Disrupting Healthcare
October 24, 2014 3:49 PM Subscribe
'We Are Going For Change': A Conversation With 23andMe CEO Anne Wojcicki. 'After spending seven months in the Food & Drug Administration’s penalty box, the consumer genetics testing firm 23andMe recently submitted a new health-related test for FDA approval.'
'It was a significant step following last November’s FDA slapdown of 23andMe’s genetic tests, which included health reports outlining customers’ chances of getting a wide variety of diseases from celiac to melanoma. In a sharply worded warning letter, the FDA said the $99 tests, analyzed from a vial of customers’ saliva, constituted a medical device under its regulations, and the company needed to get explicit approval for providing risks of getting specific diseases.'
'Q: Some doctors and regulators think people need a doctor or a geneticist as a middleman to interpret genetic tests, but you have been opposed to requiring that, right?
A: We want to give you the option to talk to a doctor or genetic counselor, but making it mandatory is antithetical to my philosophy. There was also a time period when physicians felt you should not tell somebody is they had a cancer diagnosis. There was a time period when people felt women should not have an at-home pregnancy test because they might do something radical.
Q: Do you ever regret trying to be a pioneer in healthcare, which is about as establishment as you can imagine? Did you ever anticipate it would be this much of a challenge?
A: The only thing I regret is I’d like to have more time with my children. For me, it would either doing 23andMe or doing nothing. I’m driving for change. On a weekly basis, people come up to me and talk about how I’ve saved their life. So I can’t get a better job.'
'Q: Some doctors and regulators think people need a doctor or a geneticist as a middleman to interpret genetic tests, but you have been opposed to requiring that, right?
A: We want to give you the option to talk to a doctor or genetic counselor, but making it mandatory is antithetical to my philosophy. There was also a time period when physicians felt you should not tell somebody is they had a cancer diagnosis. There was a time period when people felt women should not have an at-home pregnancy test because they might do something radical.
Q: Do you ever regret trying to be a pioneer in healthcare, which is about as establishment as you can imagine? Did you ever anticipate it would be this much of a challenge?
A: The only thing I regret is I’d like to have more time with my children. For me, it would either doing 23andMe or doing nothing. I’m driving for change. On a weekly basis, people come up to me and talk about how I’ve saved their life. So I can’t get a better job.'
Oh, and this bit about their funding model should terrify you:
Q: How does 23andMe make money? From the data more than the spit kits, surely?
A: We have partnerships with companies all the time–with pharma partners, with academic groups–and one of our main goals is to dramatically accelerate the pace of research.
One of the ways I can do that is by enabling individuals, instead of having to start a study de novo and recruit a thousand people with Parkinson’s, to get access to the database. Someone might come to us and say, we want to see if there are any genetic variants associated with people with asthma. Or someone might hypothesize that a variant in a gene is associated with a disease, and we can do a data query. Instead of actually having to do clinical trials the old-fashioned way, we could enable researchers to get their answers instantaneously. And they pay us for that.
You use her service, you're not the customer - you're the product.
posted by NoxAeternum at 3:58 PM on October 24, 2014 [39 favorites]
Q: How does 23andMe make money? From the data more than the spit kits, surely?
A: We have partnerships with companies all the time–with pharma partners, with academic groups–and one of our main goals is to dramatically accelerate the pace of research.
One of the ways I can do that is by enabling individuals, instead of having to start a study de novo and recruit a thousand people with Parkinson’s, to get access to the database. Someone might come to us and say, we want to see if there are any genetic variants associated with people with asthma. Or someone might hypothesize that a variant in a gene is associated with a disease, and we can do a data query. Instead of actually having to do clinical trials the old-fashioned way, we could enable researchers to get their answers instantaneously. And they pay us for that.
You use her service, you're not the customer - you're the product.
posted by NoxAeternum at 3:58 PM on October 24, 2014 [39 favorites]
There are many ways that America's Healthcare System desperately needs disrupting. 23andMe does NONE of them.
posted by oneswellfoop at 4:04 PM on October 24, 2014 [16 favorites]
posted by oneswellfoop at 4:04 PM on October 24, 2014 [16 favorites]
I have a scientific question about that bit, as well. Aren't new SNPs-of-interest being discovered all the time? If they aren't sequencing the whole genome, aren't the researchers going to be looking at pre-filtered data?
Or are they sequencing the whole thing for later use, but only analyzing known SNPs? In which case great googly moogly is that troublesome.
posted by muddgirl at 4:07 PM on October 24, 2014 [1 favorite]
Or are they sequencing the whole thing for later use, but only analyzing known SNPs? In which case great googly moogly is that troublesome.
posted by muddgirl at 4:07 PM on October 24, 2014 [1 favorite]
Everything this Corp asserts needs to be tempered with a healthy dose of the Mermaids Tale. http://ecodevoevo.blogspot.ca
posted by clvrmnky at 4:09 PM on October 24, 2014 [2 favorites]
posted by clvrmnky at 4:09 PM on October 24, 2014 [2 favorites]
You use her service, you're not the customer - you're the product.
Do you have a 23andme account? I do, and I respectfully think your binary assessment is pithy but not really accurate — or useful.
I think it's pretty damn awesome that in 2014, I can buy and download my raw SNP variant data for under a hundred bucks, and I can analyze it and explore it with readily-available open-source genomics tools. It is a pretty amazing time to be alive.
posted by a lungful of dragon at 4:10 PM on October 24, 2014 [64 favorites]
Do you have a 23andme account? I do, and I respectfully think your binary assessment is pithy but not really accurate — or useful.
I think it's pretty damn awesome that in 2014, I can buy and download my raw SNP variant data for under a hundred bucks, and I can analyze it and explore it with readily-available open-source genomics tools. It is a pretty amazing time to be alive.
posted by a lungful of dragon at 4:10 PM on October 24, 2014 [64 favorites]
I think I felt my brain trying to flee when I read this. This is the problem with the cult of the amateur, folks.
I get your point, on the other hand, I and people I know have been given far too much medical attention that turned out to be either unnecessary or even downright harmful but always expensive to have complete faith in the medical profession. I don't know what the right balance is, but I've come to the point of considering doctors guilty until proven innocent.
posted by IndigoJones at 4:11 PM on October 24, 2014 [18 favorites]
I get your point, on the other hand, I and people I know have been given far too much medical attention that turned out to be either unnecessary or even downright harmful but always expensive to have complete faith in the medical profession. I don't know what the right balance is, but I've come to the point of considering doctors guilty until proven innocent.
posted by IndigoJones at 4:11 PM on October 24, 2014 [18 favorites]
a lungful of dragon: " It is a pretty amazing time to be alive."
Yeah, that part is awesome. But the part where your data is aggregated and sold is maybe not so hot.
posted by boo_radley at 4:15 PM on October 24, 2014 [18 favorites]
Yeah, that part is awesome. But the part where your data is aggregated and sold is maybe not so hot.
posted by boo_radley at 4:15 PM on October 24, 2014 [18 favorites]
I'm okay with being the product in a situation where my data is being used for medical research that could save lives. I definitely plan to pay 23AndMe or a similar service for a test at some point in the future, with anything medical related being discussed with my doctor.
posted by Drinky Die at 4:16 PM on October 24, 2014 [10 favorites]
posted by Drinky Die at 4:16 PM on October 24, 2014 [10 favorites]
A lack of faith in doctors and a lack of faith in for-profit genetic profiling companies need not be a balance. One can distrust both completely, and with good reason.
posted by graymouser at 4:16 PM on October 24, 2014 [12 favorites]
posted by graymouser at 4:16 PM on October 24, 2014 [12 favorites]
What are some of these "readily-available open source genomics tools"? And are they still compatible with whatever it is you can get out of 23 and Me post-FDA smackdown?
posted by Flunkie at 4:18 PM on October 24, 2014 [1 favorite]
posted by Flunkie at 4:18 PM on October 24, 2014 [1 favorite]
Oh, and this bit about their funding model should terrify you [ . . .] You use her service, you're not the customer - you're the product.
I dunno. It doesn't terrify me and it strikes me as a potentially very useful and pragmatic way to accelerate research.
What's so scary and/or pernicious about it? Is that I'm paying them and yet they're using my data when, perhaps, according to some views, I should be compensated for that use instead? Is there a fear that they'll share the data with insurance companies and/or government entities who will use it to personally discriminate against me?
posted by treepour at 4:19 PM on October 24, 2014 [8 favorites]
I dunno. It doesn't terrify me and it strikes me as a potentially very useful and pragmatic way to accelerate research.
What's so scary and/or pernicious about it? Is that I'm paying them and yet they're using my data when, perhaps, according to some views, I should be compensated for that use instead? Is there a fear that they'll share the data with insurance companies and/or government entities who will use it to personally discriminate against me?
posted by treepour at 4:19 PM on October 24, 2014 [8 favorites]
I'm okay with being the product in a situation where my data is being used for medical research that could save lives.
But you don't control how it's used - 23andMe does. There isn't any use for your data that you would object to? Military research? Studies to attempt to determine a genetic basis for african-americans and hispanics having a lower IQ?
Also, I haven't been able to find an answer to what, exactly, 23andMe stores in their databases for medical researchers.
posted by muddgirl at 4:20 PM on October 24, 2014 [4 favorites]
But you don't control how it's used - 23andMe does. There isn't any use for your data that you would object to? Military research? Studies to attempt to determine a genetic basis for african-americans and hispanics having a lower IQ?
Also, I haven't been able to find an answer to what, exactly, 23andMe stores in their databases for medical researchers.
posted by muddgirl at 4:20 PM on October 24, 2014 [4 favorites]
Do you have a 23andme account? I do, and I respectfully think your binary assessment is pithy but not really accurate — or useful.
I think it's pretty damn awesome that in 2014, I can buy and download my raw SNP variant data for under a hundred bucks, and I can analyze it and explore it with readily-available open-source genomics tools. It is a pretty amazing time to be alive.
