Final Version October 1998

REPORT OF THE NHLBI COUNCIL
WORKING GROUP ON EPIDEMIOLOGY ⁽¹⁾

Contents

Summary Statement     Conclusions and Proposed Recommendations     Goals     Strategies Preamble Program I: Importance of Epidemiology/Prevention, and Resource Value of Cohort Studies Program II: The Need for More Clearly Defined Opportunities in Translational Research Program III: Investigator-Initiated vs. Institute-Initiated Research (Support-Review-Funding) Program IV: Training of New Investigators Process I: Scientific Advisory Subcommittee of the Council Process II: Structure of DECA Process III: Staff Role

SUMMARY STATEMENT

    Concerns of the Institute:

• Epidemiologists differ in their support for continuation of long-term observational studies vs. the exploration of newer approaches to epidemiology, e.g., genetic epidemiology or subclinical disease detection.
• "Doing everything that everybody wants is very expensive" and may distort "allocation of resources among all the Institute's programs."
• "Staff should not attempt to resolve the matter."
• "Council should investigate and review the matter . . because it is one of its responsibilities to oversee the programs of the Institute."

WORKING GROUP ON EPIDEMIOLOGY

EXHIBITS AND REFERENCE MATERIAL (Not Included in the Web Version)

• Table of Organization of DECA
• Letter from Manolio regarding trends in applications to NHLBI (No evidence of decline)
• Portfolio of Research Applications
  • Summary of NHLBI Cohort Studies
  • Epidemiology and Biometry Programs Active Grants >$300,000 Direct Costs/Year
  • DECA Inventory on Genetic Research
  • Clinical Application and Prevention Program Active Contracts and Grants >$300,000 Direct Costs/Year
• NIH Policy on $500,000 (March, 1998)
• Task Force Reports and SEPs
  • NHLBI Report of the Task Force on Research in Epidemiology and Prevention of Cardiovascular Diseases
  • Report of the Task Force on Behavioral Research in Cardiovascular, Lung, and Blood Health and Disease
  • Minutes of the NHLBI Special Emphasis Panel on Longitudinal Cohort Studies

CONCLUSIONS AND PROPOSED RECOMMENDATIONS

The Working Group recognizes that the NHLBI has been an outstanding resource for research and training in cardiovascular disease epidemiology and prevention. Our goal is to suggest ways to maintain and enhance the Institute's leadership role in the field. To that end we wish first to reinforce two previous NHLBI Task Force Reports. The first is the NHLBI Report of the Task Force on Research in Epidemiology and Prevention of Cardiovascular Diseases (Aug. 1994)⁽²⁾. This Task Force chaired by Al Oberman included 18 leading experts in Epidemiology and Prevention and 4 consultants. The second is the NHLBI Report of the Task Force on Behavioral Research in Cardiovascular, Lung, and Blood Health and Disease (Feb. 1998)⁽³⁾. The report was based on deliberations of 16 leading experts and 4 consultants, chaired by Stephen Manuck.

The Working Group urges further implementation of the recommendations detailed in those two reports.

ln reviewing our charge it became apparent that there were important concerns among the leadership of the Epidemiology and Prevention Community that had to be addressed and shared with the Council and the Institute.

Therefore this report represents not only a response to our charge but also a sincere attempt to address constructively issues that are essential to the future of research in Cardiovascular Epidemiology and Prevention.

These issues are presented under 4 General Goals and 7 Programmatic and Procedural Strategies for consideration by the Council and the Institute.

GOALS

1. Explore Novel and Unique areas in Cardiovascular Epidemiology
   • Use data from longitudinal cohort studies as a resource for (1) new investigator-initiated proposals through R01 or other mechanisms; (2) inter-Institute NIH-wide initiatives; (3) cross-institutional collaborations
   • Identify genetic risk factors that influence the response to traditional and non-traditional cardiopulmonary risk factors and to interventions of these risk factors
   • Identify subclinical disease for early intervention
   • Enhance training opportunities
   
   
   
2. Provide more leadership in the development and fostering of prevention science in the area of cardiovascular diseases
   • Identify behavioral and social factors that negate prevention
   • Assure effective, meaningful translation
   • Include both individual and population-wide strategies for research
   • Address primordial prevention
   • Promote investigator-initiated clinical trials
   
   
   
3. Foster maximum application of epidemiologic and biostatistical methods as enabling strategies to the spectrum of basic and translational research in heart, lung, and blood diseases
   
   
4. Establish a scientific advisory subcommittee of the NHLBI Council to work with staff on recommending priorities and initiatives

STRATEGIES

In an effort to achieve these goals our Working Group offers four specific programs and three procedural opportunities for review by the NHLBI Council and Institute.

