open-discussion
42 Subscribers 
open-discussion > IRB Process
Showing 1-2 of 2 posts
Dec 17, 2009 09:12 PM | David Kennedy
IRB Process
Hi; Can you say a little about what IRB process and hurdles you
faced in the process of making this data publicaly available? That
may provide valuable information for future data sharing efforts
such as this.
Dec 17, 2009 10:12 PM | Michael Milham
RE: IRB Process
David,
Below you can find the privacy policies that were established for the consortium. It includes in point #2 the text that was recommended for usage in IRB requests for approval to share data. Key to approval with IRB committees was the fact that the database is anonymized (no identifying information in file headers or accompanying documents; all face information removed).
Hope the text below helps anyone interested in pursuing public sharing of their data, whether through our site or others.
Best,
Mike
PRIVACY POLICIES
==================
1) Each contributor is responsible for confirming with their ethics committee or institutional review board (IRB) that approval is granted or that an exemption from having to obtain such approval has been recognized, for open distribution of the data. Please find below sample text that can be used for your IRB or ethics committee:
2) The 1000 Connectomes Project data sharing effort will provide the research community with open access to datasets contributed by labs around the world. Datasets provided to the 1000 Connectomes Project are to be de-identified prior to deposition of the data with the project (i.e., removal of any personal identifying information from header/support files). Upon arrival, datasets are automatically organized and header files are replaced with novel header files to guarantee that any identifying personal information within the header or supporting files is removed. Prior to open-access sharing via web-based repository, all datasets will be further de-identified and anonymized by the removal of face information from the image to prevent any inappropriate reconstruction of the image that could lead to identification of a participant. Further, each individual participant's dataset will be assigned a randomized five-digit participant identifier, along with a site identifier (two letters which map to the site providing the data). The relationship between the anonymized code and the original subject ID will be destroyed so as to assure that the dataset will be truly anonymized. For each dataset, demographic information provided via web-archive will be limited to (when available): age (18 and up), gender (male, female) and handedness. This information will serve to facilitate more careful characterization of the data, without entailing risk of violation of confidentiality. Datasets will only be included in the repository upon receipt of written expressed permission for usage of the dataset freely by the general public, without limitation.
Below you can find the privacy policies that were established for the consortium. It includes in point #2 the text that was recommended for usage in IRB requests for approval to share data. Key to approval with IRB committees was the fact that the database is anonymized (no identifying information in file headers or accompanying documents; all face information removed).
Hope the text below helps anyone interested in pursuing public sharing of their data, whether through our site or others.
Best,
Mike
PRIVACY POLICIES
==================
1) Each contributor is responsible for confirming with their ethics committee or institutional review board (IRB) that approval is granted or that an exemption from having to obtain such approval has been recognized, for open distribution of the data. Please find below sample text that can be used for your IRB or ethics committee:
2) The 1000 Connectomes Project data sharing effort will provide the research community with open access to datasets contributed by labs around the world. Datasets provided to the 1000 Connectomes Project are to be de-identified prior to deposition of the data with the project (i.e., removal of any personal identifying information from header/support files). Upon arrival, datasets are automatically organized and header files are replaced with novel header files to guarantee that any identifying personal information within the header or supporting files is removed. Prior to open-access sharing via web-based repository, all datasets will be further de-identified and anonymized by the removal of face information from the image to prevent any inappropriate reconstruction of the image that could lead to identification of a participant. Further, each individual participant's dataset will be assigned a randomized five-digit participant identifier, along with a site identifier (two letters which map to the site providing the data). The relationship between the anonymized code and the original subject ID will be destroyed so as to assure that the dataset will be truly anonymized. For each dataset, demographic information provided via web-archive will be limited to (when available): age (18 and up), gender (male, female) and handedness. This information will serve to facilitate more careful characterization of the data, without entailing risk of violation of confidentiality. Datasets will only be included in the repository upon receipt of written expressed permission for usage of the dataset freely by the general public, without limitation.