Key issues in Conflict of Interest for scientific, engineering, and educational research. (Feature).

Introduction

Conflict of interest (COI) issues are gaining increased attention due to concerns about the objectivity with which research is conducted or is influenced by financial support and other external influences. National Science Foundation (NSF) Office of Inspector General (OIG) receives and evaluates all allegations of conflict of interest made in connection with NSF activities. The allegations we investigate have principally focused on issues related to institutions involved in technology transfer, reviewer conflicts, ineffective institutional COI policies, and university researchers involved in outside companies that obtain Small Business Innovative Research (SBIR) grants and our investigations have resulted in actions against institutions and sanctions against individuals. Through the investigation of these allegations, we have also gained perspective on how conflict of interest and perceived conflict of interest issues affect NSF and its awardees.

On a basic level, conflict of interest deals with perceived or actual bias. Such bias may erode the foundation of scientific research, which rests on a bedrock of impartiality. Unmanaged conflict of interest may call into question the validity and weight of data or the integrity and reputation of the principal investigator (PI) or institution. The public, Congress, other Federal agencies, and the scientific community rely on and need to trust the findings and the quality of data resulting from Federally funded research.

For this reason, NSF requires that recipients of its funding manage conflict of interest. Properly disclosing and managing conflict of interest goes a long way to dispelling perceived bias in research. However, deciding how to manage conflict of interest is a difficult task. There is little information to help institutions identify, much less address, conflict of interest. This article seeks to identify factors institutions may consider in developing a conflict of interest policy by discussing NSF's requirements, expectations, and institutional responsibilities, while highlighting these issues with case examples. The parallel between these expectations and NSF processes for its own employees and reviewers is also discussed.

What Is a Conflict of Interest?

There are many definitions of conflict of interest. Conflict of interest may concern individual or institutional financial interest, collaborations, commitments and relationships, and can broadly be defined as:

A conflict of interest is when one's actual or imputed interest compromises or appears to compromise one's ability to impartially perform one's duty.

NSF focuses on collaboration, commitment, and financial interest with regard to its reviewers and employees. It focuses on financial interest for investigators participating in NSF funded activities.

NSF requires awardee institutions to have a conflict of interest policy in place for its investigators as described in NSF Grant Policy Manual (GPM) section 510. It focuses on financial issues and defines a conflict of interest as:

a significant financial interest that could directly and significantly affect the design, conduct, or reporting of NSF-funded research or educational activities. (GPM [section] 510(d))

What does NSF require of its awardees regarding conflict of interest?

NSF requires each awardee institution with more than fifty employees to:

1. Establish and enforce a written COI policy.

2. Define commitments that must be disclosed.

3. Designate at least one individual to review financial disclosures and resolve conflict.

4. Certify on each NSF proposal that an institutional COT policy is in place and all disclosures have been made.

5. Manage, reduce, or eliminate all identified conflict prior to expending NSF funds.

6. Inform NSF of all unresolved conflict. (See GPM [section]510(a).)

NSF will not review a proposal of an institution employing more than 50 people, unless the institution's Authorized Organizational Representative (AOR) signs a conflict of interest certification form. By signing the form, the AOR certifies that the institution has established and enforces a COT policy that conforms to NSF's requirements. A false certification is a violation of 18 U.S.C. ?? 1001 (False Statements) and can be prosecuted criminally or lead to adverse administrative actions. In the past, OIG has investigated several cases in which institutions have falsely certified to the existence of a COI policy. In egregious instances, we have referred these cases to the Department of Justice for criminal prosecution, or discussed them with NSF management, which can suspend existing NSF awards. In less serious instances, we have worked with institutions to ensure that they develop and implement a policy.

