Yesterday was National DNA Day. In addition, 23andMe decided to hold a massive sale on their “complete” scan for $99.
That in and of itself is not so exciting to me. What is, however, is the ensuing commentary that I found on SlickDeals. It is rare you see such a massive discussion with hundreds of contributors from the general public commenting about the risks/benefits/limitations of 23andMe in such a public forum.
Of course there’s the general trolling “I got my results back last week, I’m half German shepherd! Thanks, 23andme!”, and the people who are just excited by the price “OMG! This is a really hot hot deal!”.
But then you get some deep discussion about the value of the information, such as this snippet from poster travfar:
“The poster’s comment was that a false positive/negative would make someone worry about something needlessly for the rest of their lives. Telling people that smoking will give them lung cancer is the same thing. The message that goes out is that if you smoke, you’ll get lung cancer. But the reality is that if you smoke, there’s an 80% chance you won’t get lung cancer. Life causes lung cancer. It happens to nonsmokers as well.”
Go take a look, mine through the pages. This would contribute amazingly to someone interested in looking at public perceptions of DTC, in a format that is free from study bias.
Gizmodo has some interesting commentary too!
oi, life has been busy indeed. as i delay my packing for the American College of Medical Genetics meeting in Albuquerque yet another hour past midnight, here is something i have been working on that has been particularly interesting to me: risk communication in public health. as of my public health courses, i of course bring in my knowledge and interests around genetic testing and emerging genetic technologies into play. here is a sample of a short piece i wrote discussing the idea of emphasizing incremental risk…
Incremental Risk Communication and Common Genetic Variant Risks
The framing of numerical risk comparisons as incremental values when compared to showing risk figures as simply a comparison of absolutes risks has been show to better focus the target audience’s attention on specifically the change from baseline risk (Zikmund-Fisher B, Fagerlin A, Roberts A, Derry H, et al 2008). In the context of genetic risk communication and genetic counseling for many highly-penetrant and/or mendelian-inherited disorders, the risk of developing symptoms and risk of inheritance, respectively, garner considerably less utility in highlighting incremental risks given the often extreme differences that exist in these risks when compared to the general population. In stark contrast are those risks conferred to individuals when common genetic variants, such as those ascertained through many recent genome-wide association studies imparting minimal to modest risks, many with a relative risk under 2 (Eng C and Sharp R, 2010). Among a slew of complex ethical, legal, and social issues surrounding the reliability and clinical utility of communicating these common variant risk modifications to the public, an underlying methodological issue has been how a genetic counselor can actually communicate such small relative risk figures without misrepresenting the actual changes in baseline risks and increase in risk due to a variant. A potential solution lies in the communication of incremental risks in the methods that the authors of Zikmund-Fisher et al. 2008 demonstrate, focusing on comparison of baseline, absolute risks and the incremental changes attributed to presence of a common genetic variant.
In order to properly frame the utility of incremental risk in this scenario, a hypothetical discussion of an individual with a baseline heart disease risk of 14%, and a testable, validated common genetic variant will exist, with an relative risk of 1.3 affecting heart disease in individuals of this population, will be presented. One technique normally found in the genetic counselor’s repertoire in risk communication has been to state baseline risks, and then to provide the odds ratio of possessing the common variant in multiple formats. These multiple relative risk formats could be framed as “1.3 times greater chance of heart disease”, shown that the variant provides “30% more risk”, or provided as the absolute risk of an individual with the common variant to be 18.2% for heart disease. None of these techniques, however, highlight the “marginal increment in risk while simultaneously providing information about baseline risk level for context” (Zikmund-Fisher et al. 2008). Instead, framing the change in risk of an individual that possesses our proposed common variant through incremental risk, may highlight the change in risk more accurately without the need for more high numeracy-requiring processes of mental math: multiplying relative risks against baseline or subtracting absolute final risks from baseline.
In this scenario of highlighting incremental risk, the communicated risks would thus be provided as a baseline, absolute risk for heart disease of 14%, while individuals possessing the common variant experience an incremental absolute increase of 4.2% risk for heart disease, providing a total of 18.2% absolute risk. While the authors of Zikmund-Fisher et al. 2008 acknowledge that providing incremental risk altered the degree of worry in participants, without knowing what the appropriate degree of worry should be among these individuals, the focus on the specific change in absolute risk more accurately distinguishes for audiences the most important element of the risk communication.