Except that the price of getting that data wasn't just $99 - it was giving them the data for their use to sell. Which is something that the CEO openly admits.
But you are right - I don't have a 23AndMe account. Because I don't trust the company.
posted by NoxAeternum at 4:20 PM on October 24, 2014 [6 favorites]
I think it's pretty damn awesome that in 2014, I can buy and download my raw SNP variant data for under a hundred bucks, and I can analyze it and explore it with readily-available open-source genomics tools. It is a pretty amazing time to be alive.
Except that the price of getting that data wasn't just $99 - it was giving them the data for their use to sell. Which is something that the CEO openly admits.
But you are right - I don't have a 23AndMe account. Because I don't trust the company.
posted by NoxAeternum at 4:20 PM on October 24, 2014 [6 favorites]
But the part where your data is aggregated and sold is maybe not so hot.
The way genome-wide analysis studies ("GWAS") work is by aggregating (anonymously) all these genetic variations, to try to find patterns. The more datapoints, the stronger the association. This is how these and other genomic tests are able to tell you are at greater risk for various ailments in the first place.
posted by a lungful of dragon at 4:20 PM on October 24, 2014 [9 favorites]
The way genome-wide analysis studies ("GWAS") work is by aggregating (anonymously) all these genetic variations, to try to find patterns. The more datapoints, the stronger the association. This is how these and other genomic tests are able to tell you are at greater risk for various ailments in the first place.
posted by a lungful of dragon at 4:20 PM on October 24, 2014 [9 favorites]
What are some of these "readily-available open source genomics tools"?
F.ex.: Promethease
posted by VikingSword at 4:23 PM on October 24, 2014 [9 favorites]
F.ex.: Promethease
posted by VikingSword at 4:23 PM on October 24, 2014 [9 favorites]
And what guarantee do you have that they will stick to that sort of analysis?
posted by NoxAeternum at 4:24 PM on October 24, 2014
posted by NoxAeternum at 4:24 PM on October 24, 2014
"There was a time period when people felt women should not have an at-home pregnancy test because they might do something radical." I think I felt my brain trying to flee when I read this. This is the problem with the cult of the amateur, folks.
What is the problem with the cult of the amateur? Where do you draw the line about what information it's appropriate for an individual to have?
posted by the jam at 4:26 PM on October 24, 2014 [27 favorites]
What is the problem with the cult of the amateur? Where do you draw the line about what information it's appropriate for an individual to have?
posted by the jam at 4:26 PM on October 24, 2014 [27 favorites]
And are they still compatible with whatever it is you can get out of 23 and Me post-FDA smackdown?
For SNPs, you get a tab-delimited text file you can post-process to your heart's desire with any number of command-line UNIX utilities and genomics tools, and look up against SNP databases, like 1000 Genomes. There is also a RESTful API for ancestry-related and other queries, which spits out JSON, which is a text-based format you can parse with any number of frameworks in modern scripting and procedural languages.
posted by a lungful of dragon at 4:29 PM on October 24, 2014 [7 favorites]
For SNPs, you get a tab-delimited text file you can post-process to your heart's desire with any number of command-line UNIX utilities and genomics tools, and look up against SNP databases, like 1000 Genomes. There is also a RESTful API for ancestry-related and other queries, which spits out JSON, which is a text-based format you can parse with any number of frameworks in modern scripting and procedural languages.
posted by a lungful of dragon at 4:29 PM on October 24, 2014 [7 favorites]
... genetic variants associated with people with asthma.
Have asthma by what measure? Someone checked a box?
I am not excited about research based on self-reporting.
posted by Lesser Shrew at 4:34 PM on October 24, 2014 [4 favorites]
Have asthma by what measure? Someone checked a box?
I am not excited about research based on self-reporting.
posted by Lesser Shrew at 4:34 PM on October 24, 2014 [4 favorites]
And what guarantee do you have that they will stick to that sort of analysis?
It's up to either the courts to find GINA unconstitutional, or for legislators to enact a law that overrides GINA. Until then, this FAQ entry explains how 23andme's sharing of the data they collect is restricted by GINA. You can also read their privacy policy and consent policy. If these policies change, you (the end user) get to agree to or decline new terms.
Have asthma by what measure? Someone checked a box?
Drug companies that make asthma medicine, for instance, collect this phenotypic data as a matter of course, when doing clinical trials.
To the larger point about privacy, when aggregating GWAS data for larger bioinformatics analyses in academia, there is massive amounts of paperwork that must go before a committee board for approval — further, the GWAS data you request must be tied to a specific clinical research project with goals that are tied to the association. You can't just ask for this stuff out of random interest.
posted by a lungful of dragon at 4:41 PM on October 24, 2014 [7 favorites]
It's up to either the courts to find GINA unconstitutional, or for legislators to enact a law that overrides GINA. Until then, this FAQ entry explains how 23andme's sharing of the data they collect is restricted by GINA. You can also read their privacy policy and consent policy. If these policies change, you (the end user) get to agree to or decline new terms.
Have asthma by what measure? Someone checked a box?
Drug companies that make asthma medicine, for instance, collect this phenotypic data as a matter of course, when doing clinical trials.
To the larger point about privacy, when aggregating GWAS data for larger bioinformatics analyses in academia, there is massive amounts of paperwork that must go before a committee board for approval — further, the GWAS data you request must be tied to a specific clinical research project with goals that are tied to the association. You can't just ask for this stuff out of random interest.
posted by a lungful of dragon at 4:41 PM on October 24, 2014 [7 favorites]
Information is a "medical device"? Fuck the FDA.
posted by telstar at 4:52 PM on October 24, 2014 [19 favorites]
posted by telstar at 4:52 PM on October 24, 2014 [19 favorites]
I am not excited about research based on self-reporting.
With large enough numbers and fun analysis, you can get a lot of utility out self-reporting.
posted by the jam at 4:56 PM on October 24, 2014 [6 favorites]
With large enough numbers and fun analysis, you can get a lot of utility out self-reporting.
posted by the jam at 4:56 PM on October 24, 2014 [6 favorites]
But you don't control how it's used - 23andMe does. There isn't any use for your data that you would object to? Military research? Studies to attempt to determine a genetic basis for african-americans and hispanics having a lower IQ?
I've had several family members, including a baby niece, who have died of cancer during my lifetime. I'm not really worried about hypothetical rich racists paying for studies or whatever.
posted by Drinky Die at 5:09 PM on October 24, 2014 [29 favorites]
I've had several family members, including a baby niece, who have died of cancer during my lifetime. I'm not really worried about hypothetical rich racists paying for studies or whatever.
posted by Drinky Die at 5:09 PM on October 24, 2014 [29 favorites]
Oh, and this bit about their funding model should terrify you:You are straining at gnats here . . .
Q: How does 23andMe make money? From the data more than the spit kits, surely?
A: We have partnerships with companies all the time–with pharma partners, with academic groups–and one of our main goals is to dramatically accelerate the pace of research.
One of the ways I can do that is by enabling individuals, instead of having to start a study de novo and recruit a thousand people with Parkinson’s, to get access to the database. Someone might come to us and say, we want to see if there are any genetic variants associated with people with asthma. Or someone might hypothesize that a variant in a gene is associated with a disease, and we can do a data query. Instead of actually having to do clinical trials the old-fashioned way, we could enable researchers to get their answers instantaneously. And they pay us for that.
You use her service, you're not the customer - you're the product.
posted by NoxAeternum
...
But you are right - I don't have a 23AndMe account. Because I don't trust the company.
posted by NoxAeternum
and the camels you are swallowing, apparently all unawares, are the size of dinosaurs:
The Tissue-Industrial ComplexThe ethical standards of the 23AndMe business model are far, far superior to those of the business model of almost any medical center or physician you might choose to go to.
Today most Americans have their tissue on file somewhere. In 1999 the RAND Corporation published a report (the first and, so far, the last of its kind) with what it called a "conservative estimate" that more than 307 million tissue samples from more than 178 million people were stored in the United States. This number, the report said, was increasing by more than 20 million samples each year. These samples come from routine medical tests, operations, clinical trials and research donations. They sit in lab freezers, on shelves or in industrial vats of liquid nitrogen. They're stored at military facilities, the F.B.I. and the National Institutes of Health. They're in biotech companies and most hospitals. Biobanks store everything from appendixes, ovaries and skin to sphincters, testicles and fat. Not to mention blood samples taken from most children born in the United States since the late 60's, when states started mandating screening newborns for genetic diseases.
Scientists and surgeons use these tissues to develop everything from flu vaccines to penis-enlargement products. They put cells in culture dishes and expose them to radiation, drugs, cosmetics, viruses, household chemicals and biological weapons and then study their responses. They remove DNA to examine it — and therefore the person it came from — gene by gene. Without those tissues, we would have no tests for diseases like hepatitis and H.I.V.; no vaccines for polio, smallpox, measles; none of the new promising drugs for leukemia, breast cancer, colon cancer. And without tissue samples, the developers of those products would be out billions of dollars.
...
Then Golde appealed and won. In 1990, nearly seven years after Moore filed suit, the Supreme Court of California ruled against Moore on 11 counts in what has become the definitive statement on this issue: any ownership you might have in your tissues vanishes when they are removed from your body, with or without consent. When you leave tissues in a doctor's office or a lab, you abandon them as waste. Anyone can take your garbage and sell it — the same goes for your tissues. Most important, the court said, Moore couldn't own his cells, because that would conflict with Golde's patent. Golde had "transformed" those cells into an invention. They were, the ruling said, the product of Golde's "human ingenuity" and "inventive effort."