Although most institutions have a COT policy, not all policies conform to NSF requirements. For example, we investigated a large university that had a policy that was inadequate for a number of reasons. Although the institution did have an individual designated to review financial disclosures and evaluate conflict of interest issues, no one submitted disclosures, thus violating the requirement that "the Institutional policy must include adequate enforcement mechanisms, and provide for sanctions where appropriate." (GPM [section]510(e)) Furthermore, the timing of required disclosures was prior to the receipt of funds, rather than ensuring "that investigators have provided all required financial disclosures at the time the proposal is submitted to NSF." (GPM [section]510 (c)). The Principal Investigators (PIs) had not submitted financial disclosures because they were unaware that the policy existed. NSF informed the institution that it planned to suspend all awards to it unless it brought its COI policy into co mpliance with NSF requirements. The university corrected its policy.

NSF rules require that an institution's policy:

should require that each investigator disclose to a responsible representative of the institution all significant financial interests of the investigator (including those of the investigator's spouse and dependent children) (i) that would reasonably appear to be affected by the research or educational activities funded or proposed for funding by NSF; or (ii) in entities whose financial interests would reasonably appear to be affected by such activities. (GPM[section] 510(b))

NSF defines a significant financial interest as: anything of monetary value, including, but not limited to, salary or other payments for service (e.g., consulting fees or honoraria); equity interest (e.g., stocks, stock options, or ownership interest); and intellectual property rights (e.g., patents, copy rights and royalties from such rights). (GPM [section] 510(b))

It is important to note that investigator is broadly defined and includes:

the principal investigator, co-principal investigator, and any other person at the institution who is responsible for the design, conduct, or reporting of research activities funded or proposed for funding by NSF. (GPM [section]510(b))

In short, an institution's policy must ensure that students, collaborators, and others participating substantially on a project all disclose significant financial interests.

What types of conflict might an institution consider in its COI policy?

An institution should have a definition of conflict of interest that reflects the goals it wants to achieve, so long as it is at least as inclusive as the definition in NSF's policy. An institution may want to consider other types of conflicts in addition to financial conflict of interest, since the same taint of bias can occur when an investigator has a conflict based on familial or organizational relationships as occurs when financial interests are at stake. Although an institution's policy must incorporate the concepts articulated in NSF's policy, the institution has the final responsibility for defining its own policy.

Other information an institution might consider requiring an investigator to disclose includes commitments made by the investigator to serve as a board member, trustee, consultant, or contractor to any outside organizations. An institution might also require an investigator to disclose relationships with involved family members, as well as collaborators, advisors and mentors, to ensure that decisions made by investigators are made as objectively as possible.

We investigated an allegation on conflict of interest in which a P1 was alleged to have hired her spouse to work on her NSF grant. This arrangement had not been reviewed or approved by the institution. We found that the institution, a small community college, did not have a COI policy, and this grant was its only NSF award. Because this was a small institution with little experience handing federal awards, we worked with this institution to ensure that it developed a policy that complied with the NSF requirements. Through this work and review of the records and the work on the grant, we determined that the spouse's work on the grant was reasonable.

Similarly, at another institution where two PIs had NSF-funded projects, we discovered the PIs had filed false and duplicative travel claims in conjunction with the PIs' outside employment with a large corporation. Both PIs had received compensation from the corporation. The PIs failed to disclose the employment and their resulting financial interest to the university, as was required by university policy. As part of investigating this allegation, an audit of the grant revealed $71,277.65 in unallowable expenditures. The university reimbursed NSF for the full amount. The PIs provided a written commitment to abide by federal conflict of interests requirements. This latter example highlights the need for institutions to ensure that disclosures by PIs with small businesses, or institutional conflicts created by technology transfer opportunities are carefully evaluated and monitored.

How should institutions handle disclosures?

In practice, institutions vary as to how conflict of interest disclosures are made. Some have conflict of interest committees which meet regularly to discuss conflict of interest issues. Others have a departmental or an administrative official review disclosures and make determinations about managing conflict of interest situations.