Program I:	Importance of Epidemiology/Prevention: Resource Value of Cohort Studies
Program II:	The Need for More Clearly Defined Opportunities in Translational Research
Program III:	Investigator-Initiated vs. Institute-Initiated Research
Program IV:	Training of New Investigators
Process I:	Scientific Advisory Subcommittee of the Council
Process II:	Structure of DECA
Process III:	Staff Role

We look forward to sharing and discussing the report.

NHLBI Council Working Group on Epidemiology

PREAMBLE:

Our ultimate goal as NHLBI Council members, and as physicians and scientists is to advise our government on the best way to decrease the people's sufferings, morbidity, and mortality from Cardiovascular Disease.

In the midst of staggering advances in molecular biology and genetics it is our collective responsibility to maintain a balanced perspective and to recognize the continuing enormous importance of the science of Epidemiology and Prevention of CVD among the spectrum of approaches to achieve our ultimate goal.

Bridges between approaches:

Progress in the understanding and control of cardiovascular diseases must be predicated on an appreciation that no single approach or methodology can provide the sole answer to our questions. An explanatory model that continually focuses at an increasingly molecular level, while providing important insights, in itself may provide incomplete information relevant to the control of cardiovascular disease. Furthermore, it must be recognized that social, behavioral, and economic forces are the distal causes of the proximal, molecular mechanisms in which there is so much interest. That means that a full understanding of the causes and means of prevention of cardiovascular disease is only present when there are bridges built between all levels of investigation.

For example the different molecular/genetic defects that can cause various types of hypertension or obesity will have much greater significance once they can be coupled to specific phenotypes first in transgenic mouse models but eventually and most importantly in human population studies. Another example may be the phenotypic expression of a defective gene that becomes evident only if enhanced by nutritional factors related possibly to ethnicity.

The bridge is not just a one-way bridge from basic research to epidemiology. Not infrequently epidemiologic studies raise possibilities for laboratory research and molecular mechanisms, e.g., AIDS was originally described through epidemiological studies that led to basic research.

It is similarly critical that a focus on the early detection of cardiovascular disease and on its prevention in the community be retained. While some investigations are better pursued with small, unrepresentative samples they are not likely, in isolation, to lead to global truths because their true significance can only be valued within a perspective that assesses their importance within the population.

Within the preceding perspective, the population-based, prospective epidemiologic study provides a unique window on the potential causes and prevention of CVD. It provides a context in which the molecular and the social/behavioral knowledge can be linked, in which the determinants of risk factors can be studied, and in which the significance and importance in the population of etiologic findings can be measured. Without such studies we cannot hope to obtain the complete level of understanding necessary to realize our greatest potential in the prevention and cure of heart disease.

Overall, we view epidemiologic research as an enabling science in the broadest sense and would hope that this broad view would be shared by the NHLBI leadership and staff. As noted by Drs. Califf, Pearson, and Hennekens in their July 22, 1998 letter to Dr. Lenfant, as circulated to the SPARK Working Group, epidemiologic and biostatistical methods have a unique and complementary role in advancing knowledge of heart, lung, and blood diseases not only through the traditional route of population research (e.g., descriptive and longitudinal observational epidemiologic studies and randomized clinical trials), but also through patient-oriented clinical research and health services research.

Program I

Importance of Epidemiology/Prevention

Resource Value of Cohort Studies

Importance of Epidemiology/Prevention:

Longitudinal, population-based cohort studies constitute an essential component of the spectrum of scientific inquiry.

They provide unique aspects of scientific information on etiology, natural history, and consequences of disease, as well as the generalizability and import of this information for human populations, by studying both asymptomatic and symptomatic persons.