...
posted by jamjam at 5:12 PM on October 24, 2014 [49 favorites]
Also, if you're Canadian you can get all the health genetic information back at 23andMe.ca, it's back to being awesome.
posted by blue_beetle at 5:36 PM on October 24, 2014
posted by blue_beetle at 5:36 PM on October 24, 2014
To make 23andme work you needed a mailing address and a method of payment. Especially if you were ordering multiple tests (you associate each kit with an identifier you make up), the results would have been difficult to reliably link to the full details of a particular individual. That is, if you're inclined to be mistrustful of these things.
posted by tigrrrlily at 5:39 PM on October 24, 2014 [1 favorite]
posted by tigrrrlily at 5:39 PM on October 24, 2014 [1 favorite]
Studies to attempt to determine a genetic basis for african-americans and hispanics having a lower IQ?
Junk science gunna be junk science. I'm not afraid of people looking for that sort of thing.
What would be fun is running DNA tests on racists.
posted by el io at 6:06 PM on October 24, 2014 [1 favorite]
Junk science gunna be junk science. I'm not afraid of people looking for that sort of thing.
What would be fun is running DNA tests on racists.
posted by el io at 6:06 PM on October 24, 2014 [1 favorite]
Aren't new SNPs-of-interest being discovered all the time? If they aren't sequencing the whole genome, aren't the researchers going to be looking at pre-filtered data?
We actually know the locations of almost all the SNPs, and 23andme can cheaply determine their values using an Affymetrix gene chip. We don't have a clear idea what most of those SNPs actually do. When there are new discoveries about currently-known SNPs, 23andme will let you know.
They're not doing any sequencing. Wait about 10 years for that one.
posted by miyabo at 6:06 PM on October 24, 2014 [1 favorite]
We actually know the locations of almost all the SNPs, and 23andme can cheaply determine their values using an Affymetrix gene chip. We don't have a clear idea what most of those SNPs actually do. When there are new discoveries about currently-known SNPs, 23andme will let you know.
They're not doing any sequencing. Wait about 10 years for that one.
posted by miyabo at 6:06 PM on October 24, 2014 [1 favorite]
While studies show early adopters understand the data better than expected, they also show the general public does not understand the limits in regards to accuracy/predictive value. And DTC companies do themselves no favors with misleading in marketing and sales tactics This is why the government got involved at all and mandated CLIA standards, thus forcing several companies out of business. Actually just look at the entire set of entries here regarding lawsuits, countersuits, sample swaps, and companies leaving the field entirely. So while I respect what 23andME is attempting, there's clearly a lot of controversy at all levels from accuracy to privacy to education.
But this is true for even the best regarded genetic testing companies and there's a long history of people making irreversible medical decisions that retrospectively showed poor understanding of risk. But how best to put this information forth won't be known if the information isn't ever put forth. By long history, I mean current issues in clinical genetic testing and follow up such as: does a variant of unknown significance lead to breast cancer and so should a mastectomy be done? Does a positive NIPT result mean fetus has a chromosome anomaly or is further testing needed? Does a Tay Sachs screen which is geared toward the Ashkenazi Jewish population have any validity for the Sephardic Jew-is it even needed? I know what the geneticists think but it's interesting what other specialists believe, and the general public is even less likely to good basis for judging what they are told.
Essentially the marketing is driving the science right now, not vice versa. And there will always be naysayers and second guessers as how things could be better. Which is how things get better.
posted by beaning at 6:12 PM on October 24, 2014 [6 favorites]
But this is true for even the best regarded genetic testing companies and there's a long history of people making irreversible medical decisions that retrospectively showed poor understanding of risk. But how best to put this information forth won't be known if the information isn't ever put forth. By long history, I mean current issues in clinical genetic testing and follow up such as: does a variant of unknown significance lead to breast cancer and so should a mastectomy be done? Does a positive NIPT result mean fetus has a chromosome anomaly or is further testing needed? Does a Tay Sachs screen which is geared toward the Ashkenazi Jewish population have any validity for the Sephardic Jew-is it even needed? I know what the geneticists think but it's interesting what other specialists believe, and the general public is even less likely to good basis for judging what they are told.
Essentially the marketing is driving the science right now, not vice versa. And there will always be naysayers and second guessers as how things could be better. Which is how things get better.
posted by beaning at 6:12 PM on October 24, 2014 [6 favorites]
They're not doing any sequencing. Wait about 10 years for that one.
Whole exome sequencing is currently fairly widely available with varying degrees of "physician referral required."
posted by beaning at 6:19 PM on October 24, 2014 [1 favorite]
Whole exome sequencing is currently fairly widely available with varying degrees of "physician referral required."
posted by beaning at 6:19 PM on October 24, 2014 [1 favorite]
The need to ensure the accuracy of the information these guys are providing is why FDA regulation of this product is needed. Without oversight, why would you trust what a private company is telling you? How do we know their analyses are valid? How does the average consumer, presumably without college-level training in statistics or biology, know how to adequately assess the risks associated with the information? I fully support the right to have your own information, but I'm very wary of someone deciding on a medical procedure based on it without consultation from a medical professional.
Besides, it's not like this is some sort of unknown regulatory path here. This isn't a small molecule pharmaceutical that needs 15 years of Phase I, II and III trials. This is a 510(k) approval. 90 days. University start-ups do these all the time, it's not like it's strangling them at birth.
posted by Existential Dread at 6:38 PM on October 24, 2014 [8 favorites]
Besides, it's not like this is some sort of unknown regulatory path here. This isn't a small molecule pharmaceutical that needs 15 years of Phase I, II and III trials. This is a 510(k) approval. 90 days. University start-ups do these all the time, it's not like it's strangling them at birth.
posted by Existential Dread at 6:38 PM on October 24, 2014 [8 favorites]
I fully support the right to have your own information, but I'm very wary of someone deciding on a medical procedure based on it without consultation from a medical professional.
Could you expand on this? It doesn't make any sense to me and I'm guessing there are a few ideas you've smooshed together to get to what parses as "I'd hate for someone to get services from a medical professional without consulting with a medical professional."
Do you mean you'd prefer people see a genetic counselor before working with a medical professional who'd deliver a particular procedure on that counselor's advice?
posted by mph at 6:48 PM on October 24, 2014 [3 favorites]
Could you expand on this? It doesn't make any sense to me and I'm guessing there are a few ideas you've smooshed together to get to what parses as "I'd hate for someone to get services from a medical professional without consulting with a medical professional."
Do you mean you'd prefer people see a genetic counselor before working with a medical professional who'd deliver a particular procedure on that counselor's advice?
posted by mph at 6:48 PM on October 24, 2014 [3 favorites]
FFS -- if you don't want to "be the product" or participate in the "cult of the amateur," whatever the hell that means, then don't. But why do people always want to take the choice away from others?
posted by Behemoth at 7:04 PM on October 24, 2014 [16 favorites]
posted by Behemoth at 7:04 PM on October 24, 2014 [16 favorites]
I don't think the goal is to remove the choice-it's more to make sure everyone is clear on (1) what one is or is not getting back as far as personal medical information, and (2) that your personal DNA information is going beyond your control. If you know what you're getting involved in, have at it. But don't believe for an instant that the dollars behind 23andMe aren't out to squeeze this information for everything they can. 23andMe has been very clear about wanting to be "the Google of DNA research."
posted by beaning at 7:31 PM on October 24, 2014
posted by beaning at 7:31 PM on October 24, 2014
Meaningful choice, rather than just a catchphrase, requires education on the implications of choices you make or they are no choice at all, just freedom to be fucked by the more powerful amongst us
see: all of human history
posted by lalochezia at 7:32 PM on October 24, 2014 [3 favorites]
see: all of human history
posted by lalochezia at 7:32 PM on October 24, 2014 [3 favorites]
Could you expand on this? It doesn't make any sense to me and I'm guessing there are a few ideas you've smooshed together to get to what parses as "I'd hate for someone to get services from a medical professional without consulting with a medical professional."
Well, I was thinking of the BRCA gene example. Many women have been tested and diagnosed with this gene, which has frightening implications for breast cancer risk. Many have opted for major surgery to mitigate that risk. It's not entirely clear whether they are opening themselves up for greater risk by undergoing surgery (hemorrhage, infection, etc) in the absence of disease than they would by monitoring to see if cancer developed or not.
Prior to the FDA review, 23andme was providing risk probabilities for certain genetic diseases and disorders to customers, with the option for genetic counseling. What's more, these risk probabilities weren't reviewed by any oversight or regulatory body. My worry, specifically, is that people would interpret these results incorrectly, or the results themselves would be flawed, and the customer would decide on a course of action or treatment that would fail them. In this case, they would have the information but not the correct interpretation.
posted by Existential Dread at 7:38 PM on October 24, 2014 [5 favorites]
Well, I was thinking of the BRCA gene example. Many women have been tested and diagnosed with this gene, which has frightening implications for breast cancer risk. Many have opted for major surgery to mitigate that risk. It's not entirely clear whether they are opening themselves up for greater risk by undergoing surgery (hemorrhage, infection, etc) in the absence of disease than they would by monitoring to see if cancer developed or not.
Prior to the FDA review, 23andme was providing risk probabilities for certain genetic diseases and disorders to customers, with the option for genetic counseling. What's more, these risk probabilities weren't reviewed by any oversight or regulatory body. My worry, specifically, is that people would interpret these results incorrectly, or the results themselves would be flawed, and the customer would decide on a course of action or treatment that would fail them. In this case, they would have the information but not the correct interpretation.
posted by Existential Dread at 7:38 PM on October 24, 2014 [5 favorites]
We want to give you the option to talk to a doctor or genetic counselor, but making it mandatory is antithetical to my philosophy. There was also a time period when physicians felt you should not tell somebody is they had a cancer diagnosis. There was a time period when people felt women should not have an at-home pregnancy test because they might do something radical.
I think I felt my brain trying to flee when I read this. This is the problem with the cult of the amateur, folks.--NoxAeternum
This one really burns my barnacles. Cult of the Amateurs?? And the example you quote is doctors not telling patients they have cancer and not allowing at-home pregnancy tests??
posted by eye of newt at 7:49 PM on October 24, 2014 [9 favorites]
I think I felt my brain trying to flee when I read this. This is the problem with the cult of the amateur, folks.--NoxAeternum
This one really burns my barnacles. Cult of the Amateurs?? And the example you quote is doctors not telling patients they have cancer and not allowing at-home pregnancy tests??
posted by eye of newt at 7:49 PM on October 24, 2014 [9 favorites]
To build on Existential Dread's statement that FDA regulation on an entity like 23andme as a medical device manufacturer is much less taxing than what would be required of a pharmaceutical, I would also state that if they are offering such a test, they should have done all the science required for a 510k submission waaaaaay before ever offering.
the 510k submission itself should just be a formalization of work they have already done, i.e., preparing the paperwork. the work required for a 510k submission is not crazy, overly detailed protocols; it is characterization that should be done on any assay, and doubly so on an assay that will be used by people to make life choices.
I feel that seven months is a pretty speedy timeline to complete all the work for a 510k (or to complete the evaluation of any assay), so i kind of suspect they already performed a lot of the characterization, so great. good advisors should have recommended submitting this data before they ran into problems... I am also a little leary of the data storing, etc. they want to do on top of the genetic screening; *that* sounds like a trial to get through FDA approval.
Disclaimer: I work for a medical device manufacturer that is subject to government regulation. So far, none of the science I've done to meet FDA requirements has ever seemed superfluous.
posted by Tandem Affinity at 8:06 PM on October 24, 2014 [4 favorites]
the 510k submission itself should just be a formalization of work they have already done, i.e., preparing the paperwork. the work required for a 510k submission is not crazy, overly detailed protocols; it is characterization that should be done on any assay, and doubly so on an assay that will be used by people to make life choices.
I feel that seven months is a pretty speedy timeline to complete all the work for a 510k (or to complete the evaluation of any assay), so i kind of suspect they already performed a lot of the characterization, so great. good advisors should have recommended submitting this data before they ran into problems... I am also a little leary of the data storing, etc. they want to do on top of the genetic screening; *that* sounds like a trial to get through FDA approval.
Disclaimer: I work for a medical device manufacturer that is subject to government regulation. So far, none of the science I've done to meet FDA requirements has ever seemed superfluous.
posted by Tandem Affinity at 8:06 PM on October 24, 2014 [4 favorites]
Well-defined phenotypes are important for GWAS. It's not snarky to ask how, e.g., asthma is defined in the 23andme database. If it's based on self-reporting in a binary (yes or no) fashion then I suspect you're going to see a lot of false positives, which is a well-known problem.
posted by wintermind at 8:19 PM on October 24, 2014
posted by wintermind at 8:19 PM on October 24, 2014
How many of those who think this is cool invested in CDOs eight years ago?
posted by carping demon at 8:34 PM on October 24, 2014
posted by carping demon at 8:34 PM on October 24, 2014
Winded after she rode an ElliptiGO bike...
Stop right there. That company is headed straight to the toilet.
posted by gurple at 8:41 PM on October 24, 2014 [8 favorites]
Stop right there. That company is headed straight to the toilet.
posted by gurple at 8:41 PM on October 24, 2014 [8 favorites]
In the first place, I'm tired of the idea that the majority of people are ignoramuses and anything with more than five letters is over their heads, thus all medical material should be left to the brilliant experts - meaning anyone with an M.D. behind their name - and the average Joe should just shut up and do what the doctor says. Also, the doctor's too busy to answer questions, so if you're going to ask one be sure it's pertinent and shows that you at least understand the basics - if you're not sure, best to just keep quiet.
It's amazing to me how many people in this country think they're unable to grasp medical information; it's all baloney - all it really takes is interest and study and an average intelligence can understand whatever it is they're trying to learn. With the medical dictionaries available on the internet and good sources of information that are so prevalent from major medical facilities and support groups, it's just not the mystery it's cracked up to be. What keeps people from pursuing answers to all those "dumb" questions is the intimidation factor, nothing else.
If my mother had stopped trying to get medical proof that my sister had grand mal epilepsy when the doctors told her to (even though she had major tonic clonic seizures daily) they never would have kept going until they found it. She had been treated in Dallas, Phoenix, Tucson, at the Los Angeles Children's Hospital - no one could get it to show up on an EEG; when she got to Stanford University they put leads in her nose and BINGO! How 'bout that, huh?
If we think about it, and most of us try not to spend too much time doing that, we've all seen the failure of doctors and tests and misdiagnosis and missed diagnosis and false positives and false negatives and medication mistakes and incompatibilities and ... death as a result of those failures. Yes, I'd be willing to bet that a loved one has died in every family from a medical error, even if that's not information that's been revealed. My friend thinks his heart doctor is amazing because she saved his life; actually, when he had his first heart attack and she placed a stent she pushed a fat embolus up higher into the artery, which resulted in a full arrest a week or so later, at which time a visiting friend, home temporarily from Iraq, performed CPR and kept him alive until they could get him under electrical stimulation and then place a pacemaker/defibrillator. So was it a screw-up by the stent doctor or just one of those things? I don't say it's an error, but it's certainly an open question - but my point is, he doesn't think so at all.
A friend died at age 38 from raging sepsis; they sent him home from the ER on Wednesday night with a diagnosis of "some arthritis" and a few pain pills; on Friday morning he went by ambulance to the hospital and was dead at 8:00 that evening.
We must - absolutely must - get it out of our heads that our medical system is so complex we can't understand it and thus we have no right to question it because when we question a private company that's doing DNA studies or genomic research or full-body CAT scans or open lab work just because we "don't trust 'em because they're not part of the medical mainstream" and if that distrust results in the unavailability of these private services, we're simply contributing to a closed system with little accountability and little oversight.
If I decide I think I need a certain lab test and I'm willing to pay for it, if I can go to a lab and get the test done without having to ring up $200 in bills and physical and mental exhaustion trying to get my physician to agree to the test, that's a good thing. If the test comes back positive I can then go to my doc with the test results and he can decide if he wants to retest me through a different lab or pursue some other angle instead; if the test comes back negative, then my worry is mostly alleviated. If the symptoms later escalate, I can then discuss being retested with my doctor - or maybe a different test - whatever he thinks.
Every test we have, every diagnosis we get, is being fed into the massive database at one level or another that will one day have all our medical data available to every insurance company and medical practice - there's no way that's ever going to be stopped now that we have the electronic medical records and the big push to consolidate hospitals and medical clinics together into corporations of their own. If my 23andme data goes into the same database, so? No one is (yet) being forced to submit their saliva for DNA analysis.
If people misunderstand their test results from private testing, or get scared they have cancer or - that happens anyway, everyday, in the doctor's office. And the opposite, also - people don't seek medical help until it's way too late to help them because they're absolutely certain there's not a damn thing wrong with them.
Times are changing and it's time some things did. I think my medical journey should be up to me as much as possible.
posted by aryma at 9:48 PM on October 24, 2014 [46 favorites]
It's amazing to me how many people in this country think they're unable to grasp medical information; it's all baloney - all it really takes is interest and study and an average intelligence can understand whatever it is they're trying to learn. With the medical dictionaries available on the internet and good sources of information that are so prevalent from major medical facilities and support groups, it's just not the mystery it's cracked up to be. What keeps people from pursuing answers to all those "dumb" questions is the intimidation factor, nothing else.
If my mother had stopped trying to get medical proof that my sister had grand mal epilepsy when the doctors told her to (even though she had major tonic clonic seizures daily) they never would have kept going until they found it. She had been treated in Dallas, Phoenix, Tucson, at the Los Angeles Children's Hospital - no one could get it to show up on an EEG; when she got to Stanford University they put leads in her nose and BINGO! How 'bout that, huh?
If we think about it, and most of us try not to spend too much time doing that, we've all seen the failure of doctors and tests and misdiagnosis and missed diagnosis and false positives and false negatives and medication mistakes and incompatibilities and ... death as a result of those failures. Yes, I'd be willing to bet that a loved one has died in every family from a medical error, even if that's not information that's been revealed. My friend thinks his heart doctor is amazing because she saved his life; actually, when he had his first heart attack and she placed a stent she pushed a fat embolus up higher into the artery, which resulted in a full arrest a week or so later, at which time a visiting friend, home temporarily from Iraq, performed CPR and kept him alive until they could get him under electrical stimulation and then place a pacemaker/defibrillator. So was it a screw-up by the stent doctor or just one of those things? I don't say it's an error, but it's certainly an open question - but my point is, he doesn't think so at all.
A friend died at age 38 from raging sepsis; they sent him home from the ER on Wednesday night with a diagnosis of "some arthritis" and a few pain pills; on Friday morning he went by ambulance to the hospital and was dead at 8:00 that evening.
We must - absolutely must - get it out of our heads that our medical system is so complex we can't understand it and thus we have no right to question it because when we question a private company that's doing DNA studies or genomic research or full-body CAT scans or open lab work just because we "don't trust 'em because they're not part of the medical mainstream" and if that distrust results in the unavailability of these private services, we're simply contributing to a closed system with little accountability and little oversight.
If I decide I think I need a certain lab test and I'm willing to pay for it, if I can go to a lab and get the test done without having to ring up $200 in bills and physical and mental exhaustion trying to get my physician to agree to the test, that's a good thing. If the test comes back positive I can then go to my doc with the test results and he can decide if he wants to retest me through a different lab or pursue some other angle instead; if the test comes back negative, then my worry is mostly alleviated. If the symptoms later escalate, I can then discuss being retested with my doctor - or maybe a different test - whatever he thinks.
Every test we have, every diagnosis we get, is being fed into the massive database at one level or another that will one day have all our medical data available to every insurance company and medical practice - there's no way that's ever going to be stopped now that we have the electronic medical records and the big push to consolidate hospitals and medical clinics together into corporations of their own. If my 23andme data goes into the same database, so? No one is (yet) being forced to submit their saliva for DNA analysis.
If people misunderstand their test results from private testing, or get scared they have cancer or - that happens anyway, everyday, in the doctor's office. And the opposite, also - people don't seek medical help until it's way too late to help them because they're absolutely certain there's not a damn thing wrong with them.
Times are changing and it's time some things did. I think my medical journey should be up to me as much as possible.
posted by aryma at 9:48 PM on October 24, 2014 [46 favorites]
miyabo: "We actually know the locations of almost all the SNPs"
That seems like a strong claim. We probably know most of the widely prevalent ones, but given the rate of transcription errors, every living person probably has a few of their own. In fact, it appears that between build 141 and build 142, the SNPs known to man doubled. That's a pretty high growth rate for a current state of 'almost all.'!
Assuming we actually know anything at all.
posted by pwnguin at 10:16 PM on October 24, 2014 [4 favorites]
That seems like a strong claim. We probably know most of the widely prevalent ones, but given the rate of transcription errors, every living person probably has a few of their own. In fact, it appears that between build 141 and build 142, the SNPs known to man doubled. That's a pretty high growth rate for a current state of 'almost all.'!
Assuming we actually know anything at all.
posted by pwnguin at 10:16 PM on October 24, 2014 [4 favorites]
Information is a "medical device"? Fuck the FDA.You know, there actually is a device here, the SNP microarrays, the scanners that read the microarrays, a huge amount of analysis software, not to mention all the devices involved in the sample preparation to extract DNA from saliva, etc. We're used to thinking of digital information like DNA as perfectly accurate, able to be copied with error, but that's simply not the case here. There's a heavily amplified analog signal that's being converted into a digital DNA base call, and there's plenty of noise, plenty of ways to screw up the DNA extraction, plenty of ways to screw up the DNA hybridization to the chip, plenty of ways to screw up the scan, and even plenty of ways to swap labels. Something like 1% of cancer pathology slides end up being for the wrong person. I get mismatched samples all the time.
You may only see the medical "information," from 23andMe, but it most definitely comes from a device, and that it's so well hidden that you can't even imagine a device being involved with the production of the information is a pretty good sign that the FDA is in the right here.
I've said this before, but 23andMe played this pretty much the worst way possible. It's not an undue burden to show that their medical results are scientifically valid. And instead of meeting this modest burden for scientific honesty, 23andMe tries to play it like the FDA is trying to keep information out of patients' hands. I think that's kind of shady.
And what does "information" constitute in this setting? What got 23andMe in trouble was making risk calls off of SNPs, usually by reading studies that report GWAS results and have come up with a model for the risk. The problem with this whole endeavor is that it has been repeatedly shown to be a near fool's game: due to the huge number of SNPs tested, the ones that show the best association with risk are quite often more likely to be noise than the real causal SNPs. The field has dealt with the problem, but it takes careful review of the literature to make sure that somebody like 23andMe is doing it correctly. And it doesn't need to be reviewed for one disease, it needs to be reviewed for every single call that 23andMe makes. That 23andMe is now willing to submit to this type of external validation is a good sign that 23andMe might be doing something useful; when they rejected this type of analysis it was a good sign that 23andMe was doing something shady.
We're not talking about 6TB hard drives here, where you're bound to write a bad bit when you fill the thing up, and an mp3 file gets corrupted. We're talking about knowing what the error rates are in 23andMe's SNP calling process (and there is always a non-zero error rate), knowing when an individual's analysis has gone wrong and needs to be redone, and what the scientific basis of their medical "information" consists of.
I don't doubt that 23andMe has been doing these things all along, but they need to open up and show that they are doing it. That's just good science.
posted by Llama-Lime at 10:39 PM on October 24, 2014 [24 favorites]
"What's so scary and/or pernicious about it? Is that I'm paying them and yet they're using my data when, perhaps, according to some views, I should be compensated for that use instead? Is there a fear that they'll share the data with insurance companies and/or government entities who will use it to personally discriminate against me?"Its about who owns the information, who gets to use it, for what purpose and at what price.
Genetic data is an entirely new kind of information for us to generate and disseminate. Before the early 90s there was a solid consensus that the data we generate should be belong to mankind and be used only in service to all of us, held and used in trust by the kinds of academics capable of creating it. That is, until Craig Venter fucked that up the first time by building the first generation of sequencing machines capable of doing what dozens of post-docs couldn't in a fraction of the time in the context of a for profit company. He wasn't working in the interest of mankind, at least not directly, and the data he generated didn't belong to anyone but him. Thankfully, at least at the time, there wasn't really a way to gate access to the data he generated and keep it valuable - so it was as worthless to him as it was invaluable to mankind - and thus hopefully he didn't get too good of a price from Bill Clinton and the US federal government when he held it for ransom. We have GenBank and the NCBI and the last 20 years of amazing and profound progress because the people of the United States bought it for the benefit of us all and broke down the gates.
Now, again, we have these venture capital fuckers going around and talking about their data in really creepy ways that indicate they don't really understand it, don't really care about its real value, don't really care about its potential for harm, don't really care about misrepresenting its value, and don't really care about us. Genetic data has no business being owned by people who want to make us consumers of it, if it is generated at all, it should belong to all of us.
posted by Blasdelb at 3:28 AM on October 25, 2014 [17 favorites]
" Times are changing and it's time some things did. I think my medical journey should be up to me as much as possible."As hard as these assholes are trying to obfuscate the issue, we're not really talking about withholding information from you, we're talking about preventing them from giving you shit information that they cannot validate no matter how much they wave their hands. The kinds of data they we're able to generate would require the kinds of deep expert level contextual understanding and bioinformatic skills that most MDs do not have to get anything out of. It is not reasonable to expect you to be able to perform your own analysis of the results of a SNP array, with statistical rigor and review level awareness of biological context, and they've been banned from their bullshit because they've been unable to show that their own analysis means a goddamn thing.
If we're going to put the new tools of molecular biology into the hands of the public we should at least make sure that the tools fit the hands that the public has available. This was never actually the goal of these arrogant and transparently incompetent corporate fucks, their goal had always been to generate a data set that should belong to all of us in order to be as evil as they possibly can with it.
posted by Blasdelb at 6:41 AM on October 25, 2014 [7 favorites]
You guys do realize that, when you have a 23andMe account, you have to explicitly OPT IN to research use of your data, right? They directly force you to make the choice one way or the other, they do not assume your consent, and if you say no, you can still use the service at exactly the same price. I don't remember them mentioning that they charge the researchers for the data, but it's simply false to claim that they pass your data around without telling you.
If you want something to get hysterical about, if I remember correctly they did default you, or at least strongly guide you, into "find relatives" or something like that. That's actually a million times more likely to mess up your life by finding relatives you didn't want to know about. But of course they don't charge for that service separately. And I also seem to vaguely remember that they changed that default.
posted by Hizonner at 7:21 AM on October 25, 2014 [10 favorites]
If you want something to get hysterical about, if I remember correctly they did default you, or at least strongly guide you, into "find relatives" or something like that. That's actually a million times more likely to mess up your life by finding relatives you didn't want to know about. But of course they don't charge for that service separately. And I also seem to vaguely remember that they changed that default.
posted by Hizonner at 7:21 AM on October 25, 2014 [10 favorites]
Cult of the Amateur--interesting term. It's mixed bag. Allowing people to know they have cancer and home pregnancy tests--good things. People opting to not vaccinate their children because they "know" that these vaccines cause autism--bad thing.
Providing information is good, provided the information itself is good. This gang seems to have gone beyond just providing nucleotide sequences to making shaky risk assessments developing diseases. Definitely the kind of thing that I would like the FDA to prevent.
posted by haiku warrior at 7:44 AM on October 25, 2014
Providing information is good, provided the information itself is good. This gang seems to have gone beyond just providing nucleotide sequences to making shaky risk assessments developing diseases. Definitely the kind of thing that I would like the FDA to prevent.
posted by haiku warrior at 7:44 AM on October 25, 2014
If I decide I think I need a certain lab test and I'm willing to pay for it, if I can go to a lab and get the test done without having to ring up $200 in bills and physical and mental exhaustion trying to get my physician to agree to the test, that's a good thing. If the test comes back positive I can then go to my doc with the test results and he can decide if he wants to retest me through a different lab or pursue some other angle instead; if the test comes back negative, then my worry is mostly alleviated. If the symptoms later escalate, I can then discuss being retested with my doctor - or maybe a different test - whatever he thinks.
It's a probability issue. Others have mentioned the BRCA mutation upthread. There is a reason that women are only recommended to get tested if they have certain risk factors--a family history of ovarian or breast cancer at a young age, Ashkenazi Jewish background, etc-- and that reason is (1) false positives (if the test is 99.99% accurate, if you do the test a million times, there's 100 people who got the wrong results) and (2) the fact that having a BRCA mutation doesn't necessarily mean that you'll get breast or ovarian cancer. So, screening everyone in the country for BRCA would likely cause more health problems than it would solve.
And the "Oh, the doctors further down the line will make sure nothing bad will happen" deal? There's all sorts of evidence piling up that we're over screening for certain types of cancers now (specifically, breast and prostate) and that's having a bad effect: people are getting operations and treatment for tumors and pre-cancers that in all likelihood, will not cause problems further down the line. And any time people in the know suggests pulling back on screening-- say, starting regular mammograms at 50 instead of 40-- it's met with a huge backlash. Things like 23andMe will just make this sort of situation worse.
posted by damayanti at 8:20 AM on October 25, 2014 [3 favorites]
It's a probability issue. Others have mentioned the BRCA mutation upthread. There is a reason that women are only recommended to get tested if they have certain risk factors--a family history of ovarian or breast cancer at a young age, Ashkenazi Jewish background, etc-- and that reason is (1) false positives (if the test is 99.99% accurate, if you do the test a million times, there's 100 people who got the wrong results) and (2) the fact that having a BRCA mutation doesn't necessarily mean that you'll get breast or ovarian cancer. So, screening everyone in the country for BRCA would likely cause more health problems than it would solve.
