AMIA 2011 - Round 1: Data Sharing
This is the first in a series of posts I'll be doing to recount the recent AMIA 2011 Symposium in Washington D.C. What can I say - WOW! It was a week ago already that I arrived, and after a few days of having my brain filled I (needless to say) needed a few days to relax and decompress.
I began AMIA on Sunday night with a Late Breaking session about the Sorrell v. IMS Health case in Vermont where it was outlawed for pharmacies from selling prescription information to data mining and pharma companies. While I make it no secret the details of legal proceedings are often over my head (I initially attended to see a fellow OHSU classmate present), this was a great discussion and raised some important considerations about the sharing of data.
While many of the details of the discussion revolved around the legalities of how data miners were justifying the acquisition of the data under "Freedom of Speech", I found the discussion around how data could and should be used very interesting. My personal opinion is that we do need to continue the ability for researchers to have access to data. This doesn't necessarily mean full, open data available at will, but definitely access to proper data under proper institutional regulations. The ability for aggregating data can and has been shown to improve results, most directly in my experience through the eMERGE project where sites pooling data were able to increase power in genome wide association studies. This may not be the most directly applicable example, as the sites were a collaboratory working towards a shared goal, the basic tenets of benefiting from data sharing are shown. In addition, eMERGE sites deposited information into dbGaP to facilitate the future use of this information.
So why might I be against the "free sharing" of data for all - namely pharmaceutical companies and data mining? I guess for me it stems to personal opinion again instead of pure logical reasoning. I see the use of information for marketing purposes as serving no great benefit to humanity, whereas the use of information for conducting scientific research does.
Steering things towards the more technical side, how would this apply to those of us focusing on technology instead of legislation? Obviously first we have to share the idea of who should have access. I doubt everyone will necessarily agree with my viewpoint, and I respect that. For those of us who do agree models like dbGaP and HIEs which support sharing data to benefit healthcare, proper security in the systems we develop is key. This definitely has bigger implications for development - security isn't easy - but I also like to remind people that security holes are bugs. By developing bug-free software, in many cases we can avoid some security flaws. The addition of security ensures that the appropriate people will only have access to the data we are providing.
My question back to readers then is not only where you stand on this issue, but also how have you or how do you think we can facilitate better flow of information?
I began AMIA on Sunday night with a Late Breaking session about the Sorrell v. IMS Health case in Vermont where it was outlawed for pharmacies from selling prescription information to data mining and pharma companies. While I make it no secret the details of legal proceedings are often over my head (I initially attended to see a fellow OHSU classmate present), this was a great discussion and raised some important considerations about the sharing of data.
While many of the details of the discussion revolved around the legalities of how data miners were justifying the acquisition of the data under "Freedom of Speech", I found the discussion around how data could and should be used very interesting. My personal opinion is that we do need to continue the ability for researchers to have access to data. This doesn't necessarily mean full, open data available at will, but definitely access to proper data under proper institutional regulations. The ability for aggregating data can and has been shown to improve results, most directly in my experience through the eMERGE project where sites pooling data were able to increase power in genome wide association studies. This may not be the most directly applicable example, as the sites were a collaboratory working towards a shared goal, the basic tenets of benefiting from data sharing are shown. In addition, eMERGE sites deposited information into dbGaP to facilitate the future use of this information.
So why might I be against the "free sharing" of data for all - namely pharmaceutical companies and data mining? I guess for me it stems to personal opinion again instead of pure logical reasoning. I see the use of information for marketing purposes as serving no great benefit to humanity, whereas the use of information for conducting scientific research does.
Steering things towards the more technical side, how would this apply to those of us focusing on technology instead of legislation? Obviously first we have to share the idea of who should have access. I doubt everyone will necessarily agree with my viewpoint, and I respect that. For those of us who do agree models like dbGaP and HIEs which support sharing data to benefit healthcare, proper security in the systems we develop is key. This definitely has bigger implications for development - security isn't easy - but I also like to remind people that security holes are bugs. By developing bug-free software, in many cases we can avoid some security flaws. The addition of security ensures that the appropriate people will only have access to the data we are providing.
My question back to readers then is not only where you stand on this issue, but also how have you or how do you think we can facilitate better flow of information?
Comments
Post a Comment