They define the predictive role of both short- and long-term risk factors in human disease and the potential approaches to prevention of disease likely to be effective for populations and subsets of the population.

They provide the opportunity to address important multidisciplinary questions about complex biologic processes as well as to understand the multifactorial origin of chronic diseases, and lay the basis for identification of previously unrecognized risk factors with fewer sources of bias and spurious results.

They permit understanding of the complex interaction between genetic and environmental/behavioral factors and the subsets of the population that may be at risk for disease or related adverse outcomes.

Thus, they not only provide a critical link in scientific understanding of how molecular biologic research could affect the future health of populations, but more broadly the range of most effective approaches for prevention of disease.

As reported in the 1995 NHLBI Special Emphasis Panel on Longitudinal Cohort Studies⁴ such prospective studies remain essential for identification of new risk factors for, and consequences of, cardiac, pulmonary, renal, cerebrovascular, peripheral vascular, and hematologic diseases. They provide insights into potential areas for most effective prevention, most cost-effective approaches to screening and intervention, determining the impact of medical care on changes in outcomes over time and the effectiveness of implementation of recommended medical therapies and preventive interventions.

Areas in which such research is particularly needed:

Areas in which research is particularly needed have been identified by two NHLBI task forces. Their reports review recent advances in research, identify high priority research opportunities, and make specific recommendations regarding future research. These reports have been referred to previously, in the Summary Statement, Conclusions and Proposed Recommendations. Staff should be encouraged to consider how these recommendations help to fulfill the NHLBI mission and how best to integrate them into the ongoing planning on DECA as well as Institute-wide activities.

Longitudinal cohort studies are important to helping us understand:

• the study of subclinical disease (prognostic import, measurement, and risk factors for its occurrence and progression);
• triggering of acute events;
• development of better, standardized definitions and characterizations of angina, congestive heart failure and renal disease; pulmonary and hematologic conditions;
• the role of diet and physical activity in these health outcomes (and opportunities for improving measurement), as well as other understudied behavioral factors including job strain, social networks and support;
• cross-disciplinary areas;
• the differences in risk factors, natural history and disease outcomes for different racial, ethnic, and socioeconomic groups, as well as the impact of differential receipt/use of medical care;
• role of environment and culture in adoption of health behaviors;
• utilization of genetic and molecular biologic studies within such longitudinal cohort studies to define, in particular, the nature of gene-by-environment interactions, and genetically at-risk subsets of the population, as well as enhance understanding of etiology;
• understanding the effects of cardiovascular, pulmonary and blood diseases on physical disability and cognitive function; the potential for prevention of vascular dementia;
• understanding complex biologic processes that result from interaction of several diseases, e.g. inflammation and diabetes; atherosclerosis and hypertension; dementia, aging and cerebrovascular diseases.

Resource Value of Cohort Studies:

Results of longitudinal cohort studies also provide a very rich substrate for important new questions that were not necessarily anticipated at the initiation of the study. Such a use of established populations is a potentially cost-effective way to develop new scientific areas. To maximize the yield from established cohorts a range of approaches, building on SEP recommendations, is proposed:

1. 
   Obtain at regular intervals, the scientific view of experts in the field and related fields to review the insights that have been provided by recently completed epidemiologic studies and analyses, clarify areas appropriate for direct progression to translational research and define important questions to be pursued in further laboratory or population studies.
   
   
2. Ensure maximal analysis of data sets.
   
   
   • Core investigators dedicated to answering the initial objectives of a study would be, if productive, most likely to address new goals because of their knowledge of the data set. These potential efficiencies should be considered. Time for scientifically mining these data sets should be considered a legitimate cost in support of these studies.
     
     
   • A formal mechanism needs to be established to make these data more widely available to the scientific community. It is unlikely that just placing a database on a web site will lead to substantial- or meaningful-usage. The data bases for these studies are generally highly complex and thus preclude most investigators from being able to accurately conduct analyses on their own. Therefore, mechanisms to enhance the analyses of these data by external investigators might need to include core support to a unit most effective in supporting these needs; for example the study's biostatistical coordinating center could be charged with working with these investigators, and providing necessary analytic support. Such a centralized approach would help ensure the validity of the conclusions reached by investigators not directly involved in the database creation.
     