And the "Oh, the doctors further down the line will make sure nothing bad will happen" deal? There's all sorts of evidence piling up that we're over screening for certain types of cancers now (specifically, breast and prostate) and that's having a bad effect: people are getting operations and treatment for tumors and pre-cancers that in all likelihood, will not cause problems further down the line. And any time people in the know suggests pulling back on screening-- say, starting regular mammograms at 50 instead of 40-- it's met with a huge backlash. Things like 23andMe will just make this sort of situation worse.
posted by damayanti at 8:20 AM on October 25, 2014 [3 favorites]
They're not doing any sequencing. Wait about 10 years for that one.This is going to come much faster than you think. 23andMe started at something like $500, and whole genome sequencing will be pretty close to that very soon. The raw cost is ~$1000 at most centers that have their new x10 from Illumina, but most places are charging around $1500. If you have a rare mendelian disease, this is a fantastic deal. In a year or two, should any challengers to Illumina appear, sequencing will be a couple hundred dollars, or about the same cost as the hard drive it will take to store the sequence data.
If you just want to look at your DNA, maybe spending $1500 today for whole genome sequencing is not such a great deal. Because we really don't know what it means, unless you have a really strong phenotype that's likely caused by gene dysfunction, like muscular dystrophy or retinitis pigmentosa. And if do have a really strong phenotype likely caused by gene dysfunction, you don't want a SNP chip like 23andMe, you want a whole exome or whole genome sequence.
23andMe is great at the hype, great at the marketing and the PR, which is the true strength of far too many Silicon Valley companies these days. But I don't really see much general use for their product, beyond early adopters. It's like a ZX80 or Apple II, without the programmability. There's remarkably little that a smart amateur can do with this data, just as there's remarkably little that a smart professional can either. Aggregation is the best hope to make it useful, but even them, I feel a gigantic meh. The best possible hope for SNP data to become useful is aggregating everybody's full medical records (and good luck getting those in an interchangable format!!!), which even then, probably not that useful.
deCODE has done this for all of Iceland. They not only have all the SNPs, they now have pretty much all the full genomic sequence by sequencing just a fraction of the population of Iceland and then using Iceland's family tree to impute the non-sequeneced individuals' genome. This is really kind of an amazing dataset, and they churn out papers about various things now and then. But it has not been a panacea. It has not disrupted health care. And there's little reason to think it should. Pharma and others pay to see all of Iceland's genotypes, try to figure out how to maybe go after drugs for addiction or whatever, and it has lots of real benefits, but it's facilitating the slow and gradual advancement of knowledge and biotech, it's not a revolution.
Honestly, deCODE makes 23andMe look like an antiquated snake-oil salesman, and 23andMe is just the best marketed of many many nearly identical consumer SNP chip data providers. Yet how many people here have even heard of deCODE? How can Anne Wojcicki give an interview and talk about disrupting health care with SNP chips when any knowledgable observer knows that's not true? Unlike most of MetaFilter, I'm friendly to talk about disrupting industries, and friendly to Silicon Valley, and really think that technology has the potential to change everybody's lives for the better, and that most of Silicon Valley is focussed towards that end. But I see 23andMe and think that the marketing is just way too far ahead of reality. And unlike pure computer tech, the reality of biology takes a really long time to unfold, and it's not like the biology is going to just all of a sudden catch up to the marketing hype.
All this is to say, after all the extreme skepticism I see on MetaFilter about anything Silicon Valley related, I'm thinking that this is an awfully undeserving company for so many people to find interesting and worthwhile.
posted by Llama-Lime at 9:18 AM on October 25, 2014 [19 favorites]
I suppose then there is no way at present, other than having the skill and equipment to do so yourself, to actually get the ~200GB of data comprising the entire genome in a truly anonymous way, or for that matter, knowing for sure that the data you have is actually your data without having it done again by someone else and compare it. (data size info found here, including all the short-reads and quality scores, and most precise ~715MB genome file possible)
I'd like to have a copy of my entire genome just out of my own curiosity, and though I would be at least a bit concerned about who would have access to it or the knowledge I had the sequence done. As for what I would 'do' with it, granted there is not much now other than saying I have a very accurate self-portrait, but as time goes on, there is all sorts of things that the data could be used for, from artistic projects, to my own education, to possibly being useful to my own health at some point.
posted by chambers at 10:37 AM on October 25, 2014 [1 favorite]
I'd like to have a copy of my entire genome just out of my own curiosity, and though I would be at least a bit concerned about who would have access to it or the knowledge I had the sequence done. As for what I would 'do' with it, granted there is not much now other than saying I have a very accurate self-portrait, but as time goes on, there is all sorts of things that the data could be used for, from artistic projects, to my own education, to possibly being useful to my own health at some point.
posted by chambers at 10:37 AM on October 25, 2014 [1 favorite]
People make choices that have deleterious long-term consequences all the time. I mean, people smoke cigarettes, for the love of god, which—unless you are actually standing in front of a firing squad—is not regarded as a particularly good long-term/short-term tradeoff.
And yet, most people basically agree that's a personal decision, at least as long as you aren't subjecting other people to it. So I'm pretty suspicious of the whole "we can't let people know their genetic results because they might do something regrettable" canard, given that we let people do regrettable things all the goddamn time. And it seems suspiciously convenient that in this particular case, not letting them do this particular, entirely hypothetical, potentially-regrettable thing happens to help a lot of multibillion-dollar biotech companies who desperately want to keep genetic testing Rx-only, and therefore expensive.
I am not entirely in love with 23andMe, but I'll take them over any traditional biotech or pharma company, with their insulting paternalism, offensively high prices, and lets-not-call-it-corruption regulatory capture, any day of the week.
posted by Kadin2048 at 10:50 AM on October 25, 2014 [2 favorites]
And yet, most people basically agree that's a personal decision, at least as long as you aren't subjecting other people to it. So I'm pretty suspicious of the whole "we can't let people know their genetic results because they might do something regrettable" canard, given that we let people do regrettable things all the goddamn time. And it seems suspiciously convenient that in this particular case, not letting them do this particular, entirely hypothetical, potentially-regrettable thing happens to help a lot of multibillion-dollar biotech companies who desperately want to keep genetic testing Rx-only, and therefore expensive.
I am not entirely in love with 23andMe, but I'll take them over any traditional biotech or pharma company, with their insulting paternalism, offensively high prices, and lets-not-call-it-corruption regulatory capture, any day of the week.
posted by Kadin2048 at 10:50 AM on October 25, 2014 [2 favorites]
Every test we have, every diagnosis we get, is being fed into the massive database at one level or another that will one day have all our medical data available to every insurance company and medical practice - there's no way that's ever going to be stopped now that we have the electronic medical records
I suspect significant liabilities for unauthorized disclosure -- including accidents -- might do something to make data holders think twice about possessing data.
Not that we'll probably do that, for the same reasons that we don't do all kinds of sensible privacy related things, but it's not that there's no way.
Also, let's be clear about one thing: if there isn't a way to stop it being available to every insurance company and medical practice, then in practice there's no way to stop it being available to anyone at all who wants it.
posted by weston at 11:22 AM on October 25, 2014
I suspect significant liabilities for unauthorized disclosure -- including accidents -- might do something to make data holders think twice about possessing data.
Not that we'll probably do that, for the same reasons that we don't do all kinds of sensible privacy related things, but it's not that there's no way.
Also, let's be clear about one thing: if there isn't a way to stop it being available to every insurance company and medical practice, then in practice there's no way to stop it being available to anyone at all who wants it.
posted by weston at 11:22 AM on October 25, 2014
People make choices that have deleterious long-term consequences all the time. I mean, people smoke cigarettes, for the love of god, which—unless you are actually standing in front of a firing squad—is not regarded as a particularly good long-term/short-term tradeoff. And yet, most people basically agree that's a personal decision, at least as long as you aren't subjecting other people to it.
Sure, we as a society say "Sure, smoke if you want", but we have also made it difficult either buy or sell cigarettes- you have to be of a certain age, there are special taxes on them, they're required to have special warnings on them about the consequences of smoking, and the government runs extensive advertising and in-school programs talking about the dangers, places all over are implementing smoking bans etc., etc.
The same thing is being done with genetic testing: if 23andMe wants to sell this potentially dangerous product, it needs to do so under certain regulations given to it by the FDA. And I think it needs to be done because I think while most people know that smoking is bad, most people don't know that screening for diseases can have negative consequences: see the outcry over attempting to change the recommended age of first mammograms.
posted by damayanti at 11:47 AM on October 25, 2014 [1 favorite]
Sure, we as a society say "Sure, smoke if you want", but we have also made it difficult either buy or sell cigarettes- you have to be of a certain age, there are special taxes on them, they're required to have special warnings on them about the consequences of smoking, and the government runs extensive advertising and in-school programs talking about the dangers, places all over are implementing smoking bans etc., etc.
The same thing is being done with genetic testing: if 23andMe wants to sell this potentially dangerous product, it needs to do so under certain regulations given to it by the FDA. And I think it needs to be done because I think while most people know that smoking is bad, most people don't know that screening for diseases can have negative consequences: see the outcry over attempting to change the recommended age of first mammograms.
posted by damayanti at 11:47 AM on October 25, 2014 [1 favorite]
So I'm pretty suspicious of the whole "we can't let people know their genetic results because they might do something regrettable" canard, given that we let people do regrettable things all the goddamn time. And it seems suspiciously convenient that in this particular case, not letting them do this particular, entirely hypothetical, potentially-regrettable thing happens to help a lot of multibillion-dollar biotech companies who desperately want to keep genetic testing Rx-only, and therefore expensive.
