Hacker News new | past | comments | ask | show | jobs | submit login
[flagged] How medicine discriminates against non-white people and women (economist.com)
16 points by colinprince on April 15, 2021 | hide | past | favorite | 16 comments



I was reading that testing often happens on men for medications because their biochemistry is fairly constant in a month. Fluctuating hormones complicated drug testing.

This seemed to me to miss testing a scenario that affects about half the population.


It's not a single scenario; that seems to minimise the complexity.


There are many scenarios because there are variations in our genetics.

I was reading (I think it was the book The Innovator's Prescription) that a lot of medical R&D happens in the US. I wonder how that impacts this sort of testing. I bet that medical R&D happening in a more distributed nature around the world would engage more diversity for testing.


> I bet that medical R&D happening in a more distributed nature around the world would engage more diversity for testing.

Sure. Distributing the research and testing internationally would be the best case. But if you're only testing in one single place then not many locations are as ethnically diverse as US metropolitan areas.


I mean this as a sincere question and I get that their might not be an answer readily available but I've always wondered:

How onerous is it really? What an I missing here?

Not that pfizer is on trial here (maybe they do a good job, I don't know) but they have presence in diverse metropolitan cities[1]? I'm sure many others must too.

There must be something more to this.

[1] : https://www.pfizer.com/science/research-development/centers


A dermatologist friend of mine trained in a virtually all-white area and then got a job in a place that wasn't. Apparently it was a significant adjustment caring for patients with darker skin.


There's also a false belief, common among doctors and other medical professionals, that darker skin is thicker/tougher. (to be 100% clear, it only differs from lighter skin in pigment and no other characteristics).


Your comment and the one you replied to don't make sense to me. How can skin only differ in pigment, but it also require additional training to care for skin of different color?


Because it is a different colour, meaning that all the training you received learning focussed on different shades of pink no longer applies. Darker skin means the symptoms you are looking for are lower contrast and a different colour. It’s a simple case of repeating all that training against different skin colours.

It’s often difficult for someone trained exclusively with pale skin to recognise something as obvious as sunburn on darker skin, for example.


I'm guessing that identification of different skin conditions on skin of much higher pigmentation that what a dermatologist has been trained on is an obstacle.



[flagged]


In the very first paragraph of that story it reported on the failure of a common and important diagnostic tool to function properly on people with darker skin.

The Economist is catching up


Is that why The Economist has a whole separate category for reporting on "Race and sex bias in medicine"?


What is your complaint?

Do you feel that medical science is pure and unbiased and this category of reporting is not required?


This comment seems to show a pretty fundamental misunderstand of what CRT is (or isn't), or this article, or both.

The article is a bit fluffy, but some of the problems they are describing are well known in the industry (less agreement on what to do about it) and have been showing up in the medical literature for decades at this point.


[flagged]


> everything in medical literature is fact and in no way fiscally motivated

It's hard to imagine a net financial benefit for many of the studies.

Of course it's not perfect. But dismissing something like this out of had for what, (vague ideological reasons?) is irrational. There is good research in the area, and you are free to read it. There is some poor research in the area too, such is life.

It's not even suggesting anything malicious, just that in many cases if you follow the easiest path in design and testing, you will end up biasing your inputs in ways that have impact. I have seen this play out.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: