3. recommend steps that the organizations and institutions that conduct research and that the federal government should take to collect and analyze data to monitor and evaluate how well the system for protecting human research participants is operating.

This report therefore provides recommendations for core standards with which to initiate pilot accreditation programs for HRPPPs, specific comments on standards under development, and suggested interim actions that can be used to initiate and monitor the impact of accreditation on the system and its ability to achieve the intended goals. The recommendations, listed below, appear in Box 1, at the conclusion of the Executive Summary, according to how they relate to the three broad categories; that is, whether they respond to the goal of developing an accreditation program, standards, or a system of evaluation. However, all comments are made in the context of the current policy and existing regulatory structures and without the benefit of a full examination of the underlying issues and possible solutions.

The structures and processes constituting an accreditation system are only coming into being and still need to be tested. Therefore, the committee's recommendations are aimed at a moving target. Its recommendations about accreditation standards in particular presume that those standards will evolve substantially, especially with the benefit of feedback from initial pilot tests. The committee recommends standards for pilot testing of accreditation programs, but the committee did not itself formulate those standards. It neither could nor should have done so, for several reasons. First, the accreditation standards should be formulated in a “bootstrap” process, with strong feedback between the formulation of standards and direct experience with the implementation of HRPPP standards. Second, accreditation bodies should be accountable for their standards as well as their accreditation processes. Reliance on “IOM standards” would thus undermine this alignment between authority and responsibility for standard setting at a critical point in the development of (an) accreditation program(s). Finally, the standards will evolve over time and will do so rapidly during initial pilot testing. This iterative process would not be possible with a set of IOM standards produced at this time. As the committee formulated its recommendations, no pilot testing had taken place, and reliance on standards in advance of and independent of such testing runs contrary to early experience with the development of new oversight mechanisms in general and past models of accreditation in particular.

In accordance with its task, the committee reviewed available draft accreditation standards at the time of its deliberations. For this purpose, materials developed by Public Responsibility in Medicine and Research (PRIM&R) and,

subsequently, the National Committee for Quality Assurance (NCQA), were provided to the committee. To assess those materials, the committee found it useful to use the following general criteria: (1) their scope and focus; (2) their relationship to the existing regulatory standards; and (3) the extent to which the standards can be consistently implemented, measured, and enforced, as well as their inclusion of various key elements. For more discussion on the review and elements considered, please see Chapter 3.

Finding 1: The standards proposed by NCQA for VA facilities appear promising for use in the accreditation of VA facilities. Those same standards are the strongest basis for use in the accreditation of other research institutions (see Table 1). The committee regards the standards prepared by NCQA to be more suitable than those prepared by PRIM&R for not only pilot testing in VA facilities but also, with modification, for the accreditation of other research institutions.

Finding 2: Neither set of proposed standards applies readily to the full range of research involving human participants or to the diversity of research institutions that conduct it. Both sets of standards understandably and reasonably start from the kinds of research and the types of research organizations where recent problems have been best documented. It is not clear, however, how standards should be applied to nonbiomedical research settings, contract management organizations, clinical trials cooperative groups, independent IRBs, central IRBs, site management organizations, or units of research sponsors that conduct human research (e.g., research units within federal agencies and private pharmaceutical, biotechnology, and device companies).

How the proposed standards can be adapted to the large and growing fraction of research not conducted in the framework of biomedical research institutions will be an important question to be addressed in pilot tests. This is problematic in two respects. First, many institutions performing research with humans are not primarily focused on clinical research, yet the standards have clearly been formulated with medical research in mind. Second, the accreditation system must cover all types of research organizations. A very large fraction, probably a majority, of clinical research is privately sponsored and conducted outside traditional medical research institutions for which both sets of standards were developed. Failure to include privately sponsored research reviewed by independent IRBs would not only exclude a significant fraction of research with humans but would also call into question whether the accreditation process was skewed in favor of academic health centers. It is premature to judge how accreditation can work for these organizations, but it is critical to include them in any credible accreditation system.