The problem is less that 23andMe is giving out information 'they don't want you to have" and more that 23andMe's *risk assessments* are very shaky, and perhaps constitute *bad* information. SNP, from what Blasedelb, Llama-Lime, and a few others have posted, doesn't seem to be a terribly thorough or informative type of genetic "testing" for most medical purposes. And you'll note that both of them have pointed to the likelihood that in-depth screening will become cheap and available soon...it's just that 23andMe isn't really offering that.
For that matter, if the risk assessments 23andMe offers are misleading and suggest higher levels of risk for some conditions than actually exist, then the fact that they can sell the data sets is not exactly going to improve the multibillion-dollar insurance, biotech, and medical industries. For instance, the tendency of SNP assessments to superficially exaggerate the risk of breast cancer not going to reduce the problem of cancer overscreening mentioned by damayanti, nor that of unnecessary surgeries. Both of those give more money, not less, to those industries of which you're so skeptical, and at the cost of unnecessary, additional risks to health care consumers.
posted by kewb at 11:51 AM on October 25, 2014 [4 favorites]
The problem is less that 23andMe is giving out information 'they don't want you to have" and more that 23andMe's *risk assessments* are very shaky, and perhaps constitute *bad* information. SNP, from what Blasedelb, Llama-Lime, and a few others have posted, doesn't seem to be a terribly thorough or informative type of genetic "testing" for most medical purposes. And you'll note that both of them have pointed to the likelihood that in-depth screening will become cheap and available soon...it's just that 23andMe isn't really offering that.
For that matter, if the risk assessments 23andMe offers are misleading and suggest higher levels of risk for some conditions than actually exist, then the fact that they can sell the data sets is not exactly going to improve the multibillion-dollar insurance, biotech, and medical industries. For instance, the tendency of SNP assessments to superficially exaggerate the risk of breast cancer not going to reduce the problem of cancer overscreening mentioned by damayanti, nor that of unnecessary surgeries. Both of those give more money, not less, to those industries of which you're so skeptical, and at the cost of unnecessary, additional risks to health care consumers.
posted by kewb at 11:51 AM on October 25, 2014 [4 favorites]
Be wary of people who get hyperbolic about privacy violations that don't appear to happen, while maintaining at the same time, somehow, that the scientific data that is being passed on to unnamed wealthy, powerful entities is also so effectively poor as to be useless.
posted by a lungful of dragon at 11:54 AM on October 25, 2014 [1 favorite]
posted by a lungful of dragon at 11:54 AM on October 25, 2014 [1 favorite]
Those of you am who are concerned bout the medical judgement 23andme is making about the meaning/risk factor of SNPs, I'm curious how you view promethease and SNPedia with it's open source structure? All the information is based off studies, but anyone can modify the meaning of different SNPs.
posted by [insert clever name here] at 12:47 PM on October 25, 2014
posted by [insert clever name here] at 12:47 PM on October 25, 2014
It's interesting to contrast the prevalent public view of genetic privacy with the view of genetic privacy that is emerging from the Global Alliance for Genomics and Health, a consortium that hopes to be a standards body for all things genomic. In particular, the rare disease advocacy groups in the GA4GH have more fear that genetic data will hidden in silos than that genetic data may escape.
In response, the GA4GH is motivating their patient consent form tools from completely different direction than traditional research consent. Instead of trying to restrict what can be done with the data to protect people from potential harm, it is guided by:
Also, it seems that nearly all diagnostic and lab test providers these days provide the same data to the patient as they do to the provider. Thus, I think that the days of medical paternalism in this regard are pretty much done. This will continue into the genomic era, though lugging around a 750MB VCF file is not a trivial thing to do with our country's weak Internet infrastructure. (At least, the VCF files that we produce are around 750MB, instead of the 125MB from chambers' article, because we add much more annotation about the effect of particular genetic variants.)
posted by Llama-Lime at 12:48 PM on October 25, 2014 [1 favorite]
In response, the GA4GH is motivating their patient consent form tools from completely different direction than traditional research consent. Instead of trying to restrict what can be done with the data to protect people from potential harm, it is guided by:
... Article 27 of the 1948 Universal Declaration of Human Rights. Article 27 guarantees the rights of every individual in the world “to share in scientific advancement and its benefits” (including to freely engage in responsible scientific inquiry), and at the same time “to the protection of the moral and material interests resulting from any scientific...production of which [a person] is the author.”I think this change of view has great potential to improve access to research data. Right now, data is hidden behind access committees that make sure that genetic data in research studies is only viewed by those with proper credentials and research goals. If some smart teenager in China wants to try new machine learning methods, they won't have the chance, unless more people consent to have their data viewed more widely.
Also, it seems that nearly all diagnostic and lab test providers these days provide the same data to the patient as they do to the provider. Thus, I think that the days of medical paternalism in this regard are pretty much done. This will continue into the genomic era, though lugging around a 750MB VCF file is not a trivial thing to do with our country's weak Internet infrastructure. (At least, the VCF files that we produce are around 750MB, instead of the 125MB from chambers' article, because we add much more annotation about the effect of particular genetic variants.)
posted by Llama-Lime at 12:48 PM on October 25, 2014 [1 favorite]
The ethical standards of the 23AndMe business model are far, far superior to those of the business model of almost any medical center or physician you might choose to go to.
At the moment their ethics may be superior. But when they get bought by some other globo corp will they still be?
The issue I have is that people are both providing samples and paying for the analysis (with medical waste it is highly unlikely that anyone is bothering).
Say the Republicans successfully dismantle Obamacare and pre-existing conditions once again become something insurance companies can use. How much would access to a patient's genetic susceptibility be worth to a either an actuarial or claim denial department? I think probably enough to tempt any company (or even more likely an employee within the company who can steal the data).
Then you are at least doubly fucked by your own helix.
posted by srboisvert at 12:51 PM on October 25, 2014 [1 favorite]
At the moment their ethics may be superior. But when they get bought by some other globo corp will they still be?
The issue I have is that people are both providing samples and paying for the analysis (with medical waste it is highly unlikely that anyone is bothering).
Say the Republicans successfully dismantle Obamacare and pre-existing conditions once again become something insurance companies can use. How much would access to a patient's genetic susceptibility be worth to a either an actuarial or claim denial department? I think probably enough to tempt any company (or even more likely an employee within the company who can steal the data).
Then you are at least doubly fucked by your own helix.
posted by srboisvert at 12:51 PM on October 25, 2014 [1 favorite]
"Be wary of people who get hyperbolic about privacy violations that don't appear to happen, while maintaining at the same time, somehow, that the scientific data that is being passed on to unnamed wealthy, powerful entities is also so effectively poor as to be useless."Yes the conspiracy, it seems, extends to metafilter's molecular biologists!
More seriously though, wealthy and powerful people are only rarely especially intelligent and basically never actually experts in things that are genuinely complicated like molecular genetics or bioinformatics. As an embarrassing portion of synthetic biology have discovered, there seems to be a lot of money to be made in convincing people they understand molecular biology without the hassle of actually learning anything and these guys are masters of the con. The only place they slipped up is in how it's illegal to sell half assed bullshit and call it medical advice. There are certainly interesting questions ordinary people can ask of the kinds of data they provide, and get valid enough answers for, but they are manifestly incapable of 'disrupting' anything other than idle curiosity of people with more money than family tree.
SNP diagnostic and screening analysis is certainly capable of a lot more value than it currently has, but creating that value would involve the kinds of slow, expensive, and not really disrupt-able molecular biology that only the Feds and charities invest in, and doing that now would be like investing in cassette music in the mid 90s.
posted by Blasdelb at 2:18 PM on October 25, 2014
The ethical standards of the 23AndMe business model are far, far superior to those of the business model of almost any medical center or physician you might choose to go to.
Citation needed.
posted by Existential Dread at 2:50 PM on October 25, 2014
Citation needed.
posted by Existential Dread at 2:50 PM on October 25, 2014
Which is to say, that's an awfully broad swathe of doctors, nurses, and hospitals to tar with the brush of tissue banking.
posted by Existential Dread at 3:01 PM on October 25, 2014 [3 favorites]
posted by Existential Dread at 3:01 PM on October 25, 2014 [3 favorites]
It's mixed bag. Allowing people to know they have cancer and home pregnancy tests--good things. People opting to not vaccinate their children because they "know" that these vaccines cause autism--bad thing.
Those severely misguided people who are not vaccinating their kids would only act in even more irrational ways if they suspected information was being withheld from them. What is the alternative? Just surprise them at the doctor's office with unnamed syringes of "something or other don't worry it's good for you"? The only way to battle bad information is with good information. But even then, there will be people who make the bad choices. Unfortunately, it has to be their right to make that bad choice. It would be oppressive to live in a world where someone else had the ultimate say about what happened to your body.
posted by the jam at 3:21 PM on October 25, 2014 [2 favorites]
Those severely misguided people who are not vaccinating their kids would only act in even more irrational ways if they suspected information was being withheld from them. What is the alternative? Just surprise them at the doctor's office with unnamed syringes of "something or other don't worry it's good for you"? The only way to battle bad information is with good information. But even then, there will be people who make the bad choices. Unfortunately, it has to be their right to make that bad choice. It would be oppressive to live in a world where someone else had the ultimate say about what happened to your body.
posted by the jam at 3:21 PM on October 25, 2014 [2 favorites]
And it seems suspiciously convenient that in this particular case, not letting them do this particular, entirely hypothetical, potentially-regrettable thing happens to help a lot of multibillion-dollar biotech companies who desperately want to keep genetic testing Rx-only, and therefore expensive.