     
   • Encourage the development, and fiscal support, of collaborative research groups -- from diverse universities -- organized to address related research questions.
   
   
   
3. Encourage the development of an RO1 mechanism to extend longitudinal cohort studies at single sites. This will likely involve ensuring that study sections are receptive to this, and understand how to evaluate manuscript production in such studies.
   
   
4. Use the cohort studies as an NIH-wide resource. Establish a cross-Institute mechanism to examine potential of using results of certain cohort studies to address important scientific questions generated from other Institutes.

Program II

The Need for More Clearly Defined
Opportunities in Translational Research

Better Translation of Research to the Community

There is an increasing gap between basic knowledge and applied prevention.

Important information about prevention and treatment of CVD already available (NEJM 1997;337:1360) must be applied more systematically and broadly.

Many of the current preventive measures, known to be effective, are not applied widely or effectively; for example, the inadequate control of hypertension and inadequate use of lipid lowering agents coupled with an estimated 50% long-term adherence to a drug regimen for those who do initiate treatment.

The NHLBI should continue to develop and accelerate strategic initiatives to maximize (1) the evaluation, in clinical and community populations, of promising primary and secondary prevention hypotheses derived from observational studies and (2) the scientific investigation of ways to address problems in prevention, i.e., stimulating the necessary science at whatever level is appropriate to the question.

Behavioral and life style factors are major causes of morbidity and mortality, contributing to all the leading causes of death in the US. Influencing people to adopt healthy behaviors cannot be achieved without support for basic, clinical, and applied research efforts.

Emphasis is shifting to emerging nontraditional risk factors, outcome measures, and patients in the community. Research must be incorporated more effectively and quickly into preventive medical practice in the community and public health practice at the state and local level.

Research in genetic epidemiology incorporating both fundamental biology and population science is likely to be very rewarding in the coming decade. It requires demonstration and education research to achieve the safety and cost-effectiveness. The true value of genetic advances to human health can only be understood through application in studies of human populations. However, this should not be applied wholesale, but only when ready for translation.

Ultimately the medical implications and economic savings can be huge with better translation of research to the community.

The Working Group recommends the following:

• Obtain sustained and broad external advice on both clinical trial planning and prevention research and policy. Workshops can not fulfill this need because of their limited scope and duration.
• Coordinate research efforts between observational studies and intervention trials. This is best done at the NHLBI level because monitoring exciting new findings in one domain could be picked up as appropriate to other types of studies.
• To the extent that there may be concerns about possible overlap with the missions of other NIH Institutes or Federal agencies when stimulating or undertaking research to meet certain prevention goals, clarify the real and perceived boundaries on what constitutes appropriate research for the NHLBI and seek opportunities for formal partnerships with agencies and institutes that have complementary missions (e.g., NHLBI-CDC; NHLBI-Am Lung Assoc., etc.)
• The Institute should ensure that the recommendations on prevention from major studies are synthesized and disseminated.
• Develop a brief section in "continuing" applications where investigators could identify findings likely to be helpful to other domains.
• Address behavioral and social issues that limit prevention. Several such issues can be addressed much more effectively if the scientific interactions between the various programs of DECA were enhanced and the research plans of the scientific research group were better integrated or coordinated, e.g.,
  • Difficulty in generating epidemiology studies that include substantial and novel behavioral input/hypotheses.
  • Infrequent or delayed use of novel cohort findings to promote the objectives and design of studies in prevention and clinical trials.
  • Social and environmental epidemiology needs to focus on both group-level variables (e.g., policies, community characteristics) and individual-level variables and interactions between them.
  • A process should be planned for keeping abreast of emerging science in non-epidemiology disciplines (other than nutrition) and their incorporation into the DECA activities.

Program III

Investigator-Initiated vs. Institute-Initiated Research
(Support-Review-Funding)

The future of the field and the vision for research come from the scientific community, that is, investigators who are on the front line of the clinical and public health problems to be resolved through CVD research and have the motivation and vision to ask creative, relevant questions on a broad range of CVD topics. Success will depend however on a strong partnership with a scientifically astute and managerially skilled, committed and responsive staff. Staff will continue to be essential for the administrative oversight and facilitation of an efficient research program of the highest possible quality.

