Agency Recognition of Multiple Principal Investigators on Federally Funded Research Projects

AGENCY:

Executive Office of the President, Office of Science and Technology Policy (OSTP) and Office of Management and Budget (OMB), Office of Federal Financial Management.

ACTION:

Notice of policy on recognition of multiple Principal Investigators (PIs) on awards made under Federal research and research-related programs.

SUMMARY:

Many areas of today's research require multi-disciplinary teams in which the intellectual leadership of the project is shared among two or more individuals. To facilitate this team approach through recognition of the contributions of the team leadership members, OSTP issued a memorandum to all Federal research agencies on January 4, 2005, requiring them to formally allow more than one PI on individual research awards. The Federal agencies then sought input from the research community—scientists, research administrators, and organizations that represent components of the scientific community—on how best to implement this policy. This input was sought via a Request for Information published in the Federal Register on July 18, 2005 that posed a series of questions around core elements that will comprise each agency's implementation plan. The six core elements, to be posted on the Research Business Models (RBM) Web Site, include: (1) Statement of what constitutes a PI; (2) designation of contact PI; (3) application instructions for listing more than one PI; (4) PIs at different institutions; (5) access to award and review information; and (6) identification of all PIs in public data systems. The Supplementary Information section of this Notice provides background on the Research Business Models (RBM) Subcommittee of the Committee on Science (COS), the plan to recognize multiple PIs on Federal research projects, a summary of the responses to the Request for Information, and the government response to the comments submitted. The final policy on the recognition of multiple PIs is contained in the Policy Section.

SUPPLEMENTARY INFORMATION:

I. Background on RBM

This project is an initiative of the Research Business Models (RBM) Subcommittee of the Committee on Science (COS), a committee of the National Science and Technology Council. The RBM Subcommittee's objectives include:

• Facilitating a coordinated effort across Federal agencies to address policy implications arising from the changing nature of scientific research, and
• Examining the effects of these changes on business models for the conduct of scientific research sponsored by the Federal Government.

The Subcommittee used public comments, agency perspectives, and input from a series of regional public meetings to identify priority areas in which it would focus its initial efforts. In each priority area, the Subcommittee is pursuing initiatives to promote, as appropriate, either common policy, the streamlining of current procedures, or the identification of agencies' and institutions' “effective practices.” As information about the initiatives becomes available, it is posted at the Subcommittee's Internet site http://rbm.nih.gov .

II. Background on the Recognition of Multiple PIs on Federal Research Projects

Many areas of research, in particular, translations of complex discoveries into useful applications, require multi-disciplinary and inter-disciplinary teams. Innovation and progress still spring from and depend on creative individual investigators, but collaborative synergy plays an increasingly important role in advancing science and engineering.

Multi-disciplinary research teams can be organized in a variety of ways. Research teams vary in terms of size, hierarchy, location of participants, goals, and structure. Depending on the size and the goals, the management structure of a team may include: A director and/or multiple directors, assistant or associate directors, managers, group leaders, team leaders, investigators, and others as needed. Regardless of how a research team is organized, a pertinent and important question is how to apportion credit fairly if multiple individuals provide the intellectual leadership and direction of the team effort.

Acting on the recommendation of the RBM Subcommittee, the COS concluded that team research would be enhanced if all Federal agencies allowed more than one PI on individual research awards. Some agencies already do this, either formally or informally, but the COS action, which led to a directive to all research agency heads by the Director, OSTP, dated January 4, 2005, extends the practice to all research agencies as a matter of policy.

Request for Information

A Request for Information soliciting input from the research community on several core issues related to recognizing multiple PIs was published in the Federal Register on July 18, 2005 to guide the agencies as they developed their plans for implementing the policy on recognizing multiple PIs.

Respondents: A total of sixty-three comments were received from twenty-nine biomedical scientists, twenty-three universities (Office of Sponsored Projects or Vice President for Research), nine professional associations, one small business, and one unknown affiliation.

