NINDS Clinical Research Collaboration Workshop

NINDS Clinical Research Collaboration Workshop
Exploring the Options

October 7 - 8, 2002
Bethesda, MD

• Brief Summary
• Background
• Workshop Participants
  
• Specific Goals and Features of the NINDS CRC
• The Need for a CRC and General Comments
• Incentives for Physicians to Participate
• Options for structuring a CRC
• Post-Workshop Comments by the Co-chairs
• Appendix I: Key issues

NINDS Clinical Research Collaboration (CRC): linking patients and their physicians to clinical research advancing treatment of neurological diseases.

CRC Mission: To advance expeditiously the treatment of neurological diseases by facilitating NINDS-sponsored clinical research.

Brief Summary

Background

Development of an infrastructure to efficiently execute NINDS-sponsored clinical research was approved by the NINDS Extramural Science Staff at its meeting of May 20, 2002. Subsequently, funding was solicited and approved for a workshop to explore the range of options for creating an infrastructure, as well as to identify potential problems and pitfalls. Participants representing a broad range of expertise in clinical trial design, in clinical trial networks and consortia, in minority issues, in human subject protections, in medical ethics, in web-based continuing medical education, in data management and in web-based applications were invited. In addition, advice was sought from advocacy groups and medical subspecialty groups, including internal medicine, family medicine, adult neurology and child neurology. Care was taken to include both academic and non-academic physicians. A representative of the American Academy of Neurology participated. The Workshop was envisioned as the first step of an ongoing process of soliciting input and collaboration from all parties and organizations interested in advancing NINDS clinical research in neurological diseases.

Top

Workshop Participants

National Institute of Neurological Disorders and Stroke (NINDS):
	John R. Marler, M.D., Associate Director, Clinical Trials and Co-chair.
	Robert G. Hart, M.D., Clinical Trials Cluster, Co-chair.
	Peter R. Gilbert, Sc.M, Statistician and Project Manager, Clinical Trials, Co-chair.

Other NIH Institutes and Federal Agencies:
	Donna T. Chen, M.D., M.P.H., Division of Clinical Bioethics, NIMH.
	Lawrence Friedman, M.D., Special Assistant to the Director, NHLBI.
	George Gasparis, Director, Division of Assurances & Q/A, OHRP.
	Richard S. Kaplan, M.D., Chief, Clinical Investigations Branch, NCI.
	James P. Kiley, Ph.D., Director, Division of Lung Diseases, NHLBI.
	Sylvia Scherr, M.S., R.N., Executive Director, Continuing Medical Education, NIH.

Other Invited Participants (alphabetically):
	Stephen Ashwal, M.D., pediatric neurology, Loma Linda Univ, AAN QSS.
	Richard Atkinson, M.D., practicing neurologist, Sacramento, CA.
	Joseph P. Broderick, M.D., Professor of Neurology, Univ. of Cincinnati (Stroke)
	Lucie Bruijn, Ph.D., Science Director & Vice-President, ALS Association.
	Robert M. Califf, M.D., Director, Duke Clinical Research Institute, Durham, NC.
	L. Matthew Frank, M.D., Pediatrics/Neurology, E Virginia Medical School, Norfolk, VA.
	Jacqueline French, M.D., Professor of Neurology, Univ. of Pennsylvania (Epilepsy).
	Cynthia L. Joyce, Director, AAN Education and Research Foundation.
	Walter N. Kernan, M.D., Associate Professor of Medicine, Yale University.
	Kurt Kroenke, M.D., Society of General Internal Medicine, Univ Indiana School of Med.
	Anne S. Lindblad, Ph.D., Vice-President, EMMES Corp., Washington DC.
	Dr. Richard I. Lindley, Department of Clinical Neurosciences, Edinburgh, UK.
	Kevin A. Peterson, M.D., AHRQ Primary Care Research Network, Minneapolis, MN.
	Thomas Sullivan, President, Rockpointe Broadcasting, Washington, DC.
	Janice E. Williams, Ph.D., M.P.H., Neurology Dept., Emory Univ., Atlanta, GA.

