Research Using Existing or Secondary Data
Research that involves the use of existing data, documents, records, or specimens from living individuals usually must be reviewed by the IRB in advance of the investigator receiving or analyzing the data. If the data contain individual identifiers, the research may be eligible for an expedited review. If the data are recorded so that participants cannot be identified, either directly or through identifiers linked to the subject, the research may be reviewed by the IRB through exempt procedures. For procedural guidance on these types of projects, see Which Projects Require Review for a discussion of direct versus indirect identifiers as well as which secondary data projects require review (most do).
Consent issues require special attention in secondary data projects. Ordinarily, when a person uses data collected by someone else for another purpose, the consent of the participants must be sought again. For example, if researcher A has interviewed a number of persons for project A, the interview cannot be released later to researcher B for project B. The participant who consented for his or her data to be used in project A might disapprove heartily of project B and might not have provided consent for the latter activity. The wording of the original consent form is critical. If a participant consented to allow his or her blood sample to be available to persons studying blood diseases, his or her sample could be shared with many researchers without additional consent. The original researchers who received consent can re-work the data without new consent provided it is for a related purpose and the original consent form informed participants of this possibility.
Having said that, if the data from the original research are truly anonymous or the data are pooled in a form ensuring anonymity, then consent for secondary use may be waived by the IRB. It is also possible to request that the IRB waive consent when it will be impossible or highly impractical to go back and obtain consent. Issues pertaining to waivers of consent and documentation of consent are further discussed in the page on Consent.
Publicly-available datasets are defined by the IRB as data that have already been compiled and are now available in a library or on the internet, and no permissions are required to access them. Privately-held datasets are those for which some type of permission is necessary to access the data and studies involving these typically require IRB review. (The IRB has made an exception for a few well-known, commonly used repositories of secondary data, such as the InterUniversity Consortium for Political and Social Research-ICPSR.) Research involving data that is both completely de-identified and publicly available without permission or application is not reviewed by the IRB.
Studies using only existing or secondary data must be submitted for IRB review using the Cayuse IRB-Human Ethics application system. More information on this system can be found by visiting Cayuse IRB.