Core Elements of Agency Implementation Plans, RFI Questions, Comments From Respondents, and Agency Responses to Comments

General: Overall opinions on the Multiple PI policy (if stated in the comment) were overwhelmingly favorable: 45 in favor, 8 opposed. Answers to the individual questions in the RFI were listed and categorized only if the respondent addressed that issue specifically. Many respondents did not reply to the questions individually or address some of the issues at all. Numbers in parentheses indicate multiple responses citing the same issue or suggestion.

#1: Statement of What Constitutes a PI

Q 1: Will listing more than one individual as a PI present any difficulties for you or your institution?

Comments:

• Need explicit criteria, give examples of what is and is not a PI. (7)
• PI means and needs to be just one individual. (12)
• Keep Co-PI or Co-I titles. (9)
• Possible abuse—too many PIs. (6)
• Maintain maximum institutional flexibility and autonomy in designating PIs. (7)
• Institutions will have to revise processes and databases. (7)
• Concerns about accountability. (3)
• New investigators named as PI might lose status as new investigator. (4)
• May be administratively cumbersome. (2)
• Increased administrative burden. (2)
• Concern about decision-making; if no one is in charge, nothing gets done. (2)
• Harder to evaluate departments for grant ranking.
• Should be reserved for large, complex projects, not R01-type.
• Should allow use for just two close collaborators on R01-type.
• Require minimum percent effort (e.g., 20%). (2)
• Do not require minimum effort.

Agency Response: The Research Business Models Subcommittee Task Group on Multiple PIs considered these comments. The task group viewed most of these as concerned with the basic role and definition of what it means to be a PI. The agencies have agreed on a common basic definition that is suitable across all agencies and research institutions. (See Policy Section of this Notice.) In their implementation plans, agencies may elaborate on the criteria for PIs in their respective areas of science, giving examples of what does and does not qualify as a PI for particular kinds of projects, as well as the specific nomenclature that will be employed in implementation of the multiple PI concept, e.g., Project Coordinator, PI and Co-PIs, or Coordinating PI.

Institutions have the option to name one or more than one PI for each project. It is the prerogative and responsibility of the applicant organization to designate PI(s) for projects.

All PIs will be named in the official award. There will be no Federal-wide limit to the number of PIs per project; however, an agency may impose a limitation as part of their implementation plan.

#2: Designation of Contact PI

Q 2: Do you see any difficulties that would be created by designation of one PI as the Contact PI? Are there institutional issues that the agencies should consider?

Comments:

• Contact PI may become the de facto chief PI. (6)
• Favor since it is important that institution/project speak with one voice. (3)
• Most junior PI may be assigned this role and/or may feel put upon. (4)
• Must be able to enforce communication responsibilities. (2)
• Create Chief Operating/Admin Officer. (2)
• Create Lead PI or Project Director for management and regulatory compliance issues.
• Agency or institution could set up e-mail group for all PIs. (2)
• Diffusion of accountability. (2)
• Not practical if awards to more than one institution.
• Should be able to switch over course of grant.

Agency Response: All comments addressed the need for a single point of contact between the institution and the Federal agency on issues concerning scientific and technical aspects of the project. There was some concern that either the designated Contact PI would become the de facto overall PI on the project or the most junior PI would be assigned this as a largely clerical role. It is the prerogative of the applicant organization to designate the single point of contact. The agencies consider this “Contact PI” role to be primarily for communication purposes on the scientific and related budgetary aspects of the project (see Agency Implementation section below.)

#3: Application Instructions for Listing More Than One PI

Q3: What issues should the agencies consider in developing their instructions for applications naming more than one PI?

Comments:

• Management plan a good idea, but only when needed by the type of project. (15)
• Need detailed description of each PI's role and why that justifies PI status; give examples of contributions that do or do not justify PI status. (15)
• When is agency approval needed for budget reallocation. (3)
• Grants.gov form allows only one PI. (3)
• Uniform criteria should be adopted across agencies; definition in RFI is adequate. (2)
• Limit # of PIs.
• Need guidelines for compliance, coordination, decision-making, publication.

Agency Response: Each agency will specify how its standard application procedures will be modified, if necessary, to reflect the overall policy accommodating multiple PIs. This may include instructions for describing, within the research plan, the specific areas of responsibility for each PI and how the team will function. The government-wide policy does not mandate a formal management or leadership plan, but a specific agency funding opportunity or solicitation may require it.

