Ethical Considerations for Research Involving Prisoners (2007)

To comprehensively address the committee’s overarching task of reviewing ethical considerations for protection of prisoners involved in research, the committee cast a broad net for the collection and assessment of information. These sources included commissioned papers (Box A-1), open sessions and workshops, telephone interviews and e-mail surveys to the state departments of corrections (DOCs), a survey of recent literature (to assess basic characteristics of research with prisoners), and two site visits to the correctional facilities.

In addition, a liaison panel of former prisoners and prisoner advocates was assembled for the committee to consult with throughout the project (Box A-2). The committee organized two meetings with the liaison panel to receive their expert advice and guidance in framing the issues, identifying important sources of information, and ensuring a comprehensive analysis. A summary description of the committee’s evidence gathering activities and results follows.

Over the course of the study, the committee sought and received input from former prisoners, representatives of the prisoner advocacy community, bioethics researchers, health professionals, prison services researchers, and other organizations involved with research in prisons. To help accomplish this, the committee held three open meetings. The first was part of the first committee meeting on March 16, 2005. Staff from the Office for

Human Research Protections (OHRP) discussed the current federal regulations and their goals for this Institute of Medicine (IOM) project. Perspectives on the current federal regulation and needed changes were also provided by representatives of the prisoner advocacy community, bioethics researchers, prison services researcher, and a representative from the federal Bureau of Prisons (BOP). The second was a workshop in Washington, D.C., on May 4, 2005. This public workshop focused on the ethical, legal, regulatory frameworks that underlie research involving prisoners. The committee also heard from representatives of the corrections industry about the practicalities of conducting research in correctional settings. A panel of

former prisoners and prisoner advocates talked about needed protections for research involving prisoners. The third workshop was held in San Francisco on July 18, 2005. This workshop focused on topical research areas and methodological issues related to conducting research with correctional populations. Former prisoners and prisoner advocates also presented their views of needed protections. The organizations and individuals that addressed the committee in these open sessions are listed in Box A-3. In

addition, many other individuals attended and participated in the three public meetings (Box A-4).

To help characterize the landscape of published research with prisoners (i.e., who is doing what type of research in what type of prisoner settings), the committee conducted an assessment of prisoner research published in peer-reviewed journals. The preliminary search consisted of English language articles published since 1990, using the following databases: MedLine, PsychLit, Sociological Abstracts, Cumulative Index to Nursing & Allied Health Literature, Criminal Justice Abstracts, Education Resources

Information Center, National Technical Information Service, and Excerpta Medica Database. Search terms used included IRB composition, multisite study/studies, risk-benefit, informed consent, undue influence, vulnerable populations, payment, biomedical research, behavioral research, environment, clinical trials, medication development, FDA, data storage, record keeping, privacy, placebo-control trials, standard of care, follow-up care, follow-up monitoring, data monitoring, HIV/AIDS, tuberculosis, infectious diseases, substance abuse, mental health, women, females, juveniles, adolescents, and mental illness. These search terms were cross-matched with the following subject terms: inmate(s), prisoner(s), incarcerate(d), jail(s). The preliminary search resulted in more than 14,000 articles. The search was then limited to the past 5 years, which resulted in a selection of 1,870 articles.

A random sampling of 20 percent of the 1,870 articles was selected as the final sample. Of these 374 studies, a total of 327 were studies that included human subjects. The remaining 47 included the following types of articles: review articles, commentaries, introductions to special editions, letters to the editor, position pieces, editorials, theory articles, news articles, legal reviews, opinion pieces, discussion pieces, and news type articles.

All of the articles were reviewed and coded using the standard criteria. The results follow.

The results of the survey to assess the general characteristics of published research with prisoners are summarized in the following figures.

Funding stemmed from a variety of sources, including the federal government, state agencies, universities, and the private sector (Figures A-1, A-2).

