Anthony L. Asher, M.D., F.A.C.
Director, National Neurosurgery Quality and Outcomes Database
Director, Brain Tumor Program, Carolinas Medical Center
Carolina Neurosurgery and Spine Associates
225 Baldwin Road
Charlotte, North Carolina 28204
Dear Dr. Asher:
The Office for Human Research Protections (OHRP) has received your letter regarding the status of the National Neurosurgery Quality and Outcomes Database (N2QOD) with respect to the Department of Health and Human Services (HHS) regulations for the protection of human subjects in research at 45 CFR Part 46. The letter includes several specific questions related to the N2QOD, which appear to be directed by the more general question of whether and how the HHS regulations at 45 CFR 46 (hereafter “the regulations”) apply to the activities of the institutions contributing to, maintaining, or using the N2QOD. I apologize for the delay in responding.
From the information you provided, we understand the N2QOD to operate through the participation of various institutions in three ways. First, there are the practice groups, the neurosurgery practice groups and hospitals providing neurosurgery to patients as part of standard clinical care. For the first six patients who undergo certain identified types of common surgical procedures at that institution in a given week, the practice groups will be collecting identifiable private information from these patients’ medical records and from routine standard of care pre- and post-operative survey assessment procedures. The practice groups then send this identifiable private information to the institutions serving as the registry, Outcomes Sciences Inc. and the Vanderbilt Institute for Medicine and Public Health (VIMPH). The registry will receive this information and analyze it to establish new national medical practice benchmarks for the quality of the procedures, and to provide reports back to the practice groups about the quality of their services compared to the national benchmarks. It is anticipated that the N2QOD will also release de-identified data from its database to secondary analysis researchers who will use the de-identified data for various research analyses.
If this is an accurate description of the N2QOD activity, our belief is that the regulations do not apply to any of the three types of institutions. The rationale pertaining to each type of institution follows.
First, assuming that research involving human subjects is taking place (see below), OHRP would consider the practice groups to be institutions that are not engaged in human subjects research, in that the practice groups are only releasing private identifiable information obtained through the administration of standard clinical care to patients. (See Scenario B(6) in the OHRP guidance document titled “Guidance on Engagement of Institutions in Human subjects Research”, available on the OHRP website at http://www.hhs.gov/ohrp/policy/engage08.html.)
Second, the registry is performing what OHRP would consider to be nonexempt research involving human subjects; however, the N2QOD registry is not supported or conducted by HHS, and of the two registry institutions, VIMPH has not elected to extend the application of the regulations to nonfederally funded research activities on its HHS Federalwide Assurance, and Outcome Sciences Inc. does not have a Federalwide assurance. Consequently, these two registry institutions do not have to meet any of the requirements of the regulations, including review by an institutional review board (IRB), for this non-HHS funded activity.
Third, OHRP would consider the activity of the institutions of the secondary analysis researchers as not being research involving “human subjects”, according to the definition of “human subject” in the regulations, which is “…a living individual about whom an investigator (whether professional or student) conducting research obtains (1) Data through intervention or interaction with the individual, or (2) Identifiable private information….” (45 CFR Part 46.102(f)). Since the secondary analysis researchers are obtaining de-identified data, and are not intervening or interacting with the subjects in any way, the regulations do not apply to the activities of these secondary analysis researchers, and there is no requirement for review and approval by an IRB.
You specifically asked whether the N2QOD effort is consistent with a quality improvement activity; and if so, whether this effort is research that requires IRB review. We note that the regulations do not define or refer to “quality improvement”; that said, the N2QOD project description does seem to be consistent with what many people consider to be quality improvement, for example, as used in the article you reference by Lynn, Baily, and Botrell where quality improvement is defined as “…systematic, data-guided activities designed to bring about immediate improvements in health care delivery.” (p.667) Collecting data from the practice groups and providing them with reports on the quality of neurosurgical care at their own practices as compared to national benchmarks should provide valuable information to enable the clinical providers to objectively assess and improve the quality of the services they provide.