#4: PIs at Different Institutions

Q 4: Recognizing that agencies differ in the structure of their business arrangements with institutions, are there ways for the agencies to recognize PIs for a team effort involving multiple departments or institutions that would work well for your institution? What issues should the agencies consider in deciding on the most appropriate award structure?

Comments:

• Each type of award structure (subawards, separate awards) has its advantages in different situations; maintain range of award structures as appropriate to each situation. (12)
• Linked awards are a good idea, when appropriate. (5)
• Linked awards may affect institution's FAR simplified acquisition threshold.
• Need to address distribution of indirect costs among institutions/departments. (3)
• Accountability issues between institutions. (3)
• Institutions can handle these issues themselves.

Agency Response: Many respondents noted that each type of award structure (e.g., subawards or separate awards) has advantages in different situations. The agencies agree and will continue to use a range of award mechanisms. Institutions will have great latitude in proposing arrangements that will work best for the particular project and institutions involved. Agencies may, for example, use linked awards (separate awards to each research organization participating in a project), but the government-wide policy does not mandate their use.

#5: Access to Award and Review Information

Q 5: Do you favor granting access to award and review information to all named PIs, not just the Contact PI? Do you anticipate any difficulties in granting such access?

Comments:

• Favor granting access to all (27); oppose (0).

Agency Response: Since there was no controversy on this issue, the agencies will make review and award information available to all named PIs, to the extent that they currently make such information available to a single PI. Agency implementation plans will describe how and when this information can be accessed.

#6: Access to Public Data Systems

Q 6a: Do you anticipate significant benefits from listing more than one PI in agency databases? Do you anticipate any difficulties with such listings?

Comments:

• Will guarantee appropriate credit for team PIs (all comments cited this).
• Should include Co-Investigators as well as PIs. (7)
• Enable better tracking of funding by agencies and institutions.
• Will benefit junior investigators. (2)
• NIH ranking tables would be more accurate. (2)
• Harder to monitor duplicate funding. (2)
• Allows identification of potential future collaborators.
• Provides for multiple contacts per project; but not all contacts appropriate.

Q 6b: Do you anticipate using agency data systems with PI information, such that investment in alterations to such systems would be worthwhile?

Comments:

• Warrants investment (9); maybe (2); no (0).
• Numerous comments that this would be the most important single aspect of implementing the multiple PI policy.

Agency Response: The comments emphasized the benefit of giving appropriate credit for shared leadership of a team project. There was some encouragement for agencies to track the participation of scientists at less than PI level as well, but the policy will not require this. Agency data systems will eventually list all PIs on multiple PI projects. Because changes to existing data systems to accommodate this requirement may be extremely costly, there will be no mandated date for achieving these changes. Agency implementation plans will be required to address the issue of when their data systems may be expected to reflect the new policy on listing all PIs. Agencies may also consult with the Office of Management and Budget's Electronic Government (E-gov) office regarding system changes that are part of implementation plans.

Other Considerations

Q 7: Overall, do you think that the changes proposed for official recognition of multiple PIs will benefit multi-disciplinary and inter-disciplinary research?

Comments:

• The public comments uniformly reinforced the importance of official recognition of multiple PIs in facilitating multi-disciplinary and inter-disciplinary research.

Agency Response: No response is necessary; the policy will be implemented as described for the preceding core issues.

Q 8: What other suggestions do you have for facilitating the recognition of multiple PIs?

Comments:

• Apportion budgets among PIs (favor: 18, distributed evenly across PI, university, association respondents; oppose: 2, one university, one association).
• Minimize additional administrative burden of financial and programmatic management. (3)
• Need designation of responsibility for ethical conduct, human subjects, animal welfare. (2)
• Other agencies do not provide tracking data as NIH and NSF do. (2)
• Need procedures for resolving disputes.
• Should have definition of Co-Investigator.
• Urge rapid and uniform implementation across agencies.
• Provide institutions with ability to apportion responsibility along with recognition.
• Allow collaborating PIs to participate in other grant mechanisms (e.g., cap on number of grants/PI).