In order to establish common background and terms of reference for the wide range of expert participants, the workshop commenced with brief presentations about currently ongoing clinical trial networks at other NIH institutes (Drs. R. Kaplan from NCI, L. Friedman and J. Kiley from NHLBI), non-NIH research networks (R. Califf, Duke Clinical Research Center), and large, simple trial collaborations (R. Lindley, Edinburgh, UK).

Dr. J. Marler (NINDS) then presented information about NINDS clinical studies relevant to the design of a future collaboration and described a tentative vision. His major points and charges to the Workshop included:
- NINDS-sponsored clinical research needs to be more efficient. This will allow testing of more treatments, more expeditiously.
- We need to assess if there is sufficient NINDS-sponsored clinical research to justify a CRC infrastructure in addition to the existing grants and cooperative agreements. Currently funded studies testing interventions, if completed, will include 27,000 participants in over 100 studies (including 28 phase III trials) with a total combined annual budget over $90 million.
- We need to focus on long range planning and strive to design a CRC that will be relevant and effective in 5-10 years, in addition to addressing immediate needs.
- A successful CRC will involve participation of hundreds of neurology and non-neurology physicians and be adaptable to the needs of individual clinical research projects.
- The CRC should not be limited to clinical trials, but include other types of clinical research that could be expedited by a CRC infrastructure.
- There would not be a group of CRC core investigators that would propose the research. In order to ensure quality and wide input, all NINDS-sponsored clinical research studies undertaken by the CRC would be developed initially as independent investigator-initiated proposals that undergo the usual peer-review.

Top

Specific Goals and Features of the NINDS CRC

1. expeditious recruitment (achieving goals in months rather than years)
2. involvement of a wide spectrum of investigators, especially practicing physicians
3. recruitment of a broader range of participants, particularly minorities
4. infrastructure makes clinical study of "small diseases" economically feasible
5. to minimize the cost of the infrastructure, so that overall cost/participant are not increased
6. to make participation in NINDS clinical research more accessible to patients
7. to expedite transfer of research results to clinical practice, especially to community settings
8. to encourage large, streamlined trials, especially for primary prevention of neurological diseases
9. to maintain quality by peer-review of competitive investigator-initiated studies

A tentative vision of the structure of a CRC:

• Internet patient screening

• - Accurate and complete selection estimate

  - Alternative pathways to enrollment

• Physician incentives

• - Case by case reimbursement

  - CME and NINDS/NIH certification

• 10% of U.S. neurologists (n=1000) plus an equal number of other practitioners
• 5000+ patients per year entered into NINDS clinical research studies

Proposed pathways to achieving the CRC goals:

• Ways of matching patients to studies

  - Internet

  - Pre-selected physicians/centers

  - Collaboration with patient advocacy groups

  - Treatment of neurological emergencies

  - Multiple pathways depending on disease and specific research protocol

• Initial contacts

  - Patients

  - General practice

  - Disease specialists

• Motivation of physicians

  - NINDS/NIH Certification

  - CME credit

  - Reimbursement for extra research effort

• Quality control measures

  - Central data and quality control

• Maintaining reasonable expectations

Following these initial presentations, the comments and advice of the Workshop participants were solicited about a wide range of relevant design issues (Appendix I). Near the end of the Workshop, each participant was asked to offer final comments. Ideas and recommendations emerging from the Workshop are summarized below, reflecting the diverse views of the participants.

Top

The Need for a CRC and General Comments

Neurologists are hungry for coordinated efforts to advance treatments of neurological disorders; the devil is in the details of how best to do this in the current practice and academic environments. The urgent need for a CRC was voiced by many workshop participants. Further, NINDS is perceived as the natural leader of such an effort (rather than the American Academy of Neurology or other professional organizations).

We need "an army of neurologists involved in clinical trials attempting to answer critical questions about the care of patients with a variety of neurological diseases." "We don't need: "investigator" networks. We do need: "clinician investigator" networks."

It is important that the first project(s) undertaken by a new CRC be of substantial general interest, be of high visibility, and (hopefully) yield "positive" treatment advances. It was noted how the positive, widely heralded results of the ISIS-I trial in cardiology fuelled enthusiasm for subsequent mega-trials involving thousands of practicing cardiologists. Several participants urged that the initial CRC project(s) be carefully selected in this regard.

