1. Multi-site Research - Central/Single IRBs - Data sharing questions P. Pearl O’Rourke, MD Partners HealthCare October 26, 2015

2. Central/single IRBs Why now? What is the demand? What do we mean by ‘central’ or ‘single’ IRB? –The importance of differentiating between institutional and IRB roles and responsibilities –Models/types of central IRBs Benefits and challenges of local vs central IRB review (real and/or perceived)

3. Why now? Increase in multi-site research Frustration of sponsors and investigators Increased presence of independent (commercial) IRBs –Aka – there is an option Funding agencies’ demands ANPRM

4. BUT What do we mean by central IRB?

5. Important note Human subjects research requires –Regulatory IRB review –Institutional review – for example: PI training and eligibility Ancillary committee reviews –Pharmacy, Radiation Safety, Biomedical Engineering, COI Grants and contracts Sponsored research office negotiations Central IRB review - only IRB regulatory review The local institution remains responsible for the institutional review/s

6. 6 IRB vs. Institution Institutional Federal Wide-Assurance Responsibilities IRB Office Responsibilities IRB Review CIRB only to provide

7. Hence the institution Must retain/develop a system for fulfilling all institutional responsibilities Must develop a mechanism for reporting relevant institutional review information to the central IRB

8. CIRB Arrangements Type of IRB Scope Voluntariness Designation of reviewing IRB

9. CIRB Arrangements: Type of IRB Independent/commercial –This is their business Academic Medical Center or other research entity –This is becoming their business –Pretty-much an add-on to regular business

10. CIRB Arrangements: Scope –Single protocol review One-by-each protocol determination –Category of research; e.g., Cancer Pediatrics Defined network –All research Perhaps the local institution does not have an IRB!

11. CIRB Arrangements: Voluntariness –Protocol-by-protocol decision –Mandated By funding agency Network ‘business’ rules

12. CIRB Arrangements: Designation of reviewing IRB –Single IRB designated for all relevant reviews NCI NeuroNEXT Commercial/Independent IRB –Protocol-by-protocol decision Reciprocal reliance agreement

13. Reciprocal Reliance Agreement –Multiple institutions sign the same reliance agreement Allows any of the signatories to be the reviewing IRB Allows each institution to decide on a protocol-by- protocol basis whether or not to rely

14. Central IRB models Based on IRB regulatory review responsibilities –Initial protocol review –Continuing review –Amendments –Adverse events, unanticipated problems, deviations

15. Central IRB models Non-share model –Central IRB fulfills all IRB-regulatory review requirements Share model –Central IRB and local IRB share regulatory review requirements

16. The Options All local review Each site’s IRB conducts its own review Share review Central IRB and local site IRBs share each have review responsibility Non-Share review Central IRB does only review

17. IRB Tasks Reviewing IRB Relying IRB Reviewing IRB Relying IRB Initial Review ++ Continuing Review ++ ‘Substantial’ Amendments ++ Staff and Minor Amendments ++ Serious Adverse Events (AEs) ++ Local minor AEs ++ ‘Other’ Events +?? Non-Share ModelShare Model Relying on external IRBs

18. Regardless of the model… must also consider…

19. Getting there Selecting the CIRB Reliance Agreement –N.b., HIPAA/Privacy Board Review Robust SOPs needed Develop communication systems between relying sites and CIRB

20. Advantages of Local IRBs HRPP (Human Research Protection Program) coordination – local IRB often the ‘hub’ –Ancillary reviews Including COI assessment –CT.gov –Interface with grants and contracts and sponsored research offices –May promote one-stop shopping for investigators and sponsors Maximizes review of local context Promotes awareness of what research is being conducted and by whom (oversight and control issues) Investigators have only one system to know

21. Disdvantages of Local IRBs Cost of the infrastructure and regulatory compliance Possible loss of sponsored research as well as federal funding –When central IRB mandated For multi-site research –Duplication of review at many sites Often with insignificant changes –Lag time to activate all sites Some sites may drop out –Possible lost opportunities and lost funding Added expense?

22. Advantages of Central IRBs Depending on number and type of protocols reviewed –Possible cost savings for institutions Possibly happier sponsors and funding agencies which could lead to more research collaboration and funding Possibly happier investigators For multi-site research –Efficiency of IRB review - perhaps –Faster to approval and enrollment at all sites –Potential increase in research opportunities

23. Disdvantages of Central IRBs Possible erosion of cohesive HRPP –Distancing of research and researchers Difficulty segregating and fulfilling IRB vs institutional responsibilities New processes for fulfilling institutional-responsibilities –New system for review for eligibility to use a CIRB –New system for engaging the PI –Ancillary reviews Negotiating contract and/or reliance agreements –Possible liability issues Investigators must learn new systems If CIRB does NOT conduct HIPAA review – likely relying institution still retains a role

24. Multi-site Research - Central/Single IRBs - Data sharing questions

25. Data-sharing What is being shared? Why is it being shared? With whom is it being shared? How is it being shared? After it is shared…what then?

26. What is being shared? Type of data: –Health care – medical records, etc. Genetic data? –Insurance –Other Origin of data: –Clinical data –Research data –Combination Collected pursuant to an informed consent? –If yes, what are the specifics re: sharing and future use HIPAA status of the originating and recipient entities

27. What is being shared? Identifiability – and who determines? –Fully identifiable –Limited data set –Indirectly identifiable (e.g., linked) –De-identified Per what standard? Anonymous? Quantity of data –If large amounts with many elements – can it be de-identified? –Small cell size concerns as well

28. Why is it being shared? Research collaboration Development of a repository Condition of grant award –dbGaP Part of sponsor contract Primary research Secondary research

29. With whom is it being shared? Research collaborators –AMCs –Industry Funding entities Federal agencies Other

30. How is it being shared? Hard copy Electronic files How transmitted?

31. Once it is shared...what then? Where is it maintained (security questions) If for a single use –Destroyed or returned after use? If for secondary use –Rules for determining appropriateness of secondary use? If for a repository –Rules for data maintenance, access and use

32. At the end of the day Details matter