Lowery JC (Center for Practice Management & Outcomes Research, VA Medical Center, Ann Arbor, MI)
Baker SJ (Center for Practice Management & Outcomes Research, VA Medical Center, Ann Arbor, MI)
The objective of this study was to determine the magnitude and nature of variability in decisions made by human studies/IRB committees regarding multi-site, low risk health services research (HSR). This study is part of a larger study to identify ways to reduce variability in the protection of human subjects by establishing best precedents.
This was a cross-sectional, observational study with primarily qualitative analyses. Multi-site HSR studies initiated since 2001 were identified from several sources, including the HSR&D Annual Report database. Copies of the original protocol, IRB applications for each participating site, and all IRB correspondence/approvals for each site were obtained for each study. The study material was reviewed to determine common issues across the studies for which IRB responses varied.
Of 50 multi-site studies identified initially, 27 were determined to be appropriate for inclusion. Mean number of sites/study was 7.1 (range of 2 to 43). Studies were about evenly split between observational (13) and intervention (14). Interventions were related to the organization of care delivery, rather than to implementation of experimental treatments. The following 10 issues had variable IRB responses in more than one study (frequency shown in parentheses): type of review (full, expedited, exempt) (16); use of waiver/informed consent (6); inclusion of irrelevant language in consent form (4); training requirements (4); recruitment strategies (3); identification of suicidal patients or undiagnosed conditions (3); provider incentives (3); identification of vulnerable subjects (including employees) (2); description of risks in consent form (2); and union notification (2).
There are considerable differences among sites in interpretation of federal guidelines for protecting human subjects in low risk, HSR studies. Many of the more conservative interpretations do little to improve human subjects protection while increasing costs and delays of conducting the research.
IRB practice variation poses a significant barrier to the conduct of low risk HSR, and that variation must be reduced if this type of research is to remain practical. The first step in reducing variability is to understand the magnitude and nature of the variability. The next step will be the provision of guidance for interpretation of federal guidelines.