Not all clinical trials and other types of clinical research would be appropriate, depending on the final CRC structure. Careful attention should be given to defining types of projects which would and would not benefit from collaboration with the CRC infrastructure. It could be that fewer are appropriate than initially supposed, influencing the economy-of-scale aspect of a CRC. Some minimum number of protocols must be operative for efficient functioning of the CRC

Define and prioritize as clearly as possible the specific needs driving the creation of a CRC. Is it primarily aimed at more rapid patient recruitment? More effective recruitment of minority participants? Testing of interventions in "real-life" practice settings? The capacity to disseminate results through the CRC? Raising the general appreciation of level of clinical evidence? While several important objectives may be embedded the CRC, a firm idea of the priorities will importantly influence CRC structure in the face of several design options.

"Think ahead five years" and define the construct of the future. A tremendous amount of groundwork must be undertaken to change a culture that inhibits collaborative research. Practicing neurologists must become better educated about medical evidence and the benefits of collaboration. It cannot be done overnight. Include in the RFP organization of meetings and educational programs to this end, as well as interaction with professional societies.

"Lay the groundwork carefully before launching" was endorsed by several participants. The NINDS CRC offers the opportunity to be innovative in approach, but consequently there are many new ways to fail. Focus group studies of patients with neurological diseases, particularly including minority patients, was recommended to define interest, motivation and barriers to participation an NINDS CRC. Design a CRC for the future (not necessarily one that can be created in the short-term).

Key elements of success:

1. An important, well-defined mission.
2. Leadership (inspiration)
3. Process (perspiration)

Expert "streamlining" of clinical trial proposals to make them as close as possible to routine clinical practice will encourage participation of practicing physicians. The expertise and effort required is often underestimated. For the International Stroke Trial, several years of development and validation of appropriate outcome measures preceded initiation of the trial.

Developing common instruments/nomenclature (e.g. participant features, functional outcome measures) and the use of common data systems for all CRC-related research projects make things easier for physician participants, more efficient and offer potential research advantages.

"Multiple models, multiple pathways, multiple solutions": The concept of multiple pathways to participation by patients and physicians was proposed. For example, some could be web-based, but others not. For physicians, perhaps separate "tracks" for academic-based and practice-based physicians, who may well have different motivations and barriers to participation. There could be tiered levels of physician participation: some physicians may screen and refer patients, while others would additionally enter and follow patients.

A consistent deficiency in NCI, NHBLI and NINDS clinical research, widely acknowledged by the workshop participants, was the lack of adequate mechanisms to disseminate results to practicing physicians, to modify practice guidelines, and consequently to change clinical practice. Build-in mechanisms to disseminate the results of CRC-affiliated (and perhaps other NINDS-sponsored research) in order to expeditiously change practice. Measures of the impact of the CRC and of the methods of disseminating new treatment information should be a part of its design. Consider a program to expeditiously generate high-quality guidelines based on results of CRC-based research, perhaps collaborating with the American Academy of Neurology and other relevant organizations that develop guidelines.

Include a national registry of rare neurological diseases to facilitate research. Patient privacy protection and access issues would have to be carefully considered.

"But where's the science?" in large, streamlined trials that will allow CRC-appropriate projects to successfully compete in a peer-review process? Several comments were offered in response to this issue:

1. . review committee are becoming more sophisticated about the need for large, streamlined trials that offer definitive evidence to change neurological care and no longer dismiss them as "blue collar" research.
2. it is possible to incorporate projects in participant subgroups at selected clinical sites that involve data collection and procedures to address more basic scientific issues.
3. the results of large definitive trials frequently inform ideas and principles about disease pathogenesis and stimulate basic science more frequently than is generally appreciated. Grant review committees must be educated about this idea.

Input, endorsement and cooperation from professional societies in Neurology and Neurosurgery (e.g. American Academy of Neurology, Child Neurology Society, Association of Neurological Surgeons) should be sought and nurtured. Similarly, cooperation with leaders disease-oriented research groups to assure them of the ancillary, supportive role of an NINDS CRC, not competing with them to fund new research, is important in the early stages ("you've got to get the blessing of the established neurology research groups", per one particularly experienced participant). Early input and collaboration with existing primary care networks was encouraged.

