Recommendations from NHLBI Workshop

Improving Design and Conduct of Clinical Trials

February 8-9, 2001
Bethesda, MD

The NHLBI Clinical Trials Review Group was charged with developing suggestions for increasing the effectiveness and efficiency of the NHLBI Clinical Trials effort. The group included trialists with a primary interest in heart, lung, and blood diseases representing the clinical, epidemiological, and biostatistical community as well as staff from the NHLBI. This report represents the results of a one and a half day brainstorming session. It has been circulated to all the participants for their comments, and while it does not represent complete agreement on content, there are no major areas of disagreement.

The NHLBI has set the pace for the development of clinical trials in the past. We are now entering a new era in medicine characterized by:

1. An explosion of new diagnostic and therapeutic technologies fueled by NIH funded basic science, both intramural and extramural.
   
2. A continued focus with increased intensity on all elements of the disease spectrum: identification of genetic susceptibility, primary prevention, treatment of subclinical disease to prevent clinical manifestations, long-term treatment of chronic disease,and evaluation of therapies for life threatening acute manifestations.
   
3. A need to understand clinical issues in children and the elderly.
   
4. A need to link clinical trial findings to delivery of effective therapy, including to populations currently identified as under treated.
   
5. An environment in which the financial margin of clinical care has diminished to the point that institutions receiving grant and contract money from the NIH can no longer use departmental money and fees collected from patient care billing to subsidize clinical trials.
   
6. A legal environment in which the details of contracts compete with the time and effort required to develop scientifically excellent protocols.
   
7. A changed view of patients from passive recipients of health care to active consumers of health care and research.
   
8. An increased focus on both the role of local Institutional Review Boards (IRBs) and the conflict between the importance of releasing data for research and the protection of patient privacy in federally funded studies.
   

In order to meet the challenges and opportunities of this new environment, the NHLBI should consider a significant revamping of its approach to its clinical trials effort to include the following strategies:

I.  Given the major investment made by the United States in epidemiological studies, registries, and previous clinical trials, the NHLBI should develop better ways to use existing and new data to determine the need for trials and to plan them. Different policies would be needed for ongoing versus completed studies; clinical trials versus observational studies; published versus unpublished data; grants versus contracts; individual research project grants (R01s) versus Cooperative Agreements; and studies that have published only their major findings versus studies that have published both major findings and other results. This approach should balance the intellectual investment that comes from investigator-initiated efforts with the imperative to use projects supported by public funds for commonly identified national needs.

• The NHLBI should further refine mechanisms to facilitate access to existing data sets that it has funded to provide information to improve understanding of the need for new clinical trials and for planning them. This mechanism should include a formal agreement regarding publication policies, which should emphasize confidentiality as appropriate.
  
  
• The NHLBI should work with other government agencies to develop a mechanism for similar sharing of data sets among agencies and their grantees.
  
  
• The NHLBI should consider the use of registries, especially with regard to rapidly evolving therapeutic technologies, to evaluate the need for trials and to enhance understanding of the generalizability of trial results. Such registries might provide an efficient structure for adding a randomized component to the registry to evaluate the appropriateness of practice variations observed in the registry.
  
  
• Under some circumstances it may be valuable to add an intervention component to an ongoing epidemiological study funded by the NHLBI, in addition to the recognized practice of sometimes adding a significant epidemiological component to a large clinical trial.
  
  
• A more systematic approach should be developed for making the critical decisions about what to do at the end of clinical trials. Criteria should be developed for determining when continuation of follow-up for an extended period should be seriously considered. In such cases a plan should be enacted far in advance of the trial completion date.
  
  
• The NHLBI should support further methodologic work on the role of observational data analysis in determining the need for trials or for replacing the need for trials. This methodological work would be quite inexpensive relative to the cost of full trials and would mostly involve a pragmatic analysis of issues such as trial costs, estimates of potential benefits of interventions, and issues such as randomizing after early termination of a trial for efficacy and estimation of bias in nonrandomized treatment comparisons.
  
  

II.   There is no shortage of excellent ideas for clinical trials that would improve the public health. It is therefore imperative for the NHLBI to focus on redesigning its systems so that the greatest efficiency is gained in answering important clinical and public health questions.

• Formal clinical trials networks should be considered for some conditions when:
  
  
  • multiple interventions can be addressed by trials with similar structure (e.g., patient cohorts, similar protocols, manuals of operations, data collection forms); and
    
    
  • the time for each study is not prolonged; and
    
  • the number of patients needed for each study is small to moderate
    
    or
    
    when complex technology or complex behavioral interventions need to be employed.
    
