Robin Hanson says that his students liked the idea of publishing all medical data, eschewing privacy concerns in exchange for the ability to better examine the impact of lifestyle, environment, genetics and health care in determining health outcomes.
For researchers this could mean vastly better analysis. For consumer, it might mean being able to accurately shop the best doctors, hospitals and most importantly, procedures.
Since, I was kid this seemed like a good idea to me. Indeed, my younger self always wondered why the government didn’t just give out free medical care in exchange for data. At the time I thought science was omnipotent and if we couldn’t cure a disease it must be because we didn’t have access to the right data. Public data I reasoned, would lead to virtual immortality.
I am not quite – quite – as naive as all that today. Nonetheless, I see the benefits as much greater than the costs. I strongly suspect that privacy concerns only really matter when its only one or a few people whose information is private.
Everyone has something “wrong” with them. Some embarrassing medical fact or potentially “damaging” information. Yet, potential employers, lovers, and lenders can’t discriminate against everyone.
If anything it seems that public medical information would work to erase the stigma associated with certain diseases. Can it be shameful that grandma is going senile when everyone knows that seven other women on the same street are too. Is it embarrassing to take Viagra when over half the men your age at your workplace are known to do so as well?
9 comments
Comments feed for this article
Wednesday ~ June 9th, 2010 at 2:58 pm
Steve Hamlin
Use of health data in the ways you suggest doesn’t require data records to be personally-identifiable at large.
Why should a researcher studying the impact between erectile dysfunction and cigarette smoking need to know that Bob Smith at 123 Main Street takes Viagra? Why should Bob’s neighbors?
The researcher only needs to know that #ABC123DE4 has attributes Viagra and Smoking. The neighbors don’t have a need to know anything. And Bob can already tell anyone he wants to about his health history.
If Electronic Medical Records become a national standard, there can be serious and effective privacy protections, while still allowing widespread research use of the anonymized health data.
Wednesday ~ June 9th, 2010 at 3:03 pm
Apex
“Everyone has something “wrong” with them. Some embarrassing medical fact or potentially “damaging” information. Yet, potential employers, lovers, and lenders can’t discriminate against everyone. ”
No just the ones who will cost them the most.
“Is it embarrassing to take Viagra when over half the men your age at your workplace are known to do so as well? ”
I would think it certainly would be. It is not very much less embarassing because others are doing it. People will still tell jokes and now they can name you specifically behind your back. We can all see that half of men are bald, it doesn’t stop many of them from doing everything they can to cover it up (wear hats) or to get hair treatment. They don’t seem to find comfort in their shared affliction, and baldness is purely cosmetic so it can’t have much to do with anything other than how you think others view you (and how you view yourself).
And publishing embarassing information will make those who appear to have it more rare because many people will simply refuse to get viagra if they know it will be published. And that will apply to some life threatening conditions as well.
I must confess that this view stricks me as not having improved much upon the earlier referenced niavety
Wednesday ~ June 9th, 2010 at 3:06 pm
Apex
@Steve,
Very good and sensible. I cannot imagine why universal personally identifiable medical information is necessary or even good and I can certainly imagine many ways in which it is not.
Wednesday ~ June 9th, 2010 at 4:04 pm
Roland
The problem of siloed health data is huge–and one of the big advantages in quality control enjoyed by large health systems (VA, Kaiser, British NHS) is due to their possession of huge amounts of data on every aspect of care. This can be done, as noted by others, without revealing specific identities.
Wednesday ~ June 9th, 2010 at 5:13 pm
Niklas Blanchard
Seems to me it would be very easy to have a coded database (as Steve said ABC123DE4), and then have a completely separate relational database that would store personally identifiable information.
Then all you would have to share with a hospital you visit is your PIN, and possibly a passcode.
Wednesday ~ June 9th, 2010 at 5:15 pm
Niklas Blanchard
and then have a number of completely separate relational databases that would store personally identifiable information.
Fixed. Each medical care facility you visit would presumably have their own database for this.
Wednesday ~ June 9th, 2010 at 5:56 pm
Niklas Blanchard
I should probably point out that my last couple posts deal directly with the issue of things that “seem easy” ;].
Thursday ~ June 10th, 2010 at 9:31 am
Robin Hanson
Interesting – I say to keep personal identification info private and most of my commenters say that doesn’t work. Karl says to reveal it alll, and most of his commenters say, no, keep the personal identification info private.
Thursday ~ June 10th, 2010 at 9:51 am
Ian Kanchay
It has a positive impact and I believe it is a question of “social construct”. People may very well say yes if proposed nevertheless I feel naked just thinking of it. Irrational? I prefer to see it as human. (I know for a fact that my grandmother would refuse.)