Regarding whether this effort is “research” according to the regulations, it is OHRP’s opinion that a part of the registry activity is research; however, as previously explained, in this instance the registry is not supported or conducted by HHS, and so the requirements of the regulations, including the requirement for IRB review, do not apply. OHRP does not see the terms quality improvement and research as referring to mutually exclusive types of activities. We consider some activities to be both quality improvement activities and research activities at the same time. For example, in the OHRP Quality Improvement FAQ titled “How does HHS view quality improvement activities in relation to the regulations for human research subject protections?”(at http://answers.hhs.gov/ohrp/categories/1569), we include the statement that “…in some cases quality improvement activities are designed to accomplish a research purpose as well as the purpose of improving the quality of care, and in these cases the regulations for the protection of subjects in research (45 CFR part 46) may apply.”
You further ask if OHRP does conclude that the N2QOD effort is “research” according to the regulations, then which characteristics of the project would lead OHRP to this conclusion. The definition of “research” in the regulations is “…a systematic investigation, including research development, testing and evaluation, designed to develop or contribute to generalizable knowledge” (45 CFR 46.102(d)). N2QOD is described as follows: “In order to effectively determine risk-adjusted benchmarks of safety, quality, and surgical efficiency as well as determine the comparative effectiveness of low-back surgical treatments, high volume data representing a multi-regional and heterogeneous population are required. Hence, a national collaboration of quality and outcomes reporting after lumbar surgery could establish, for the first time, a robust mechanism of quality reporting, risk-adjusted benchmark calculations, comparative effectiveness analysis, and evidence-based practice improvement”. (Section 3.0) The population characteristics, measures used and description of data analysis methods are consistent with the definition’s idea of being a “systematic investigation”. An intended result of the analysis will be new information about the quality of the surgical procedures for a nationally representative sample, which is consistent with the idea of being “… designed to develop or contribute to generalizable knowledge.” In contrast, the part of the registry activity that consists of providing reports back to the practice groups about the quality of the surgical procedures they are using at their own institution, compared to the national benchmarks, so that they can use that information to guide their use of those procedures, represents a quality improvement activity that is not research.
You also ask whether outside researchers who request the release of de-identified data from the N2QOD registry for secondary research analyses would require IRB review and informed consent. As explained above, such activity on the part of the secondary analysis researchers does not meet the definition of “human subject” in the regulations, and therefore the regulations do not apply to this activity, and there is no requirement for either IRB review or informed consent.
We also note that the design of N2QOD could be modified to avoid the circulation of individually identifiable information from one institution to another, thereby providing an additional protection for the rights and welfare of the subjects and further distancing N2QOD from the regulatory jurisdiction of 45 CFR 46. If the practice groups were to replace the personal identifiers with a code and use the code to link the longitudinal data about each patient when releasing the data to the registry, then OHRP would consider the registry to not be conducting research with “human subjects” and thus not subject to the regulations, according to the same rationale given above regarding the activity of the secondary analysis researchers using de-identified data. (See our guidance regarding coded data titled “Guidance on Research Involving Coded Private Information or Biological Specimens”, available on our website at http://www.hhs.gov/ohrp/policy/cdebiol.html.) The registry might even be able to preserve this status if N2QOD were to decide to add onto the project the merging of the N2QOD medical dataset with other administrative datasets with financial data, which you letter suggests might happen in the future; if a computer program can be utilized to merge two separate datasets without any of the researchers ever being able to see the identifiers used to merge the data, OHRP would again consider the researchers to not be conducting research involving “human subjects”, and the regulations would not apply for that reason. We understand that VIMPH already has such an information technology capacity for merging separate datasets within VIMPH; you might want to consider consulting with VIMPH about whether to utilize that capacity for N2QOD.
OHRP recognizes that serious questions and concerns have been raised about the appropriateness of oversight of quality improvement and research activities in health care. We share your concerns about creating unnecessary impediments to improving the quality of healthcare or to the use of data to improve healthcare. At the same time, it is of course important to provide appropriate safeguards for the protection of the rights and welfare of human subjects in research. We welcome any contributions you would like to make regarding appropriate solutions to these issues. Please feel free to contact us if you would like to pursue this discussion further.
[/s/ Ivor A. Pritchard]
Ivor A. Pritchard, Ph.D.
Senior Advisor to the Director of OHRP