There is a perceived gradual increase in the influence of NHLBI staff on the planning and direction of CVD epidemiology and prevention without appropriate involvement from the scientific community. This may have resulted from the loss of standing advisory committees to NHLBI, and a relative increase in contracts and cooperative agreements resulting in part from conversion of large investigator-initiated studies to cooperative agreements. There is also a perceived emphasis on scientific productivity by the staff.

Investigator-initiated research support continues to provide the optimal mechanism to stimulate creativity and quality from the scientific community. The tendency to view separately the epidemiologic research program of Institute-initiated studies from those that are Investigator-initiated is an artificial division and may be misleading as to the status of science in a given content area. Potentially understudied and overstudied areas as well as potential intersections and economies among funded studies in the Institute's epidemiologic study portfolios can best be identified by considering the totality of funded research devoted to the topic in question.

RFAs and RFPs should be initiated in priority areas where Investigator-initiated mechanisms fail to stimulate scientific initiatives. There seems to be a need also for a mechanism, which stimulates multicenter collaborative proposals in priority areas without the constraints of contractual or cooperative agreement mechanisms.

It is recognized that major longitudinal inter-Institutional studies are likely to be initiated by Institute staff. However, the initiation of single-site longitudinal cohort studies and the continuation of some cohort studies should be encouraged through an investigator-initiated mechanism.

Mechanisms of review and support:

In considering the role of epidemiologic studies at NHLBI, it is important to remember the many cohort studies are funded through extramural, investigator-initiated efforts in addition to those that are Institute-initiated. In many cases, these studies have provided critical and important information. The record of publication of many of these studies is comparable to that coming from the multi-center, Institute-based studies.

The requirement to seek administrative approval for studies over $500,000 limits may represent an obstacle to such initiatives. Concomitant with the administrative review by staff a scientific peer perspective may be obtained from the proposed Scientific Advisory Subcommittee of the Council. This may be necessary to counter the perceived bias against continuation of established studies. The same Council subcommittee could consider proposals of cross-discipline projects. One model might be a program project grant, with different projects funded by different Institutes and core NIH funding.

For the Investigator-initiated approach to be a viable one for support of longitudinal cohort studies, it would be appropriate to consider a period of support that is longer than 5 years but with some interim peer review for conditional continuation. Support for a longer period would greatly aid the investigators.

Careful consideration needs to be given to a more efficient mechanism by which RFPs are generated, reviewed and approved. The lengthy processes of pre-award submission and resubmission of RFPs and their redesign after a successful bid occur sometime at a great cost to the potential investigators. It is recognized that the process is mandated but there are elements that are repetitive, time consuming, and ultimately unproductive in that they involved detailed development and revision of protocols and plans that are never utilized.

Another area for consideration is the lack of suitable funding mechanisms for individual investigators to benefit from longitudinal data from cohort studies in conjunction with coordinating center personnel who can provide data and guidance about proper use of the data. Many of the fortunately still numerous publications generated by studies such as ARIC (Atherosclerosis Risk in Communities) and CHS (Cardiovascular Health Study) are based on unfunded efforts of investigators who are interested in questions that can be asked of the data. Because these efforts are unfunded they are sometimes very slow in progressing and may be compromised by the inability of the coordinating center to donate the necessary time and expertise.

Program IV

Training of New Investigators

The goal of the Institute should be to develop the needed scientific basis for epidemiology and prevention and to train the scientists necessary to create this scientific base. Such training requires emphasis on epidemiology, biostatistics, and study design -- from basic to clinical research. Population studies allow great opportunities for training clinical researchers and others by virtue of their collaborative nature involving multiple centers and by implementing of epidemiologic techniques for design, standardization of procedures and analysis. The requirement for investigators from various disciplines seems integral to the evolving concepts of CVD.

• The Need for Review of the NHLBI Training Portfolio

There is a need to assess the current population of trainees in cardiovascular epidemiology and prevention and related disciplines, both in terms of quantity and quality. It is unclear if the needs are met by current training programs. Various estimates by scientific organizations suggest that there is a deficit of formally trained epidemiologists and biostatisticians in the US. Those involved with recruiting and hiring these scientists concur with this conclusion.