    For example, sickle cell disease and congenital heart disease are conditions for which consolidated expertise could reap significant rewards. In the area of primary prevention, sentinel networks to evaluate biomarkers of atherosclerosis prevention could be helpful in sorting through the growing number of potential interventions. Because behavioral interventions, including interventions on lifestyle behaviors such as diet and exercise need to be evaluated in their complexity before their use in large, outcome-based trials, networks with behavioral expertise should be considered.
    
    These networks provide the possibility of a major public private partnership in which the infrastructure provided by the NHLBI could produce a more efficient method of doing clinical trials that would attract studies funded by industry, thus providing an opportunity for rapid resolution of the many unanswered questions about diagnostics and therapeutics.
    
    
  • Although networks are potentially of great use in the situations listed above, the NHLBI should carefully evaluate the risks and benefits of forming networks for conditions requiring large and long trials. The advantages of networks for such major problems as sudden death, heart failure, and secondary prevention of heart failure are more debatable, but a more efficient system for conducting trials in these areas is clearly needed. The network approach should not detract from the experience and record of success in the design and conduct of non-network based trials as appropriate.
    
    
  • Regardless of whether a formal clinical trial network is formed in each area, the NHLBI needs to reduce the time from the concept to trial completion by developing an organized approach to investments and standards in the following common structural elements:
    
    
  • Nomenclature -- The development of common nomenclature for case report forms would allow new trials to be initiated with considerable shortening of development of case report forms. Ideally, the NHLBI could play a central role in working towards a common nomenclature with the FDA, HCFA, the VA, and other professional organizations to reduce redundant coding by health care providers conducting clinical trials.
    
    
  • Information technology platform -- Efforts to develop common technology platforms among NHLBI coordinating centers and investigators would reduce the time to startup and enhance communication during the conduct of trials. For example, internet-based investigator meetings and joint, internet-based IRB meetings could markedly reduce time delays.
    
    
  • Pools of investigators with common training and operational procedures should be developed for each of the major disease areas
    
    
  • for certain conditions a formal network may be preferred
  • for other conditions a formal network is not needed but common procedures could significantly improve the quality and speed of the clinical trials in each area
  • common operational procedures maintained on a shared information technology platform could enhance the speed of initiation of trials by reducing the need for training
  • certification of investigators and study coordinators as "NHLBI certified investigators"
    
    

• An annual meeting for each of the major disease areas held by the NHLBI, including trialists, clinicians, patients, and policy makers could:
  
  
  • Review research priorities in that area
  • Review output of major epidemiological studies and clinical trials funded by the NHLB
  • Review appropriate nomenclature
  • Provide training about clinical research for new investigators and updates for old investigators

• The time for protocol development should be reduced by:
  
  
  • having more focused RFPs for contracts and RFAs for Cooperative Agreements
  • having a smaller group finalize the protocol after the proposed study has been funded
  • shortening time to buy in by the clinical investigators; a more effective information technology platform could allow this to happen electronically
    
    
• The NHLBI should work with OHRP, its networks, and its investigators to develop more efficient systems for centralized ethical review and effective audit of the veracity and quality of the conduct of clinical trials. This system should seek to avoid redundant and ineffective systems of review of ethics, veracity, and quality. Instead a systematic approach should be taken so that the public can be assured of high quality clinical research done with the highest ethical standards.
  
  
• The NHLBI should encourage pilot studies in certain circumstances, either separate from the main study or incorporated in the startup of the main study. Adequate and appropriate funding mechanisms should be developed to encourage pilot studies that enhance the efficiency of the main trials. This approach could hone the practical aspects of a study when existing data are not sufficient for confidence in its design or feasibility. This effort should include a reevaluation of feasibility at the conclusion of the pilot. An approach would be needed so as not to make the conduct of a pilot study an effort with intellectual or financial risk that is too great from the perspective of the investigator because of the cost of startup of a new project. The use of networks for pilots could be one such approach so that a failed pilot could be rapidly followed by another protocol.
  
  
• The NHLBI should conduct a serious evaluation of the cost of conducting clinical trials at the site level. This is an urgent need. Almost all sites in NHLBI trials believe that their activities are not compensated at a level necessary to cover the costs incurred. This leads to inordinate delays in enrollment and forces sites to conduct less desirable studies in order to subsidize NHLBI efforts.
  
  
• A standard for reimbursement for heart, lung, and blood trials would create a new ethical standard that would reduce the national worry about inducement of investigators by payments over and above reasonable levels. While "overpayment" is not typical of industry-sponsored trials, it does occur occasionally, and an equal standard could be used for both government and industry funded trials.
  