Regardless of who reviews the disclosures, the COI policy should enumerate methods for managing a conflict such as waivers, modification, divestiture, declination, disqualification, and/or independent oversight. A waiver may be granted for several reasons, such as the interest is not so substantial as to affect the integrity of decisions that must be made, or the benefit of the research outweighs the potential for conflict of interest. The institution may require divestiture of the financial interest that causes the conflict. The grant or proposal may be modified to remove or lessen the conflict. If the conflict of interest cannot be adequately managed, the institution may decide to decline the grant, withdraw the proposal, or require that the conflicted investigator be disqualified from participating in some decisions surrounding the grant, or even be completely recused from participating in the grant. Alternately, the institution could elect to have an independent third party oversee the decisions of the p arty with the conflict to ensure that the decisions are not based on undue bias. NSF's policy requires that the institution inform NSF of any conflicts that cannot be managed or reduced.

NSF also requires that disclosures be updated, and reviewed, on an ongoing basis, during the life of the grant. The institution's policy must ensure that investigators disclose all required interests at the time of submission of each NSF proposal and that they provide updates throughout the life of the award-- "either on an annual basis, or as new reportable significant financial interests are obtained" (GPM [section] 510 (c)). Because of the requirement for ongoing monitoring, it is important that the records are kept in a manner that allows easy cross-reference between the investigators and the separate grants in which these individuals are involved. For example, we investigated a situation in which a PI attempted to hire his wife as a secretary under a NSF-funded grant. The university denied the request because it violated the university's COI policy. To circumvent university policy, the PI then hired his wife through a company created by his wife and brother. However, the PI failed to report his financia l interest in the holding company. The significant financial interest of a spouse is a required financial disclosure (GPM [section]510 (b)). As a result of the investigation, the university restricted the PI's ability to sign grant-related charges to $100 per voucher for a period of five years. Because ongoing monitoring is fundamentally the university's responsibility, the university reimbursed NSF for $61,189 paid to the wife's company and implemented safeguards to prevent a similar situation from occurring again.

Since institutions review more than conflict of interest in the course of processing a grant proposal, more than one committee or individual may need to review the same materials. We have had the opportunity to assess several institutions' techniques for coordinating these reviews. We have noticed that those that depend on paper routing slips and paper copies of proposals are generally not the most effective at managing multiple conflicts. Institutions may want to consider implementing electronic filing systems to enhance timely review and management of not only conflict of interest, but also other research and funding related issues that bear on conflict of interest decisions. Recent discussions on the work of Institutional Review Boards and evaluation of conflict of interest highlight the importance of concurrent evaluation.

How does NSF handle reviewer conflicts?

In addition to being an investigator on a NSF grant, a researcher may be asked to serve as a reviewer of proposals submitted to NSF. NSF has procedures in place to address conflict of interest that might arise for either panel or ad hoc reviewers. Ad hoc reviewers are advised about conflict of interest issues on the review form on which they provide their recommendation. This notice advises reviewers to notify NSF if they have an affiliation or financial connection with an individual or an institution that could be considered a conflict of interest. If NSF is notified of such issues, an NSF official assesses the issue to determine if a conflict of interest exists, and if so, how to handle that conflict.

The rules for panel reviewers, who are special government employees, include disclosing certain conflict of interest concerning "potentially biasing affiliations or relationships" (NSF Form 1230P), including affiliations with an applicant institution, relationships with an investigator or anyone who has an interest in the proposal, and any financial interests related to the proposal, such as employment by the submitting institution or by a subcontractor on the proposal. Prior to serving as a reviewer, panel reviewers are asked to sign and return a form indicating they have no COI or to reveal any conflict to the appropriate NSF program officer. Program officers rely on written guidance provide by NSF's Designated Agency Ethics Official (DAEO), who is part of the Office of General Counsel. They may also meet with the DAEO to discuss any conflict matter. The program officer, with the advice of the DAEO as necessary, decides that the conflict can be managed by having the reviewer recused from individual proposal s or, alternatively, that the reviewer should be removed from the review process altogether.