It is also unclear if current training programs are training in subject areas in which manpower needs are likely to be the greatest. If we are going to succeed in fostering genomics, computational biology, emerging technologies, and imaging, population scientists will be needed in these areas. These new investigators will not only need skills in biostatistics, epidemiology, health services research, social and behavioral sciences, and other quantitative methods, but will also need to be able to apply these methodologic disciplines to prevention related research in addition to classical epidemiology and biostatistical methodology.

There is also a great need for training in genetics for individuals trained as epidemiologists to address the future compelling questions in genetic epidemiology.

• Use NHLBI-initiated Studies as a Training Opportunity

One creative approach to both maximizing research from large NHLBI-initiated studies and providing training opportunities is to encourage and support involvement of fellows and/or graduate students on projects sponsored by contracts and cooperative agreements. Currently, if budgets are perceived as tight, no funds are set aside to support graduate students or fellows (unlike bench-related research where this is a ubiquitous practice). The provision of additional funds (not reallocated moneys) for trainees would benefit both the projects and the trainees.

The use of such a mechanism to attract new investigators to ongoing studies for the purpose of initiating their careers is recommended. It is increasingly difficult for persons just out of training to establish careers in epidemiology and/or clinical research. While in bench sciences, an institution can commit finite resources to create a laboratory for a new investigator; it is difficult to create an epidemiology "laboratory" to allow the investigator to create a track record in epidemiology and to develop novel ideas. The attachment of such new investigators to large, established epidemiologic databases and/or prevention programs will allow such career development.

The current stipend and salary levels may be low for trainees, discouraging persons from pursuing fellowships or early faculty career development awards. Review of NHLBI programs to assure concurrence with other NIH Institutes and with the job market may be necessary.

Process I

Scientific Advisory Subcommittee of the Council

Background: Concerns Regarding Review and Long Range Planning Process

External Review of NHLBI-initiated Research. Suggestions were made to enhance the accountability of initiatives proposed by NHLBI staff. In the same way that externally initiated applications undergo peer review, there should be a mechanism to review internally initiated contracts. A standing body of advisors in epidemiology and prevention research would review major NHLBI studies for the purpose of identifying new ideas and opportunities and advise the Council. Such a group would be independent of the studies' data safety and monitoring boards.

Creativity and Novelty. One concern about long term epidemiologic studies and clinical trials is that funding is often limited to low risk applications, which, by the time the data collection is complete, no longer provide an expansion of knowledge. Mechanisms are needed to infuse creativity and risk-taking into clinical trials and epidemiologic studies. One comment made was the contrast between industry-supported trials as creators of new knowledge versus NHLBI trials which are more likely to confirm results already presented by others.

An issue frequently raised is the rule requiring staff preview of applications which had a cost of $500,000 or greater. This was felt to reduce investigator initiative and to dampen creativity in favor of "safe science".

Sustained and broad external advice is needed on both clinical trial planning and prevention research and policy. Workshops and SEPs can not fulfill this need because of their limited scope and duration. There should be a concerted effort to coordinate research efforts between observational studies, and interventions best done at the NHLBI level. Exciting new findings in one domain could be picked up as appropriate to other types of studies. Continuation applications could have a brief section where investigators could identify findings likely to be helpful to other domains.

Rationale for Creation of Subcommittee of the Council:

NHLBI staff and the quality of research generated by the Institute have suffered because of the loss of regular and sustained input from the scientific community. Such input which had worked effectively in the past is essential if NHLBI staff are to perform effectively their jobs and quality research in the interest of the health of the public is to be supported.

Composition. The committee would comprise ten members from the scientific community with recognized skills and expertise in the disciplines represented by DECA as well as other disciplines.

Selection. Professional organizations in the areas of epidemiology, public health, statistics, medicine and behavioral science would nominate individuals from their organizations. Institute officials would select from that group as well as other at-large sources, to obtain representation from various disciplines. Recommendations would be approved by Council. Council members with expertise in these areas could be asked to serve on this committee.

Four-year terms with overlapping membership would provide sustained input.

Frequency: The group would meet three times per year.

