Human Studies Considerations with Publicly Archived Data:

There are several human studies (a.k.a., Institutional Review Board — IRB) issues related to contribution of EMG data to a public archive such as the EMGlab website. Our advice in this note is derived from practices of the U.S. National Institutes of Health (NIH), but should be relevant to most other domestic and foreign institutions and funding agencies. In all cases, data contributed to the site must satisfy the IRB practices of the EMGlab web site and your local IRB — both in word and deed.

Donation of your data to a public archive is always subject to the practices and permissions of your local IRB. The Informed Consent document signed (typically) by each subject should ideally inform each subject that their data will be archived on a public web site. [See reference (1) below for some points to consider when designing your Informed Consent document.] If your Informed Consent strictly limits use of your data for your original study, then both IRB approval and (likely) re-consenting of all subjects would be required in order to contribute your data. (Thus, we would argue against including such restrictions in your Informed Consent document, if permitted by your local IRB.) If your Informed Consent document is neutral as to use of the data beyond the initial study, then it is recommended that you seek the approval of your local IRB prior to donating the data — donation of these data may be entirely appropriate.

NIH rules (and most IRBs) make utilization of archived data by other U.S. investigators much simpler/easier if strict anonymity of subject identities is maintained. In fact, if the data in a public archive can NEVER be linked back to a subject (termed by NIH as “unlinked” data), then U.S. users of the data base may NOT even need IRB approval or even IRB exemption to perform studies using these data.  [See reference (4) below.]  Basically, since the information would not be considered identifiable data, it is not regulated by the U.S. human studies rules.  This set-up is VERY convenient for researchers who download data from an archive. In addition, this option is simplest for those who maintain the data archive — if all data in the archive are “unlinked,” then IRB supervision of the archive may not be required.

The easiest way to eliminate links to subject identities is to destroy subject identity information before contributing data to a public archive. (Again, do so with the agreement of your local IRB.) This method permanently eliminates all links, which is the most complete and definitive manner to preserve subject anonymity. Note that unlinking your own data may also make it easier for you to re-use the data within your own lab! Once unlinked, you may no longer require IRB supervision to utilize the data in future studies. Alternative forms of identity control exist. Contributing researchers can request a Certificate of Confidentiality from NIH (see http://www.hhs.gov/ohrp/humansubjects/guidance/certconf.htm). This certificate can help to prevent disclosure of subject identities in any Federal, State, or local civil, criminal, administrative, legislative, or other proceedings. Alternatively, database users can request a letter or agreement from the database contributor stating that the database contributor will not reveal subject identities to the user under any circumstances.

Finally, it is always best to communicate with your local IRB about any contribution of data. U.S. rules provide a fair amount of flexibility and local interpretation to many human studies questions. Thus, it is common for different IRBs to set different guidelines and exhibit different levels of supervision over human studies. It is always a requirement that you satisfy the rules, regulations and practices of your local IRB.

Advice Relative to IRB Applications and Informed Consent Documents:

Note that there are two developing trends in U.S. policy towards use of human studies data. Locally, most IRBs are becoming increasingly strict that the use of data be clearly and precisely specified in IRB applications and subject Informed Consent documents. There is growing scrutiny and restrictions related to data use and mechanisms to preserve subject anonymity. More globally, funding agencies and the general public are recognizing the value of sharing data beyond the use of the original study. The costs and risks incurred acquiring and processing data best contribute to the public good if the broadest use possible is made from the collected data. For example, NIH grants requesting more than $500,000 in direct costs in any funding year are expected to include a plan for sharing of the collected research data (or justification for not doing so).

At first, these two trends may seem at odds with each other. However, new research studies can resolve these requirements by explicitly acknowledging (in the IRB application and in the Informed Consent document) that the collected data will be contributed to a public archive. We do not yet have experience in gaining approval of this arrangement from an IRB, however available NIH guidance (see the Reference section of this note) suggest appropriate mechanisms for doing so.

Please note that there are several different ways in which to protect the identity of subjects, which is often the prime concern in a public archive that stores recorded EMG (and associated) data. Below, we suggest IRB application and Informed Consent document wording for two of the most common cases. These are only recommended wordings. We’d enjoy receiving feedback as to how this wording is received by local IRBs.

References (All from NIH web sites):

We hope the above information is useful to you. Thank you for considering a data contribution to EMGlab.

Report prepared by: Ted Clancy, 25 July 2006