When my friend's daughter Lisa had some bleeding from her breast about four years ago (she's in her late 40s now), tests didn't show anything so they decided to do the "wait and see" thing. In the meantime, she inquired about having a BRCA test done since there was some fairly remote breast cancer in the family; the doc said it really wasn't necessary at that time, but she checked into the cost of having it done herself - it turned out that, at $5000, it was just to expensive to have the test run. There's no been no evidence of breast cancer and her problem cleared up, but my point is that the test cost $5000 to have done. Now go to the Inspire Breast Cancer forum or the Inspire Ovarian Cancer forum and read what people who have those cancers have to say. Just as a side note here - almost without fail, women with ovarian cancer aren't diagnosed until they're at a very late stage and it's almost always without any hope of cure.
It's a probability issue.
Oh, me. No it's not - it's a human issue involving very real persons who have every right to be 100% involved and invested in their own health care - which involves more than "healthy eating and no smoking," persons who own the right to learn and know exactly what the condition of their own body is right now so they can make better decisions about what to do with it in the future.
posted by aryma at 4:26 PM on October 25, 2014 [4 favorites]
When my friend's daughter Lisa had some bleeding from her breast about four years ago (she's in her late 40s now), tests didn't show anything so they decided to do the "wait and see" thing. In the meantime, she inquired about having a BRCA test done since there was some fairly remote breast cancer in the family; the doc said it really wasn't necessary at that time, but she checked into the cost of having it done herself - it turned out that, at $5000, it was just to expensive to have the test run. There's no been no evidence of breast cancer and her problem cleared up, but my point is that the test cost $5000 to have done. Now go to the Inspire Breast Cancer forum or the Inspire Ovarian Cancer forum and read what people who have those cancers have to say. Just as a side note here - almost without fail, women with ovarian cancer aren't diagnosed until they're at a very late stage and it's almost always without any hope of cure.
It's a probability issue.
Oh, me. No it's not - it's a human issue involving very real persons who have every right to be 100% involved and invested in their own health care - which involves more than "healthy eating and no smoking," persons who own the right to learn and know exactly what the condition of their own body is right now so they can make better decisions about what to do with it in the future.
posted by aryma at 4:26 PM on October 25, 2014 [4 favorites]
As hard as these assholes are trying to obfuscate the issue, we're not really talking about withholding information from you, we're talking about preventing them from giving you shit information that they cannot validate no matter how much they wave their hands. [...] their goal had always been to generate a data set that should belong to all of us in order to be as evil as they possibly can with it.
Look, you can't have it both ways. Either their data is useless or they're trying to be as evil as possible. Unless you think they're purposefully trying to be evil with useless data.
I'd wager that, at very least, they believe their science is sound. Why else do you think they'd go through the rigmarole of seeking FDA approval?
posted by evil otto at 2:01 AM on October 26, 2014
Look, you can't have it both ways. Either their data is useless or they're trying to be as evil as possible. Unless you think they're purposefully trying to be evil with useless data.
I'd wager that, at very least, they believe their science is sound. Why else do you think they'd go through the rigmarole of seeking FDA approval?
posted by evil otto at 2:01 AM on October 26, 2014
IndigoJones: "I don't know what the right balance is, but I've come to the point of considering doctors guilty until proven innocent."
Yeah, I feel the same way about startups: it's a morally bankrupt culture designed to tear the world apart to make a buck, all the while telling you it's saving you by giving you "information."
evil otto: "Look, you can't have it both ways. Either their data is useless or they're trying to be as evil as possible. Unless you think they're purposefully trying to be evil with useless data."
Nobody said they're "trying to be as evil as possible," which is kind of a silly characterization. The claim is that they're not terribly concerned about privacy - which seems true from everything I've seen. I mean, come on - they're funded by venture capital and Google. Odds are pretty darned good they don't give a shit about anybody except insofar as they can make a buck from them.
These things are not mutually exclusive. They could be incautious about privacy AND bad with handling the data they have. That combination isn't self-negating - in fact, it makes all of this worse.
"I'd wager that, at very least, they believe their science is sound. Why else do you think they'd go through the rigmarole of seeking FDA approval?"
Come on, they actually answered this one right out: because they were forced to legally. they actually said explicitly that they would never have sought FDA approval if they hadn't been forced to. Besides, who knows how hard it is to pass FDA approval? 23AndMe isn't the first company to try to make money from medical testing. I would bet there are ways through that that will allow them to keep giving tests by renaming them "recommendations" or using the well-developed dialect of marketing statistics ("chances are good that..." etc) - lord knows they're already milking that for all it's worth.
posted by koeselitz at 2:59 AM on October 26, 2014 [2 favorites]
Yeah, I feel the same way about startups: it's a morally bankrupt culture designed to tear the world apart to make a buck, all the while telling you it's saving you by giving you "information."
evil otto: "Look, you can't have it both ways. Either their data is useless or they're trying to be as evil as possible. Unless you think they're purposefully trying to be evil with useless data."
Nobody said they're "trying to be as evil as possible," which is kind of a silly characterization. The claim is that they're not terribly concerned about privacy - which seems true from everything I've seen. I mean, come on - they're funded by venture capital and Google. Odds are pretty darned good they don't give a shit about anybody except insofar as they can make a buck from them.
These things are not mutually exclusive. They could be incautious about privacy AND bad with handling the data they have. That combination isn't self-negating - in fact, it makes all of this worse.
"I'd wager that, at very least, they believe their science is sound. Why else do you think they'd go through the rigmarole of seeking FDA approval?"
Come on, they actually answered this one right out: because they were forced to legally. they actually said explicitly that they would never have sought FDA approval if they hadn't been forced to. Besides, who knows how hard it is to pass FDA approval? 23AndMe isn't the first company to try to make money from medical testing. I would bet there are ways through that that will allow them to keep giving tests by renaming them "recommendations" or using the well-developed dialect of marketing statistics ("chances are good that..." etc) - lord knows they're already milking that for all it's worth.
posted by koeselitz at 2:59 AM on October 26, 2014 [2 favorites]
aryma: "Just as a side note here - almost without fail, women with ovarian cancer aren't diagnosed until they're at a very late stage and it's almost always without any hope of cure."
Well then - let's make actual medical tests easier and more available to people - which is something our society already does for routine, well-established stuff like blood tests. We do that because they are actual medical tests. There are laws about how those tests are done, rules and regulations about consent and approval and privacy, but that doesn't stop me from walking into a phlebotomist's office and having a blood test any time I want.
We aren't living in Galt's Gulch here. There are rules about these things - not to keep people from them, but because they are important.
If 23AndMe wants to offer medical tests, let them get the approval to do so and go through the regulatory hassles just like everybody else. Those regulatory hassles are awesome; they are the reason I can get a blood test any time I want and trust the results to be correct and private. The sooner companies like this get in line with actual standards, the sooner we'll have available tests we can actually rely on.
posted by koeselitz at 3:14 AM on October 26, 2014 [2 favorites]
Well then - let's make actual medical tests easier and more available to people - which is something our society already does for routine, well-established stuff like blood tests. We do that because they are actual medical tests. There are laws about how those tests are done, rules and regulations about consent and approval and privacy, but that doesn't stop me from walking into a phlebotomist's office and having a blood test any time I want.
We aren't living in Galt's Gulch here. There are rules about these things - not to keep people from them, but because they are important.
If 23AndMe wants to offer medical tests, let them get the approval to do so and go through the regulatory hassles just like everybody else. Those regulatory hassles are awesome; they are the reason I can get a blood test any time I want and trust the results to be correct and private. The sooner companies like this get in line with actual standards, the sooner we'll have available tests we can actually rely on.
posted by koeselitz at 3:14 AM on October 26, 2014 [2 favorites]
Nobody said they're "trying to be as evil as possible," which is kind of a silly characterization.
I was quoting Blasdelb directly.
Come on, they actually answered this one right out: because they were forced to legally.
And you think they'd even bother if they didn't think they could succeed?
posted by evil otto at 3:49 AM on October 26, 2014
I was quoting Blasdelb directly.
Come on, they actually answered this one right out: because they were forced to legally.
And you think they'd even bother if they didn't think they could succeed?
posted by evil otto at 3:49 AM on October 26, 2014
These things are not mutually exclusive. They could be incautious about privacy AND bad with handling the data they have.
It strikes me that a very good analogy for this would be the personal credit ratings industry. The ratings are very buggy and often inaccurate, to the detriment of many people, and the companies make lots of money selling this often inaccurate credit data to the financial industry as well as by getting individuals to buy their own credit reports. None of this is malign conspiracy, just carelessness in the service of profit.
posted by kewb at 6:09 AM on October 26, 2014
It strikes me that a very good analogy for this would be the personal credit ratings industry. The ratings are very buggy and often inaccurate, to the detriment of many people, and the companies make lots of money selling this often inaccurate credit data to the financial industry as well as by getting individuals to buy their own credit reports. None of this is malign conspiracy, just carelessness in the service of profit.
posted by kewb at 6:09 AM on October 26, 2014
Kewb - not entirely careless. Credit reporting is to protect lenders from people with bad credit. The product is "good credit" ratings. Those are badically reliable.
When credit reporting gives J.Q. Public a bad credit in error, only J.Q. is hurt.
More than carelessness.
posted by Lesser Shrew at 6:46 AM on October 26, 2014 [2 favorites]
When credit reporting gives J.Q. Public a bad credit in error, only J.Q. is hurt.
More than carelessness.
posted by Lesser Shrew at 6:46 AM on October 26, 2014 [2 favorites]
« Older Anarchy in the Pre-K | I love tortoises, and I love to crochet. Newer »
This thread has been archived and is closed to new comments
I think I felt my brain trying to flee when I read this. This is the problem with the cult of the amateur, folks.
posted by NoxAeternum at 3:52 PM on October 24, 2014 [11 favorites]