23andMe, Science and PR

I have a form of Cystic Fibrosis. It's a genetic disorder and I have a relatively recently identified, relatively mild form of the condition.

I don't know my alleles. My diagnosis is by medical history and sweat chloride test and confirmed because treating for CF was extremely empowering for getting healthier.

So that hypothesis has been tested against reality and if it looks like a form of CF and walks like a form of CF and quacks like a form of CF, it's 99 percent likely to be CF.

I left all the CF lists years ago because knowing your allies is essential street cred for being believed.

Last I checked, there were 1600 known alleles for CF. I had two or three blood tests and I don't have any of the more common ones. They probably initially tested for the hundred or so most common.

I have a weird variation. It's to be expected I don't have the genes for the most common variations. 

My CF specialist requested a more extensive and more expensive test and my insurance denied the request. My sister at the CDC added up the tens of thousands of dollars my insurance company had likely spent on me so far that year and told me that's likely why.

I had been a military dependent my whole life. Insurance bills and medical bills weren't a part of my life.

I didn't battle the denial. I was fighting for my life and getting better and that's all I cared about or had energy for.

Because of my medical situation, I've read up on genetic stuff and tend to remember it.

A woman had a child with CF. She was a carrier. Her husband was not. Medical personnel insisted it wasn't his child and she was unfaithful. She insisted it was his child.

Additional testing showed it was his child and the child had CF due to spontaneous gene mutation in the relevant alleles.

A woman with chimerism was put through an absolute legal nightmare of having to prove the children she birthed were hers. Chimerism means she basically has two sets of DNA and what genetic profile you get depends on where you take the sample.

She likely resorbed a fraternal twin in utero and some tissues had her DNA and some had the twin's DNA. The children weren't genetically hers. They were genetically the offspring of the resorbed twin.

23andMe began as a harmless carney act. It began as the new version of astrology in a more scientific world.

It was "Gee, golly whiz! Isn't this COOL! For a cheek swab, we can tell you who you are related to."

That rapidly went sideways because it began airing a lot of dirty laundry. People got divorced as illicit affairs and the like were divulged with no hope of plausible deniability.

One couple insanely got divorced over a child he fathered out of wedlock as a teenager before he got with her. 

Because of that, 23andMe seems to have largely abandoned the marketing angle of "Find your long lost relatives! Cool!" Even though the science behind identifying your immediate relatives is mostly on solid ground, except for the occasional statistical outlier like the cases described above.

And rather than double down on legal disclaimers and up their game on the science and knowledge of edge cases like the ones above, they have taken a BULLSHIT PR spin doctoring approach to continuing to make money.

Rather than serve the customer.

They could have combed through HIPAA and talked to a lawyer and said upfront something like "This is YOUR private medical information and you have a legal right to do with it as you see fit, but be advised it also contains medical information about other people who share your genes and perhaps read it PRIVATELY and THINK about the consequences FIRST before telling immediate relatives about the results. We are NOT legally responsible for what YOU choose to do with this information."

Instead, they are trying to spin doctor it as "Explore your ethnicity!" when the science for that doesn't really adequately exist and on top of that their attempts to cover their butts legally and disavow stuff involved in accusations of calling someone's mom a HO is actively muddying the waters.

The marketing position is actively underminimg the advancement of genetic science by making up nonsense about "Your biological grandfather who is French may have contributed anywhere from 1 percent to 50 percent French ethnicity!" 

There are plenty of REAL holes in our understanding of genetics. You don't need to whole cloth invent BS spin doctoring here.

You do need to deal realistically with the fact that genetic testing is a can of worms socially because it can and often does reveal uncomfortable truths like "The man who raised you isn't your biological father."

You need to document the actual edge cases and where things can go wrong and educate your users about "Before suing us, FIRST pull your head out your butt and PRIVATELY read YOUR report before showing it to relatives and SECOND read through this list of real edge cases and see if something like that might explain your test results because in a world with billions of people, there will be thousands with a one in a million result."

Their spin doctoring is rapidly veering into outright fraud and is something the entire scientific community should vociferously object to for the widespread harm it is doing to the field of genetics.