Who can make the determination that an activity is (or is not) human subjects research?
At CHOP, the investigator is permitted to make the determination that what they are doing does not meet the definition of human subjects research. Only activities that meet the definition of human subjects research require submission to the IRB for a review and approval or for a determination of exemption. However, there are many situations where it may be unclear whether or not the proposed activities meet the definition of research and if it does, whether or not it involves human subjects. Requests for the IRB to make a determination as whether or not an activity meets the definition of human subjects research should be submitted via the eIRB system. Examples:
A QI may produce as a byproduct information that is worth sharing. Depending on the design of the project, it might or might not be human subjects research.
Receipt of a large Medicaid data set containing hundreds of thousands of records might contain birth dates and dates of service. If the investigator is uncertain, the IRB would need to determine if the individuals would be readily identifiable or not.
Is innovative care designed to enhance the well-being of an individual using an FDA approved drug or device research?
No. The FDA and the IRB do not regulate the practice of medicine. Since the drugs, biologic or device is approved and the intent is to provide care for an individual patient and not to produce generalizable knowledge, this is not human subjects research.
Is a database (registry) designed for use in future research considered research?
Yes. If the database/registry is being designed and used for the purposes of future research then even though no research activities are planned per se, this is considered research and also falls under the requirements of HIPAA. If the database will be used primarily for clinical care or hospital operations purposes, then prior IRB review is not required. However, any research that intends to use the information from the database would require IRB review.
Does a case report require prior IRB review?
No. The CHOP IRB does not consider retrospective case reports to meet the definition of research. However, prospective research involving just a single subject do not qualify as case reports just because the research is limited to one individual.
What about a case series?
It depends; a determination from the IRB may be necessary. A case series can be several notable cases reported at the same time. The IRB does not consider this to be research. In IRB SOP 407: Determining When a Proposal Meets the Definition of Human Subjects Research, the IRB excludes case series of 5 or fewer patients from review. Case series with 6 or more cases must be submitted to the IRB as research or for a determination that they do not meet the definition of research. The rationale for this requirement is that case series involving greater than 5 records frequently involve dozens of records and could be more appropriately classified as descriptive or observational research. See the IRB's webpage on Study Design.
We are conducting a Quality Improvement/Quality Assurance project and have no plans to publish; do we need IRB review?
Maybe. Publication is a consideration but it does not distinguish research from non-research QI/QA projects. The intent of QI/QA is to improve the care of patients at CHOP, often using proven care strategies and methods and most are not research. The project may still produce generalizable knowledge that clinicians would wish to share with others. If the QI/QA project intends to test an as yet unproven care strategy using a standard research methodology, then it is likely that the project is research. For more information, see the OHRP Quality Improvement Activities Frequently Asked Questions website, Quality Improvement vs Research, or the CHOP Screening Checklist for Quality Improvement (QI) Projects.
What about receipt of deidentified datasets or biospecimens?
No. Use of de-identified datasets or biospecimens is not human subjects research. In order for the data/biospecimens to be de-identified, there can be no identifiers and there can be no link to identifiers. A coded specimen that can be linked back to individuals is not considered de-identified. If the providers have access to the key but will not provide the recipient with any PHI, then from the recipients’ perspective the data or specimens are de-identified. The provider of the data/specimens must either provide a letter stating that they will not provide any identifiers, or have policies which clearly state that they will not provide identifiers. For details of the requirements for data sharing, see Guidance on Engagement of Institutions in Human Subjects Research
Can I share de-identified data or biospecimens with another investigator?
Yes. Use of deidentified data or specimens does not constitute human subjects research. However, the investigators must adhere to both the limitations (if any) specified in the consent form used to obtain the specimens and the subject's expressed preferences with regard to future use. If re-use of study data or biospecimens for future research is a possibility, then the consent form must include information about future use. For most studies, subjects should be able to opt-in or opt-out without affecting their ability to enroll in the main study. The IRB's Consent Form Templates include example language for optional future use. The subjects' preferences should be noted as part of the study record to ensure that future use of data and specimens matches each subject's decision.
If CHOP is not enrolling subjects in the research but is functioning as the data coordinating center (DCC) does this require review by the IRB?
It depends. If the research is funded by the DHHS or if the DCC will receive identifiable subject information, then the CHOP is engaged in the research and the IRB must review the activities of the DCC. If the DCC will only receive de-identified data and the funding is not from DHHS, then there is no obligation to obtain IRB approval. Having monitors (CRAs) from the DCC perform source-data verification during a site visit does not engage the DCC in Human Subjects Research unless it receives PHI from the site.