Most studies (66 percent) did not report the mechanism by which they were approved (Figure A-3). Fifteen percent indicated institutional review board approval; other entity review was 19 percent.

FIGURE A-1 Source of funding.

FIGURE A-3 Mechanism of approval.

As shown in Figure A-4, epidemiological studies were the most common (39 percent). Other common study designs included correlational studies (27 percent) and those assessing behavioral outcomes (14 percent).

FIGURE A-4 Study design.

Most of the studies (41 percent) in the sample had a sociobehavioral focus, lacked a therapeutic purpose, and had minimal risk to participants (Figure A-5). Program evaluations (26 percent) and record reviews (21 percent) were also common.

FIGURE A-5 Type of study. (Same as Figure 2-3).

NOTE: Greater than minimal risk included any biomedical (nontherapeutic) study; any medical therapeutic study (regardless of the existence of a standard of care); any social/behavioral therapeutic study; and any nontherapeutic study involving a manipulation that the research assistant judged to involve potentially serious physical or emotional stress (e.g., long sleep deprivation). Not greater than minimal risk included any study based on review of administrative records; any program evaluation study; any nontherapeutic social/behavioral study that either involved no manipulation (e.g., innocuous questionnaires/surveys) or involved a manipulation that the research assistant judged did not involve potentially serious physical or emotional stress (e.g., long sleep deprivation).

Studies were largely focused on health status (43 percent) and personality characteristics (19 percent) (Figure A-6).

More than half of the studies (53 percent) were conducted in prisons or jails (Figure A-7). Another large proportion of the studies (37 percent) were conducted in alternate settings, such as treatment programs or postincarceration settings (Figures A-7, A-8).

FIGURE A-6 Categories of research.

FIGURE A-7 Facilities/location of studies.

The number of participants in a published article ranged from 1 to 336,668. Most studies (272) included 1,000 or fewer participants (Figure A-9).

Gender More studies included male participants than female participants (Figure A-10).

Age Most studies included adult participants; few included participants younger than 18 years (Figure A-11).

FIGURE A-13 Number of studies with participants of “other” race/ethnicity.

FIGURE A-14 Number of studies with nonwhite participants.Mulatto

Race/Ethnicity Most studies included Caucasians, closely followed by African Americans, and Latinos/Hispanics (Figure A-12). Other racial and ethnic groups were represented to a lesser extent (Figures A-13, A-14).

The committee conducted two site visits to correctional facilities. On July 20, 2005, the committee visited San Quentin Prison in San Francisco and the California Medical Facility (CMF) at Vacaville, California. During their site visits, the committee had guided tours of both facilities and un-

structured discussions with peer educators (i.e., inmates who are trained to be peer educators) about research experiences and needed protections when participating in research.

As part of its data collection activities, the committee collected information from state DOCs regarding research policies and practices. The committee conducted telephone interviews with six states and sent a survey to the remaining states and District of Columbia by e-mail.

The committee collected information from DOCs in six states via telephone interviews: New York, California, Iowa, Texas, Florida, and Utah. The interviews covered the following:

• types of research that are conducted,

• number of studies that have been undertaken in recent years,

• requirements for informed consent,

• degree of risk to which research subjects are subjected,

• procedures for processing research proposals,

• credentials and qualifications of the people charged with the responsibility of approving research,

• problems or concerns that have arisen in connection with such research, and

• impact of laws and regulations on proposed or actual research projects.

Chapter 2 includes summary results of those telephone interviews.

In the interest of reaching all 50 states and the District of Columbia, similar information to that which was collected by telephone interviews was requested from the DOCs of the remaining 44 states and Washington, D.C., in an e-mail survey. The purpose of this survey was to poll the states’ DOCs about their research activities and practices. All but three (Delaware, Illinois, Wyoming) DOCs responded, bringing to 48 the total number of DOCs about which the committee had information (6 from telephone interviews, 42 from e-mail survey). Table A-1 presents the survey questions and a summary of DOC responses.