Agency Response: Most of these issues have been addressed in the previous responses to the core issues. Implementation plans to be posted on the RBM Web site for the policy on multiple PIs will use a common format to address each of the core issues. Agencies will have the latitude to expand upon the basic requirements for each issue, as appropriate for their research communities, and will address these variances in supplemental material provided through links to their own agency Web sites or through published information.

Apportionment of budgets to individual PIs is not a core implementation feature. If it is done at all, it will be addressed in agency-specific implementation plans.

Policy

All Federal research agencies will recognize multiple Principal Investigators (PIs) on research projects (grants and contracts). Proposing institutions may identify individuals as PIs in proposals when those individuals share the major authority and responsibility for leading and directing the project, intellectually and logistically. This policy does not replace the use of a single Principal Investigator when that is most appropriate for the project.

Statement of What Constitutes a Principal Investigator

A Principal Investigator is the individual(s) a research organization designates as having an appropriate level of authority and responsibility for the proper conduct of the research, including the appropriate use of funds and administrative requirements such as the submission of scientific progress reports to the agency. When an organization designates more than one PI, it identifies them as individuals who share the authority and responsibility for leading and directing the research, intellectually and logistically. The sponsoring agency does not infer any distinction in scientific stature among multiple PIs.

Discussion

It should be emphasized that naming multiple PIs for a proposed research project is solely at the discretion of the proposing institution(s). This concept is similar to the widely accepted practice of recognizing the contributions and responsibilities of business partners. The government's recognition of more than one individual as PI also is not intended to alter the working relationship between team members as they collaboratively allocate resources within the team, subject to any constraints of the awardee institution or the Federal agency under the award terms and conditions, nor as they apportion credit for research accomplishments. Compliance requirements will continue to apply to individuals and institutions, as they do today, regardless of the designation of multiple PIs.

The agencies recognize that teams frequently cut across institutional and geographic boundaries and that team efforts therefore often involve subcontracting or consortia arrangements between different institutions. Based on the experience that some agencies already have with research teams spanning multiple institutions, the agencies are confident that recognition of personnel involved in multi-institution research projects will not substantively alter these well established relationships between institutions.

Agency Implementation

In order to implement the policy on recognition of multiple PIs, each Federal research agency will post in the Research Business Models Toolkit its own plan for implementing the policy beginning in calendar year 2008. Because changes to existing data systems to accommodate the policy may be costly, there will be no mandated date for achieving these changes. Agency implementation plans will be required to address the issue of when their data systems may be expected to reflect the new policy. Agency implementation plans will be posted in the RBM website no later than February 2008. Each agency's implementation plan will include the following elements:

(1) Statement of What Constitutes a Principal Investigator

Each agency will describe if its definition of PI differs in any way from the Federal-wide definition either routinely or in special solicitations.

(2) Designation of Contact PI or Project Coordinator

Each project with multiple PIs will have a Contact PI, or Project Coordinator, to whom agency program officials will direct all communications related to scientific, technical, and budgetary aspects of the project. By recognizing a person as a Contact PI or Project Coordinator, a Federal agency will not confer any difference in scientific stature to that person. Some agencies may designate a specific term for this role in their agency-specific implementation procedures, which may differ by solicitation or type of award mechanism, for example Project Coordinator, PI and Co-PIs, or Coordinating PI.

(3) Application Instructions

Each agency will specify how its standard application procedures will be modified, if necessary, to reflect the overall policy accommodating multiple PIs.

(4) PIs at Different Institutions

Agencies will use the full range of award mechanisms currently used by each agency, and institutions will have great latitude in proposing arrangements that will work best for the particular project and institutions involved.

(5) Access to Review and Award Information

Agencies will make review and award information available to all named PIs, to the extent that they provide this information to single PIs.

(6) Identification of All PIs in Public Data Systems

Agency data systems will eventually list all PIs on multiple PI projects. Agency implementation plans will address the issue of when their data systems may be expected to reflect the new policy on listing all PIs.