Saturday, July 22, 2017

Statistical wellness: the current face of eugenics

Eugenics are back! And in some cases being enforced by governments!

Ok, so that bears some explanation. But I’m going to have to start with some basics.

The first thing to understand is that we don’t know how our bodies work. We see what they do and have some really good ideas about what’s going on, but we’re not really sure. This is why drugs still have to go through trials cause it’s all guessing. Where we differ from the age of leeches, is statistics. We started using data across populations to figure out what worked and what didn’t and we started to correlate different metrics for quantifying a person. Doing this made a leaps and bounds improvement to medicine.The problem is that the low hanging fruit of obvious and useful correlations have all been  identified. Now all the correlations are part of more complicated systems so they are much harder to figure out. The other thing to understand is that most of the work has been targeted and dealing with illness. All the metrics we use are targeted for use when the patient has a problem, as part of a process to guess what the problem is. It’s not a process that thinks it’s definitely going to get it right on the first try, so it isn’t optimized for that. It’s optimized for the statistically best outcome.

The second thing to understand is that we’re not all the same. A person whose growth has been stunted might be taller than a different person in perfect health. We vary. There are members of our population that stand out. Sometimes when we identify these people we learn something. Most of the time we just don’t understand. Averageness and healthiness are disconnected.

With our wellness programs we’ve decided to use our diagnostic tools to evaluate people. Which with my setup leads to, What happens when someone that’s not average goes in for one of these checkups?   They get identified as potentially having a problem and start to get treated for it. In all likelihood they’ll either continue to be treated to cure their healthiness for the rest of their lives, though instead of being treated for one non-problem, they may have a doctor that wants to aggressive track down the non-problem, and keeps changing up their mistreatment. What happens when you keep trying to fix something that isn’t broken? It’s going to break! Wellness programs are by design going to harm the atypical. Then why do we do it? Because it’s statistically “better”. Where better is a metric that includes lifetime cost of care. “Better” might also consider some other factors like life expectancy, but a treatment that kills half the people that take it instantly but slightly more than doubles the life expectancy of the rest is also “better”. Not to mention that some atypical people who would benefit from being checked, won’t because they’re unhealthily (for them) average. The is a system that helps the “average” at the cost of the atypical.

Which leaves us with, the question, if we keep coming back to eugenics, is it really wrong? Objectively, Yes! Favoring an ideal leads to a monoculture. Monocultures are susceptible to extinction level pandemics. We’re in the middle of the 2nd one for commercial bananas. Also for sexually reproducing populations you have the genetic bottle neck which leds to all the genetic problems from inbreeding. Beyond that, you have to remember that we don’t really understand how our bodies work. What ever statistical ideal we pursue, could be tragically flawed. We’re more likely to make a bad choice than a good one. Also having atypical people let’s us learn from their advantages.

Unfortunately, in our world, wellness programs are likely here to stay. But there is a light of hope on the horizon. Big data medicine. With access to comprehensive records of your vital stats, doctors will know how you are doing against your own baseline. And with much much larger populations of People for which there is data, docotors will be able to come up with multiple models and find one that actually fits you instead of basing your prognosis on a bogus one size fits all model.