Zikmund-Fisher BJ, Fagerlin A, Roberts TR, Derry HA, Ubel PA. Alternate methods of framing information about medication side effects: Incremental risk versus total risk of occurrence. Journal of Health Communication 2008;13(2):107-24.
here’s a funny video of Larry David’s experience with ancestry testing through the Ancestry by DNA test from DNA Diagnostics Center, as revealed on George Lopez. apparently their test uses 176 Ancestry Informative Markers, or AIMs. which ones these are and why they only test for European, Native American, Sub-Saharan African, and East Asian “BioGeographic Ancestry”, we don’t know without inquiring further. the references they cite are all from 2003 and before too..
Larry David is 63% European, 37% Native American.
“so I fail as a European?”
ain’t that interesting. it does make it seem a bit ridiculous/sensational, but funny indeed. their explanations of how these things work, however, aren’t the clearest thing:
also check out: DNA and Race FAQ via AncestrybyDNA
as a future clinician, tweets like this creep me out:
My 23andMe results are in: the worst thing I have to worry about is my 61% chance of Venuos Thromboembolism.
so is that.. a risk for superficial venous thromoboembolism? or risk for a deep venous thromboembolism (DVT)? or even a pulmonary embolism (PE)? of which the latter two are, i’d say, pretty serious…
its especially weird especially when the response by @23andme is:
meh. who am i to judge… lowly genetic counseling student.
So on my way to work today, driving across from 880 to Westbound 80, avoiding the Bay Bridge toll plaza, I saw a pair of chromosomes from my onramp in oakland that I knew immediately belonged to the 23andMe logo. Nevertheless, these were on a not-a-blimp! I tried to take a snapshot on my phone, but sadly, resolution wasn’t so great and the not-a-blimp was out of my driving trajectory, but a quick search online and lo and behold, there definitely is a 23andMe not-a-blimp!
Talk about aggressive marketing! This is taking the DTC part of personal genomics waaayyyy beyond what I had expected. Sure, trendy spit parties at fashion week that make the Times is one thing, but a not-a-blimp to harass SF commuters? C’mon. Personal genomics is not something that can be sold as a simple commodity. The procedures, risks, and implications of it can be serious, and by advertising the service as though it was a movie or typical commercial product is tricking those that are less genetically literate (unfamiliar in complex genetics) into thinking that this is just for fun.
According to Alex Hall, CEO of Airship Ventures, via ChubbyBrain
While blimps can also do advertising, the Zeppelin is 50ft larger than any of them and can provide unique promotional opportunities…
I can only speculate how much this costs in advertising. Sure, the previous contract was with Disney-Pixar promoting UP, very appropriate, especially in the Bay Area. But personal genomics?
Airship Ventures even has a time-lapse Youtube video they just posted of the application of the 23andMe logo that was uploaded yesterday.
After a bit more digging, and not to be offensive, there’s something strangely incestuous about the relationship between Google, 23andMe (which you already knew), and this Airship Ventures company. A relationship that involves founders, mothers, board members (i.e. 23andMe cofounder’ s mother, back in october of last year, writes a travelogue promoting Airship Ventures (links below). Airship Ventures’ major investor is 23andMe’s first listed board member, which for once, isn’t all that surprising… So that’s where this is coming from..
Apparently the difference between a Zeppelin and a Blimp is whether or not there is a rigid, internal structure. The former, which is the case mentioned here, is rigid, wherease a blimp is, well.. floppy ( that word’s for you, Beth).
you can track the zeppelin!!! in case you actively wanted to be visually spammed by the zeppelin….
via 23andme’s facebook page: which is.. having a contest for best picture of the zeppelin to get fifty-buck discounts? really, whatchu gonna do 23andme when its some 15 year old kid that wins the contest? oh yeah, they don’t have restrictions on age, despite established NSGC positions about childhood testing…
Prenatal and childhood testing for adult-onset genetic conditions should always include genetic education and counseling. Genetic counseling for clients considering such testing should include exploration of the psychological/social risks and benefits of early genetic identification from both the parents’ and child’s perspectives. When possible the child should be involved in the decision about whether or not to be tested. Other issues discussed should include the possibility of discrimination in insurance, education and employment for the child or family in both the immediate and more distant future.