Charge: To provide scientific input to staff and advice to Council as follows:

1. Review and recommend Institute initiated research, collaborative agreements and contracts.
   
   
2. Review and recommend to Council plans for termination of contracts and collaborative agreements.
   
   
3. Provide scientific input to staff and Council on Investigator-initiated proposals over $500,000.
   
   
4. Advise staff and Council on areas/topics of future research.
   
   
5. Provide and maintain a long-range perspective on areas of research priorities.

    Process II

Structure of DECA

The Working Group recommends a review of the structure and operations of DECA to optimize a broader representation and greater communication among related disciplines that contribute to the missions of epidemiology and prevention.

Current Structure of DECA

The Division provides administrative oversight of three programs related to epidemiology, prevention and biostatistics. The epidemiology and clinical trials as well as prevention programs related to diseases of the lung and blood are not administered through DECA, but rather through the Lung and Blood Divisions, respectively.

The Epidemiology and Biometry Program directed by Teri Manolio has several groups including the social and environmental epidemiology scientific research group; the analytical resources scientific research group; the field studies and clinical epidemiology scientific research group; the Framingham epidemiology research unit; and the genetic epidemiology research unit.

The second program on Clinical Application and Prevention directed by Jeff Cutler includes the prevention scientific research group; a clinical trials scientific research group and a behavioral medicine scientific research group. It is our understanding that the behavioral medicine research group focuses on basic research and clinical trials. The resources and perspectives of that group can be very useful for other research groups and other divisions.

The third program is on Biostatistics directed by Nancy Geller. It is a resource for all programs and overlaps in function with the analytical resources scientific research group. In addition, there is the nutrition coordinator who reports directly to the director of DECA and to the Director of NHLBI.

Comments:

The Working Group believes that the breadth of what is defined as epidemiology at the NHLBI seems too tied to the discipline of epidemiology. Future directions should be broadened to include all the goals and programmatic initiatives outlined above.

The major limitations may be both structural and operational.

The expectation is that a revised structure would allow much more frequent communication and exchange of ideas among the staff, more frequent attendance at scientific meetings and more time for interaction and consultation with the scientific constituency. If a significant change in structure is not optimal then a process for more frequent interactions and integrated program planning among the various scientific research groups within DECA and other divisions is essential.

Sharing of expertise and resources for epidemiologic and intervention studies within DECA and between DECA and the Lung and Blood Divisions would be beneficial to all, increase efficiency, and create a more cohesive environment.

A greater integrative influence should be exerted by the Director and Deputy Director over the efforts at the Program and research group levels.

        Process III

Staff Role

The role of the staff as managers of science vs. a collaborating scientists should be addressed. Under most circumstances the managerial role would be the primary responsibility of the staff. However the staff needs to maintain scientific interest and expertise to be a most effective advocate for the programs. It does seem reasonable that opportunities for collaborative science be made available to the staff.

Several concerns were expressed regarding a potential conflict of interest which may be present in the involvement of the contract or cooperative agreement manager as a participating scientist. The interaction of the staff with the investigators on a project is one of accountability and control over budget when the staff is functioning as manager of the contract or cooperative agreement. This contrasts with the collegial interaction expected from a co-investigator or a collaborative scientist. This potential conflict should be addressed by a policy to be developed by the Institute defining guidelines under which scientific collaboration would be acceptable.

Clarification of the role of and expectations from staff will promote the essential .cooperative and mutually supportive partnership between staff and investigators.

Footnotes

1. An Addendum to this Report, dated February 1999, was prepared as a follow-up to the NHLBI response to this Report. [Return to text]

2. Recommendations are listed on p. 12 and 13 of the Executive Summary. They include five strategic emphases and eight priority areas for both epidemiological research on causation and enhanced application of already available knowledge for primary and secondary prevention of cardiovascular disease. [Return to text]

3. On pages 20-25 of that Task Force Report are listed 12 recommendations including the incorporation of behavioral science into existing epidemiological and clinical studies, developing interventions to sustain improvements in lifestyle risk factors and adherence to treatment, and incorporate genetic research strategies. (February, 1998). [Return to text]

4. Refer to Minutes of the NHLBI Special Emphasis Panel on Longitudinal Cohort Studies (Sept. 5-6, 1995). [Return to text]