The importance of training aspects for physician participants was emphasized as important for the quality of the research, for human subject protection, and for physician motivation. Training courses and on-line training and CME should be a key feature of the CRC. Stress the CRC as a "clinical learning network" to physician participants; a "virtual university" of clinical training and research.

"This could get expensive" to do well. To justify the expense, the ancillary benefits of the CRC beyond enrolling patients into research studies (i.e. physician education, patient satisfaction, rapid dissemination of new neurological treatments) should be emphasize and measured. Points of reference: the cost per participant in recent clinical NCI oncology trials averaged $6500; for the NHLBI asthma network, the figures range from $8000-$12,000 per participant.

An immediate need is to put into place effective, user-friendly mechanisms (website, hot-line, etc) allowing patients and physicians to access currently recruiting NINDS-sponsored clinical research. The support and advice of trialists and advocacy groups should be sought. Anecdotes were related about the difficulty in finding clinical trials by interested patients. "NINDSclinicaltrialmatch.com" is needed by patients and physicians; a "dating service" model was mentioned, with Eeasy access a priority.

Top

Incentives for Physicians to Participate

Several participants offered the view that many practicing neurologists and other physicians would be a part of a CRC if it could be done with a minimum of hassle, additional work and liability and that financial incentives may not be the key motivating factor.

CRC Workshops at large national neurology meetings (travel sponsored only for a selected few participants who have a high level of participation). These could include training sessions in research methods and ethics and presentation new information emerging from the CRC and other NINDS-sponsored research. Updates by leading researchers would create interest.

Oncologists and cardiologists receive research training and participate in clinical trials routinely during their fellowship training, facilitating their later participation in clinical research. It was noted that 9000 of the 15,000 (60%) oncologists practicing in the U.S. participate in NCI clinical research. Attempts to inform and involve neurology residents in CRC activities might bear fruit in the long run. This might include efforts to require some training in clinical research as part of approved neurology residency training, courses coordinated by the CRC aimed at residents at major meetings, and resident newsletters.

Incentives for participation by practicing physicians could include certification as an NINDS CRC Investigator after completion of training, a minimum CME requirement, and some level of screening/recruitment activity. An attractive poster is envisioned on the wall in the waiting room, proclaiming the physician to be an NINDS CRC Investigator (perhaps for a two year period): "NINDS CRC Investigators - offering the latest treatment, advancing care...".

Top

Options for structuring a CRC

While one vision of the CRC includes involvement of large numbers of non-neurologist physicians, it was suggested that a core of practicing neurologists be recruited as the first step in CRC development, bringing in other physicians after things were running smoothly (neurologists might have more inherent commitment and better tolerate the initial bugs in the system).

Considerable discussion was given to a potential structure of "regional networks" centered on leadership by a local, prominent academic health center (e.g. the Womens' Estrogen Stroke Trial trial model). Residents who trained at the regional academic center could be contacted for participation, with training and quality control monitoring carried-out regionally. The notion of "research mentoring" by the regional academic leadership was put forth. Using this model, the CRC could be initiated in pieces, through a series of regional RFPs to build and operate CRC networks that could field-test aspects in pilot projects. The regional academic leaders could be held accountable for productivity and compared with other regions. An opposing view was that, given an internet-based structure, organization based on geographic regions seems less important. Opinion varied regarding how much espirit-de-corps for CRC participation by practicing physicians was dependent on local academic leadership. Information from focus group interviews with likely CRC physician participants was suggested to explore these issues.

WEB-based training methods with informed consent obtained by central telephone contact with CRC experts was proposed, as having been successfully used in internal medicine networks. Outcomes could be assessed using central telephone-based systems (e.g. the six month outcomes in the International Stroke Trial using the two question determination for residual disability).

IRB issues will be a concern to practicing physicians, who may only enter a handful of patients into a specific protocol, yet hours of paperwork/administrative time could conceivably be required to gain IRB approval. Many IRBs now charge $2000 to review a protocol. Could the CRC maintain its own central IRB?

