eh. I thought I was cheating myself by looking up solutions so often, but then I solved a few myself and there was still no sense of satisfaction. It’s more just testing what the environment will let you do until the problem somehow goes away.

I was intrigued enough by the story to look up the ending. If you’re looking for something with similar themes, less terror, and better puzzles, I strongly recommend The Swapper, which remains one of my favorite games of all time.


In summary: Manganese is an essential nutrient and too much will kill you.

On one hand we have people exposed to manganese is the air from industrial processes. These people definitely get Parkinson’s like symptoms, and in the aggregate score lower on various intellectual tasks. Monkey infants on soy formula (which has more manganese than milk based formulas) had mild neurological effects, so you can take in enough from food alone to be dangerous.

On the other hand, manganese superoxide dismutase is the only thing keeping your mitochondria from melting, and it does several other things besides. So we can’t just throw up our hands and say “no more manganese”, we have to actually figure out a safe level.

The World Health Organization has set the No Observed Adverse Effects Level at 11mg/day, based on the observation that some humans naturally get that much through their diet. As mentioned previously, we know you can get manganese poisoning through food alone, so this is not very reassuring to me.

The most comprehensive source I found on this was this thread on the soylent message boards (they cite several of the same things I did above, all correctly, so I trust their work ). According to their sources, most adults already eat enough Manganese to meet the RDA. I don’t want to put too much stock in the RDA, because neither I nor they could find out how the FDA (or the Linus Pauling Institute) set it. However even if you accept the number, if people are already getting it through diet, and too much is toxic, it doesn’t belong in a multivitamin.

This is worrisome enough to me that I’m tossing my multivitamin and focusing on getting individual missing nutrients through specific supplements (if anyone knows of a balanced B supplement, please let me know). It’s not so bad I’m tossing my protein/vitamin powder, because it has a lower volume of manganese and I don’t eat much at a time- however when it’s gone, I’ll probably switch to a straight protein powder.