Appendix A: SUMMARY OF ACADEMYHEALTH ANNUAL RESEARCH MEETING ROUNDTABLE DISCUSSION
JUNE 26, 2006

On June 26, 2006, members of the Exploratory Committee held a Roundtable Discussion at AcademyHealth’s Annual Research Meeting. The focus of the session was how best to balance the expectations, rights and obligations of sponsors and researchers in the context of contract-funded and grant-funded research.

The Session was moderated by Tom Rice (UCLA) and panelists included: Sara Rosenbaum (George Washington University), Robert Berenson (The Urban Institute), Douglas Kamerow (RTI International; US Editor, BMJ), Harold S. Luft (University of California, San Francisco), and Vincent Mor (Brown University).

Panelist Perspectives

At the start of the session, each panelist outlined their perspective on the issue of funder restrictions. Rosenbaum explained that the Exploratory Committee is in its evidence-gathering phase. She noted that private funders can exercise their own prerogative. Government funders, when making contract awards, are governed by the federal acquisition regulations (FAR). She noted that universities enjoy a unique protection based on the 1991 Stanford case. Institutions, such as RTI and RAND face broad policy questions when dealing with government funders and their contract become complicated when university subcontracts are involved. Rosenbaum highlighted four problematic areas, which the participants were encouraged to address during the session, including ideas of solutions:

Luft noted that increasingly, government contract provisions are becoming problematic. He cited the Betty Dong case in which UCLA refused to defend its research in a government contract dispute. Luft said that increasingly, government contract language is being written broadly and universities continue to try and narrow that language. Moreover, when government contracting officers do not understand the underlying research, contract clauses can be particularly problematic. Moreover, the variation in research projects necessitates variation in contracting language.

Mor pointed to past problems associated with restraints in Data Use Agreements, although he noted that the CMS structure seems to have improved.

Kamerow pointed to drug companies as “straw men.” He noted that the government’s problematic restraints are moving closer toward drug company-type restraints. RTI works primarily on research contracts. He finds that the questions around prior restraint arise in a variety of ways and vary from agency to agency, from contract to contract, and from contracting officer to contracting officer. Ten years ago, Kamerow worked in AHRQ’s Evidence-based Practice Program. Generally, the practice centers conducted AHRQ-funded work were required to submit products to the agency for review and approval. AHRQ argued that the review and approval process was necessary to maintain quality. He noted that they were also “seeking to get glory for AHRQ” and also because of the content of some of the products. Kamerow noted that they did receive pushback from grantees, which was generally resolved. He offered that a solution to the issue may be to develop model language which includes a specific period for review, with an “escape clause” which allow the researcher to publish, if the period for review has pass with no action, without the agency’s name.

In his role as U.S. editor of BMJ, Kamerow explained that if a manuscript is “watered down” before it reaches the journal, there is not good way for the editors to determine what changes may have been made.

Session Discussion

Before the roundtable discussion commenced, the participants were asked to identify themselves and their organizations. Represented at the session were individuals from: University of Washington, Brandeis, UCLA, Kaiser Commission, AHRQ, Kaiser Northwest, Lovelace Clinic Foundation, BCBS MN, RTI International, GWU, CDC, University of Kentucky, Oncology Research Society, Boston VA and Thompson MedStat.

A question was raised about the role of journals in soliciting disclosures about contributions made to journal manuscripts. Arnie Epstein noted that the NEJM raises questions about sources of funding. The journal is working on revising its existing disclosure statement. Hal Luft noted that this particular problem came to light when the UCSF tobacco documents became public and it was not disclosed that Phillip Morris employees were co-authors on papers. Luft said that Health Services Research will soon roll out a new author/disclosure form, similar to the form used by JAMA. The HSR form will require authors to disclose whether sponsors or other organizations contributed to the manuscript, including any review/approve requirements. The disclosure will be required at the time manuscripts are accepted for publication, not at the time of submission. However, the authors will see the form requirement at the time of submission. Doug Kamerow noted that BMJ‘s position is that transparency goes a long way toward resolving the issue of funder influence. However, he does not believe the journal asks for information about any prior approval process associated with submitted manuscripts.

Bill Marder from MedStat noted that they collaborate with researchers at AHRQ. The “review and comment” requirement is part of the collaboration process. There is a somewhat different issue if “review and approval” is required, especially if journals ask for disclosure of manuscript contributors.

One participant who has conducted some federal agency-sponsored work noted that if findings are not consistent with the agency’s agenda, or if the funder disagrees with the findings, they do not want to be named as the funding agency. This could put researchers in a difficult position if they are required by journals to disclose the sponsor’s name. A way to address this dilemma is to add a disclaimer that the work is the author’s and does not represent the views of the government agency.

Another participant noted that “review” varies from agency to agency. Editing, developing statistical strategies, etc. may go beyond the purview of the funder. A “review” may lead to months of delay in getting the work out. Such delay is more frequent when the funder is a close collaborator.

Often universities are well-protected. Other organizations do not have expertise in-house. AcademyHealth might consider developing model contract language to deal with a variety of situations and covering both offerors and awardees. Researchers are placed in a difficult position when work is completed, review by the funding agency is completed, but the product is not released from the agency. Negotiating contracts and resolving publication prohibitions includes a whole series of landmines. It can be the “battle of the boilerplates.”

Federal agencies seem to be funneling funds through small businesses. Some university faculty are responding by developing their own small businesses and accepting work outside of the university setting. This scenario is becoming more common as both universities and government funders are holding firm to the position that they own the data. Faculty are caught in the middle.

If AcademyHealth develops model language and the government agencies are not interested in adopting it, what else can we do? There has to be an “out” in any model language. If the funding agency refuses to approve publication, the researcher could publish without identifying the sponsor or indicate that the findings do not represent the views of the government.

Developing model language would be a useful exercise if it is consistent with the intent of funders, but some funders will continue to resist complete freedom to publish. AcademyHealth might develop a policy position which provides that research funded by public monies should be widely disseminated.

AcademyHealth could develop model contract language which includes multiple options. Or, it could develop a set of good practices. Journal editors could develop a practice to inquire of authors whether they followed the AcademyHealth “set of good practices.” The set of good practices or statement of principles or policy statement could articulate what the technical contract language is trying to accomplish.

It may be useful to gather intelligence from co-authors from government agencies. If publication is restricted, is it because they are concerned about the quality of the research; are they concerned about pleasing a higher authority at the agency; do they want to assure Congressional funding for a particular issue? We need to better understand the motivations behind the groups who are putting up the roadblocks. There simply needs to be a greater understanding of the underlying issues.

On the issue of authorship, some journals require the rights are responsibilities of each author - - detailing each person who had input into a manuscript. Hal Luft pointed out that as an editor, he has no concern about what level of input a funder had, as long as it clearly disclosed. Doug Kamerow agreed, noting that at the BMJ, the issue is one of “contribution.” Did the funder contribute substantially to the manuscript.

AcademyHealth might consider developing an educational Primer which describes the issues that are likely to arise in contract negotiations. The Primer could be developed with an eye toward multiple perspectives (researcher, lawyer, PI, funder). The Primer could help to dispel the current misunderstanding and/or lack of information about publication rights.

Data use agreements remain high for some researchers as a problematic issue - - can you use data after you have completed the initial work?