How much on-site data monitoring is necessary to assess quality (given that many CRC physicians will be novices at clinical research) and to discourage fraud? The importance of on-site monitoring was minimized by several workshop participants as overemphasized, antequated (sophisticated analysis of submitted data can screen for fraudulent data), and of unestablished value to trial quality. It should be done, if at all, in a small sample as a deterrent to poor data quality and fraud.

A central CRC pharmacy and event adjudication group should be part of the design.

Envision and portray the CRC Coordinating Center as a "switchboard" for protocols, data, other information, and questions.

Who decides what studies are appropriate for the CRC? Should there be a protocol review committee to makes recommendations about suitability and approves CRC streamlined protocols?

Top

Post-Workshop Comments by the Co-chairs

The 21 invited participants were an extraordinarily experienced and accomplished group representing a wide range of expertise. The workshop sessions were stimulating, and the quality and detail of the comments, criticisms and recommendations exceeded expectations. We were encouraged by the unanimous enthusiasm expressed for some form of innovative CRC by these seasoned clinical researchers. Much sound advice was offered which will influence the next steps of CRC development. General impressions following the workshop included:

1. The need - even an urgent need - for an NINDS CRC involving hundreds of practicing physicians was broadly endorsed. It will be good for neurological research, for patients, and for neurology as a specialty, vanquishing remaining nihilism about neurological treatments.
2. Implementation of a CRC that will be efficient and relevant in 5-10 years will be more complex than initially anticipated. Stepwise implementation was urged: develop and define the ideal CRC, and then move toward it in incremental steps. Some important aspects can be initiated immediately (e.g. a website to link patients and physicians to research studies), while others will require more information to properly design.
3. Flexibility in the CRC structure should be inherent in its design. There should be multiple pathways linking patients to clinical research studies, several options for physician participation, and the capacity to modify and adapt the CRC structure over time. This flexibility will be challenging to incorporate, given the funding structure, the need for multi-level oversight, and the inherent inertia that besets large projects.
4. The educational components were consistently emphasized by workshop participants as key to physician participant motivation, research quality and human subject protection.

Top

Appendix I: Key issues

A. Is the proposed CRC the optimal way to accomplish these goals? What alternative models might be superior?

B. Obstacles to physician participation.

	B-1: What are the problems inhibiting participation by practicing physicians? By academic physicians? By non-neurologists?
	B-2: Novel approaches to human subjects protections and IRB review.

C. How to make participation in the attractive and meaningful, particularly for internists and family medicine physicians?

	C-1: Academic (affiliation, writing committees).
	C-2: Continuing medical education.
	C-3: Monetary (how to compete with pharmaceutical companies?).
	C-4: Altruistic (education about importance and impact of the trials).
	C-5: Espirit-de-corps (CRC sessions at major meetings, posters, tee shirts, etc).
	C-6: Certification by NINDS CRC for good research practice standards.
	C-7: Making it easy:

	- Web-based screening & enrollment; provide recruitment ads, recruitment kits.
	- Collect essential data only.
	- Web conferencing for protocol training.
	- Design studies to fit into usual treatment paradigm.

D. Appropriate studies for the CRC.

	D-1: What types of trials/studies would be appropriate for a CRC, and which would not?
	D-2: Meaningful outcome measures that can be reliably applied in this setting.
	D-3: Special issues relevant to pediatric research.

E. Coordination and Administration.

	E-1: Single data coordinating center, common data elements, central pharmacy.
	E-2: On-site data monitoring/quality assurance.
	E-3: Who should operate the CRC? Single institution/entity?
	E-4: Role for an Oversight Board?
	E-5: Efficient training and retraining.
	E-6: Logistics of continuing medical education.
	E-7: Interactions with disease advocacy groups and established research networks.

F. Budget Issues

	F-1: What is the breakpoint for economy of scale (numbers of trials, numbers of participants)?
	F-2: How to structure the funding to keep infrastructure costs modest?
	F-3: Will it really be cheaper?

G. Can clinical trials of "small diseases" be facilitated by the CRC structure?

H. How to gauge the success of the CRC.

Top