We have evaluated allegations that both mall and panel reviewers failed to disclose collaborative relationships and were suspected of either inflating or deflating ratings to support their collaborators or institutions they were affiliated with. Our investigations have resulted in three actions: (a) program officers have been unable to use the reviews provided by the individual involved, (b) the reviewer's conflict has been considered manageable, or (c) the conflict has been deemed insubstantial or unrelated to the proposal under consideration. In working with NSF to make decisions about how to handle the proposals and reviews in question, we have been able to ensure that NSF's funding decisions are fair and objective.

What is NSF's policy for handling its own internal conflict of interest?

In addition to requiring awardees to manage conflict of interest and reviewing such conflict as necessary, NSF sets policies for and manages its own internal conflict of interest disclosures. NSF employs both a permanent and a nonpermanent workforce. Non-permanent employees are generally university faculty who agree to work for NSF for one to two years.

Because the non-permanent staff is from academic institutions, they are closely linked to at least one institution. Because NSF hires them for their expertise in science or engineering, they often have conflicts of interest that must be managed. Such conflicts include not only their relationship with their home institution, but also collaborations with other investigators in their fields and other professional or personal relationships.

NSF Manual 15, incorporating the principles of Executive Order 12674, enjoins all employees to "endeavor to avoid any action creating the appearance that they are violating the law or ethical standards." There is also a regulation requiring employees to recuse themselves from any decision surrounding a particular proposal with which they have a disqualifying conflict (18 U.S.C. [section]207).

NSF's internal procedures have three key tools to assist in managing potential conflict of interest issues: (a) mandatory annual training for employees, (b) mandatory annual reporting, and (c) a disclosure process for employees and reviewers. All of these processes are overseen by DAEO, who also informs OIG about violations as they emerge.

Employees may be required to disclose financial interests, familial ties, and commitments that might raise conflict of interest issues. Some employees are required to file annual financial disclosures; these include members of the National Science Board, program officers, experts, consultants, advisory committee members, panel reviewers (special government employees), selected executives, and employees in sensitive positions. NSF obtains financial disclosures from approximately 50% of its employees.

In addition, if N SF employees want to work outside of NSF, they must have prior written permission to work for employers that do business or may reasonably be expected to do business with NSF. For example, program officers, who are unpaid member of an institution's board of directors, would likely be required to recuse themselves from review of any grant applications from that institution. It is interesting to note that there is a split between what NSF does internally to manage its own conflict of interest and what it requires of its awardees, since it only requires awardees to disclose significant financial interests.

NSF employees receive a COI briefing during new employee orientation and must attend mandatory annual training. On both occasions employees are given an Ethics/Conflict of Interest Manual, which summarizes and clarifies the regulations. The manual is also available on the NSF internal web site. At the annual ethics/COI training the DAEO highlights conflicts rules and guidance with case studies.

The DAEO reviews these annual employee disclosures and has the authority to make decisions regarding conflict of interest within NSF. The DABO also serves an important role as a clearly identified person to whom employees can go for advice on conflict issues. If employees who bring issues to the DAEO follow the written guidance provided, they are afforded a safe-haven for disclosure and evaluation of potential COI issues. This process provides employees with an incentive for disclosing and seeking informed guidance.

NSF takes undisclosed conflict of interests very seriously. For example, we investigated an NSF program manager who oversaw SBIR grants who was alleged to have reviewed grants in which he had a financial interest. During the course of the investigation, we discovered that he owned stock in a small business that had applied for several SBIR grants. The program manager had reviewed all of the submissions. We found that he violated ethical regulations requiring that he recuse himself. It should be noted that he did not recommend the company's proposal for funding, and there is no evidence that he was anything but impartial in his review but he failed to properly disclose the conflict of interest to his superiors and failed to include the stock on his financial disclosure form. Based on the evidence in our investigation report, NSF suspended the program manager without pay. He subsequently resigned from NSF.

What happens when OIG receives an allegation of conflict of interest?

NSF sets the standards for its awardees to manage conflict of interests, and applies the government-wide regulations on managing its own conflict of interest. The NSF Office of Inspector General is responsible for investigating allegations of conflict of interest in connection with NSF activities by awardees and NSF employees. In either case, our process is the same.