  
• Large, simple trials remain an important way to answer critical questions amenable to abbreviated data collection. The NHLBI should encourage the study of approaches to reducing costs by streamlining the collection of data in outcome based trials, so that definitive answers can be obtained at a reasonable cost. Many clinical outcome questions can only be answered through international collaboration; a better system for working with international collaboration is needed.
  
  

III.   The conduct of the best possible clinical trials in heart, lung, and blood diseases is in the interests of both the nation and the world. Because of the importance and size of this mission, the NHLBI should seek methods of optimizing the potential for collaboration between NHLBI and other organizations. It is in the best interest of the public to develop a better, more efficient approach to government interaction and to public private partnerships so that adequate resources are brought to bear on important clinical trials and so that the trials are designed and executed in the most effective manner.

• The NHLBI should work with other institutes within the NIH to develop trials that measure various disease-specific endpoints (e.g., cardiovascular disease and cancer) at different times in the same populations. Cancer and rheumatic diseases represent specific areas of opportunity where long-term therapies have cardiovascular and pulmonary consequences, and the endpoints relative to one disease occur at a different time point compared with the other diseases. The intersection of mental health and heart, lung, and blood diseases represents another important chance for collaboration among Institutes.
  
  
• The NHLBI should also work more closely with the Department of Veterans Affairs Cooperative Studies Program (CSP) in conducting joint clinical trials. Heart, lung, and blood diseases represent a significant portion of the diseases of the veteran population and the CSP has conducted 30-40 percent of its trials in these disease areas over the last 25 years.
  
  
• The NHLBI should devote immediate attention to working directly with HCFA and its investigators to facilitate implementation of the new HCFA guideline on reimbursement of clinical trials. This effort should also include ongoing efforts with the Office of the Inspector General, DHHS. This effort will probably require the creation of an office to maintain appropriate communication.
  
  
• The NHLBI should work to facilitate the co-funding of clinical trials by the medical products industry, recognizing the increasing complexity of the legal and conflict of interest environment. This may require creation of template agreements with the pharmaceutical, biological, and medical products industries to address their legitimate needs and concerns. Additionally, further clarification of the principles of public private partnership is needed. In some cases inexperienced investigators will need assistance in negotiating with industry to enhance the probability that government, academia, and industry will meet their goals without breeching the principles of public-private partnership.
  
  
• The NHLBI should work with managed care health plans to achieve greater participation of plans and their patients in NHLBI-sponsored clinical trials. Important goals are agreement on coverage of routine patient care costs by health plans, negotiation on definitions of routine care costs before initiation of trials with unusually high costs, inclusion of plan-based physicians in the conduct of trials, and involvement of health plan representatives in the planning of new trials.
  
  

IV.   When an investment is made in a major clinical trial, the potential advantage of adding additional questions that could not otherwise be answered is self-evident. In order to take maximal advantage of this opportunity:

• Planning for ancillary studies should occur along with the plans for the main clinical trial. In many cases the funding may come later, but unless the substudies are planned in conjunction with the main trial, the likelihood of their successful completion may be compromised.
  
  
• The mechanistic substudy RFA program that was recently initiated by the NHLBI should continue.
  
  
• An equivalent program is urgently needed to fund methodological research on new statistical methods and the study of the conduct of clinical trials. The field is suffering from an absence of funding for such methodological research.
  
  

V.   In order to maximize the opportunity to improve its clinical trials program, the NHLBI should consider further one-day meetings to discuss individual points in these recommendations.

Submitted by Robert Califf, M.D.
Chair
E-mail: Calif001@mc.duke.edu

Robert Califf, M.D., Workshop Chair
Duke University Medical Center
Duke Clinical Research Institute
P.O. Box 17969
Durham, NC 27715

Reuben Cherniack, M.D.
Department of Medicine
National Jewish Medical and Research
Center
PPU 3, May Building
1400 Jackson Street
Denver, CO 80206

Katherine M. Detre, M.D., Dr.P.H.
Professor, Department of Epidemiology
A531 Crabtree Hall
University of Pittsburgh
Pittsburgh, PA 15261

George Williams, Ph.D.
Vice President, Biostatistics and
Research Systems
Merck Research Laboratories
P.O. Box 4
West Point, PA 19486-0004

Janet Wittes, Ph.D.
Statistics Collaborative, Inc.
1710 Rhode Island Avenue, N.W.
Suite 200
Washington, DC 20038

NHLBI Representatives

Lawrence M. Friedman, M.D.
Special Assistant to the Director
National Heart, Lung, and Blood Institute
National Institutes of Health
Building 31, Room 5A06
31 Center Drive, MSC 2482
Bethesda, MD 20892-2482

