- Table of
- What's New
- Research &
- ASA Home
Sally T. Hillsman,
Is it ethical for data collected for one purpose to be used for another entirely different purpose for which participants never gave informed consent?
In 2007, sociologist Nicholas Christakis and his co-author James Fowler published a path-breaking analysis demonstrating that a person’s chances of becoming obese increased 57 percent if their friends became obese. The data on which these widely reported findings were based were collected for an entirely different purpose by other researchers for the Framingham Heart Study. Because it was a panel study, the original researchers collected contact information on participants’ close friends and neighbors who might know how to contact a participant who moved or changed phone number between contacts. This contact information allowed Christakis and Fowler to map tens of thousands of network links between 5,124 individuals, and study health outcomes based on individuals’ positions in that network. Yet no participant in the Framingham Heart Study gave explicit informed consent to have their data used in this way.
On July 26, 2011, an Advanced Notice of Proposed Rulemaking (ANPRM) was published in the Federal Register (Vol 76, No 143) soliciting comments on U.S. Department of Health & Human Services (HHS) proposals for an extensive revision of the Common Rule, the federal rules governing federally sponsored research with human participants. The proposed revisions are intended in part to reflect significant differences between social science research and biomedical research in ways the existing rules do not. However, if the new proposals outlined in the ANPRM were in place when the original Framingham data were collected, Christakis and Fowler could have been required by their IRB to obtain written consent from each of those 5,124 individuals.
The importance of the ANPRM prompted ASA member Harry Perlstadt to submit a resolution to Council at the ASA Business Meeting in Las Vegas in August, asking the ASA to develop a response to the ANPRM. The resolution, which Council approved, coincided with efforts already under way in the ASA Executive Office to develop a response to the ANPRM in collaboration with other scholarly associations in the behavioral and social sciences. Two months later ASA added its endorsement to a white paper submitted by a group of 22 associations representing more than 350,000 members in the social and behavioral sciences (see www.asanet.org/images/press/docs/pdf/Common_Rule.pdf).
This jointly authored Social and Behavioral Science White Pages (“White Paper”) commends HHS for recognizing the need to modernize the Common Rule, and responds point-by-point to the 19 issues and 74 questions included in the ANPRM. Three key issues (among many) highlighted in the White Paper are summarized briefly below.
Gravity of harm versus likelihood of harm. The ANPRM proposes distinguishing among three types of possible risk to human subjects: physical, psychological, and informational risk. The White Paper urges a separate evaluation of the probability of a given harm from its severity. This would permit acceptable risks to include “both high probabilities of minor inconveniences, harms, and discomforts when these are of a magnitude… that characterize people’s daily existence and potentially even slightly more serious harms when the likelihood that these will occur is so extremely remote that they approximate the very low likelihoods of serious harms in daily life.” This distinction would allow a more nuanced consideration of whether proposed research requires full IRB review, and would also allow researchers and IRBs to more effectively consider the kinds of protections that should be implimented to reduce the likelihood of that harm occurring.
Risk associated with inappropriate release of personal information. In much social science research, disclosure of sensitive information is the most significant potential harm, but one that is distinct from the physical and psychological harms that are central concerns of most IRBs. To deal with this the ANPRM proposes removing informational risk entirely from the responsibility of IRBs. Instead, the proposal is that future federal regulations would apply the HIPAA Privacy Act standards to all research data. The White Paper counters that while removing data security issues from the purview of IRBs is well advised, HIPPA is an inappropriate standard to apply to non-health records (and has significant limitations even in health research). Instead, the White Paper urges HHS to examine the extensive research and practice history of “extant, time-tested, and cumulative standards, methods, measures, and advisement for protecting personally identifiable information, ensuring data protection and security, and permitting access and use” that are in place in government and major research institutes. These include the National Center for Education Statistics Restricted Use Data Procedures Manual; the ICPSR Guide to Social Science Data Preparation and Archiving, Data Use Agreement, or Instructions for Preparing the Data Protection Plan. The White Paper suggests that these tested data security systems are “more relevant, nuanced, and better adapted to the range of data protection issues that researchers confront” than is HIPPA.
Expedited, exempt and “excused” research. The ANPRM outlines three proposals to change the expedited review processes: revise the criteria and list of research activities eligible for expedited review; eliminate continuing review of these studies; and streamline the submission process. Exempt research and expedited review are a huge issue for the social and behavioral sciences. While the White Paper supports such changes, as well as the ANPRM’s plan to base revisions of the regulations on “systematic, empirical assessment of the level of risk,” it also raises some significant concerns. The White Paper cautions against the regulatory creep implicit in moving six categories of studies that are currently “exempt” from the current regulations into a new category called “excused” that—contrary to the symantic—would actually expand the government’s oversight of research by requiring all studies in these categories to adhere to the new data security protections.
The ANPRM has garnered considerable attention in higher education and research circles; it was even featured on the Kojo Nnamdi Show, a Washington-area NPR program. However, the proposals within ANPRM are not yet federal regulations. They are proposed directions for future regulations rather than proposed regulations themselves. The proposals offer many positive suggestions for change, including elimination of annual IRB reviews for many low-risk projects after initial approval, and even for higher risk projects after data collection is complete; simplification and clarification of informed consent regulations and requirements; and elimination of the requirement that multi-site research projects obtain IRB approval at each site. But as the old saying goes, “the devil is in the detail” and there is, as yet, no detail.
The social science White Paper suggests several directions for greater consideration by HHS of actual practices and proposals for regulatory change, which the social science community and its federal partners have developed since the Common Rule went into effect. Much of this appears to have been ignored by the framers of the AMPRM. Rectifying this omission, as the White Paper suggests, requires a continuing, focussed dialogue among the relevant stakeholders. We hope that will happen.
If and when HHS proposes actual regulatory changes in the Common Rule, federal law requires a Notice of Proposed Rule Making (NPRM), followed by a period of public comment, before any new regulations can be enacted. That would be a long process even without HHS responding positively to the recommendations of the White Paper for more empirical examination of some issues and greater dialogue about options. As author Zachary Schrag put it, when the current human subjects regulations were developed in 1979, “the social sciences were the Rosencrantz and Guildenstern of human subjects regulation. Peripheral to the main action, they stumbled onstage and off, neglected or despised by the main characters, and destined for a bad end” (2010). We can’t afford to repeat that scene.
ASA will continue to work closely with our colleagues in other social and behavioral science associations and with HHS and keep the membership informed. Such collaboration, we hope, will assure that this time our voices are center stage.
Sally T. Hillsman is the Executive Officer of ASA. She can be reached by email at mailto:email@example.com.