We first determine whether there is substance to an allegation. We then investigate substantive allegations. As appropriate, we notify the DAFO, the awardee institution, the Department of Justice (if there are possible civil or criminal sanctions), and/or the Office of Government Ethics, to coordinate an adequate resolution.

The conflict of interest allegations we have investigated have covered a wide range of issues, including officials' failure to safeguard the system by failing either to have or to enforce a COI policy, the receipt of gifts or honoraria, perceived conflicts with regard to mail and panel reviewers, undisclosed financial interests, and violation of post employment restrictions.

Our investigations have resulted in a range of outcomes. These outcomes included corrective action by NSF and the institution such as (a) warnings, reprimands, or mandatory ethics training; (b) restitution of the funds to NSF; or (c) actions by the institution limiting the principal investigators' access to grant funds as well as the payment of significant penalties and the reimbursement of funds to NSF.

Conclusion

Managing conflict of interest issues is becoming increasingly complex. Methodologies for managing conflicts are necessary to ensure the objective conduct of research and integrity in federally funded research. When developing or modifying a conflict of interest policy, institutions can seek guidance from similar institutions, professional societies and federal policies. If the institution receives funding from NSF, its policy must meet NSF requirements. An institution's policy needs to be written and it needs to be distributed, explained, and implemented. The institution must identify the individuals who need to provide disclosures, provide those individuals with notice and training, and document and review these disclosures. Subsequently, the institution needs to objectively assess the disclosures and manage the conflicts-or, if it finds that a conflict is unmanageable, report the conflict to NSF. Experience at NSF suggests that mandatory briefings, on at least a yearly basis, can be an integral part of effe ctive implementation of a conflict of interest policy. In addition, there should be at least one person to whom investigators and others can go to for advice. Ideally, the atmosphere should encourage questions to be raised in a free and open way, in which there can be free dialogue without fear of reprisal.

References

18 U.S.C. [section]208, 2000 is http://www4.law.cornell.edu/uscode/18/208.html

NSF Manual 15 contains Executive Order 12674 Section 101(n), 14 April 1989 can be retrieved at http://www.nsf.gov/home/pubinfo/coi/manual15.htm

NSF Grants Policy Manual (July 1985) retrieved at http://www.nsf.gov/pubsys/ods/getpub.cfm?gpm

NSF Form 1230P (August 1987) retrieved at http://www.nsf.gov/pubsys/ods/getpub.cfm?ods key=form1230p

Lee Stokes, JD, is an Investigative Attorney with the National Science Foundation/Office of Inspector General. She is responsible for investigating civil, criminal, and administrative allegations of wrongdoing, including allegations of research misconduct, involving NSF activities; Ms. Stokes also recommends policies with regard to economy, efficiency, and effectiveness in administering NSF programs and operations. Prior to coming to NSF, Ms. Stokes spent five years at the Department of Justice's Office of Community Oriented Policing Services where she provided training and technical assistance and curricula development for COPS grantees and insured effective delivery of organizational messages through the development of manuals and training. Ms. Stokes graduated from the College of William and Mary's Marshall-Wythe School of Law.

Author's Note: Earlier this year, the NSF'S Office of Inspector General conducted a review of the results of conflict of interest investigations our office received from 1989 through 2001. Through this review, presentations at conferences, information gleaned during the course of investigations, and discussions with NSF ethics officials, we have developed perspectives about conflict of interest issues that may assist institutions in evaluating and managing these issues. Portions of this information have been presented in poster and oral form at two conferences and are available in NSF-OIG brochures and on our web site at http://www.oig.nsf.gov

NSF disclaimer. Lee Stokes is an Investigative Attorney in NSF'S Office of Inspector General. Because the author prepared this article in her capacity as a government official, it is in the public domain [but permission to copy the published article must be received through SRA]. The views expressed here are not necessarily the views of the National Science Board, the National Science Foundation, or the Office of Inspector General. Contact Ms. Stokes at NSF-OIG, 4201 Wilson Blvd., Arlington, VA 22230, USA Email: [e-mail]