Carl Roth, Ph.D., LL.M.
Associate Director for Scientific Program Operation
National Heart, Lung, and Blood Institute
National Institutes of Health
Building 31, Room 5A03
31 Center Drive MSC 2482
Bethesda, MD 20892-2482

Peter J. Savage, M.D.
Acting Director, Division of Epidemiology
and Clinical Applications
National Heart, Lung, and Blood Institute
National Institutes of Health
Two Rockledge Center, Room 8100
6701 Rockledge Drive, MSC 7938
Bethesda, MD 20892-7938

Julie Buring, Ph.D.
Professor of Ambulatory Care & Prevention
Harvard Medical School & Brigham & Women's Hospital
900 Commonwealth Avenue (East)
Boston, MA 02215

Gregory Burke, M.D., M.S.
Associate Professor
Department of Public Health Sciences
Bowman Gray School of Medicine
300 South Hawthorne Rd.
Winston-Salem, NC 27103

Michaele Christian, M.D.
Associate Director
Division of Cancer Treatment and Diagnosis
Cancer Therapy Evaluation Program
National Cancer Institute
6130 Executive Blvd., Room 742
Rockville, MD 20852

Jeffrey Cutler, M.D.
Director, Clinical applicarions and Prevention Program
Division of Epidemiology & Clinical Applications
National Heart, Lung, and Blood Institute
6701 Rockledge Drive, Room 8130
Bethesda, MD 20892

Joseph Feczko, M.D.
Sr. Vice President
Medical and Regulatory Operations
U.S. Pharmaceutical Group
Pfizer Inc.
235 E. 42nd St.
New York, NY 10017-5755

Robert L. Frye, M.D.
Department of Medicine
Mayo Clinic
200 First St. Southwest
Rochester, MN 55905

Saul Genuth, M.D.
School of Medicine
Case Western Reserve University
10900 Euclid Avenue
Cleveland, OH 44106-4951

Peter Greenwald, M.D.
Director, Division of Cancer Prevention
National Cancer Institute
31 Center Drive, Room 10A52
Bethesda, MD 20892-2580

Steven Haffner, M.D.
Professor
Division of Clinical Epidemiology
Department of Medicine
University of Texas Health Science Center at San Antonio
7703 Floyd Curl Dr.
San Antonio, TX 78284-7873

Al Hallstrom, Ph.D.
Professor of Biostatistics
University of Washington
1107 N.E. 45th Street, Suite 505
Seattle, WA 98195

William G. Henderson, Ph.D.
Chief, Cooperative Studies Program
Coordinating Center (151K)
Hines VA Hospital
5th Avenue and Roosevelt Road
Hines, IL 60141

James Kiley, Ph.D.
Director, DLD, NHLBI
6701 Rockledge Drive
RCKL2/Room 10018
Bethesda, MD 20892-7952

Lewis H. Kuller, M.D., Dr.P.H.
Department of Epidemiology
Univ. of Pittsburgh, GSPH
A526 Crabtree Hall, Epidemiology
130 DeSoto St.
Pittsburgh, PA 15261

Stuart J. Pocock, Ph.D.
Professor of Medical Statistics
London School of Hygiene and Tropical Medicine
Department of Medical Statistics
Keppel Street
London WC1E 7HT
England

Nancy C. Santanello, M.D., M.S.
Senior Director, Epidemiology
Merck Research Laboratories
PO Box 4, BL 1-7
West Point, PA 19486

Joseph Selby, M.D.
Director, Division of Research
Kaiser-Permanente Medical Center
3505 Braodway
Oakland, CA 94611-5714

Russell P. Tracy, Ph.D.
Laboratory for Clinical Biochemistry Research
University of Vermont - Aquatec Bldg
Rm T-205
55 A South Park Dr.
Colchester, VT 05446

Sean Tunis, M.D., MSC.
Director, Coverage and Analysis
Health Care Financing Administration
Office of Clinical Standards and Quality
Coverage & Analysis Group
Mailstop S3-02-01
7500 Security Boulevard
Baltimore, MD 21244-1850

Joel Verter, Ph.D.
Biostatistics Center
6110 Executive Blvd., Suite 750
Rockville, MD 20853

Elliott Vichinsky, M.D.
Division Head, Hematology/Oncology
Director, Comprehensive Sickle Cell Center
Children's Hospital of Oakland
747 52nd Street
Oakland, CA 94609

Robert A. Wise, M.D.
Associate Professor of Medicine
Johns Hopkins Asthma and Allergy Clinic
301 Bayview Boulevard
Baltimore, MD 21224