NHLBI INTEGRATED BIOSTATISTICAL TRAINING PROGRAM FOR CVD RESEARCH

North Carolina State University
Department of Statistics
http://www.stat.ncsu.edu

Duke University
Duke Clinical Research Institute
http://www.dcri.duke.edu

Director: Marie Davidian, Ph.D.
Address, phone, e-mail

Co-Director: Elizabeth DeLong, Ph.D.
Address, phone, e-mail

Abstract

Starting in May 2006, the Department of Statistics at North Carolina University (NCSU) and the Duke Clinical Research Institute (DCRI) at Duke University are jointly administering an integrated program for predoctoral training in biostatistics to prepare trainees for careers in medical research, with a focus on cardiovascular disease (CVD) research. The shortage of skilled biostatisticians equipped to address emerging challenges in this exciting new era of CVD research calls for training that formally integrates (i) in-depth experience in collaboration in a multidisciplinary environment, (ii) mastery of the theoretical underpinnings of statistics required for valid application of sophisticated biostatistical techniques and for research on development of new methodology, and (iii) emphasis on communication and leadership skills. The program capitalizes on the long-standing partnership between one of the largest graduate programs in statistics in the world (NCSU) and a research institution that is the largest of its kind and home to internationally-known researchers at the forefront of CVD research (DCRI). This partnership affords trainees the unique opportunity to develop all of these skills through interaction with faculty at both universities, who themselves have a history of inter-institutional collaboration and research and who have extensive experience in training and mentoring.

The training involves formal coursework at NCSU on statistical theory, including probability, inference, linear and other statistical models, measure theory and advanced probability, and advanced statistical inference, and statistical methods, including clinical trials design/analysis, longitudinal data analysis, survival analysis, epidemiology, and cutting-edge special topics, such as causal inference; and at DCRI in fundamental aspects of CVD science and in research responsibility and ethics. There is also extensive formal and experiential training in communication and leadership skills at both institutions. Trainees are introduced to DCRI CVD research gradually and will evolve over their tenures to holding full collaborative apprenticeships in which they are fully integrated as functioning members of DCRI project teams. The Apprenticeships will provide trainees with extensive working knowledge of CVD science, the opportunity to develop collaborative skills, and the recognition of how new biostatistical methods development follows from challenges encountered in the collaborative context. This last point will be emphasized through mechanisms under which statistical methodological challenges arising in trainees' apprenticeships will lead to doctoral dissertation research in biostatistics. .

Trainees are assigned an inter-institutional mentorship team consisting of a biostatistician mentor at NCSU, with responsibility for guiding academic progress; a biostatistician mentor at DCRI, with responsibility for overseeing collaborative experiences; and a clinician mentor, who serves as a resource for experiential training in CVD science and research. A biostatistician "mentor-in training," who is one of several Assistant Professors with less than 4 years experience, will augment most teams, assisting the biostatistician mentor, and will serve as a role model close in professional age to the trainees themselves.

Areas of Special Emphasis

The overarching goal of the program is to prepare trainees to excel as both biostatistical methodologists and biostatistical collaborators with extensive knowledge of CVD science and of the types of data analytic challenges arising in CVD research. Accordingly, the training emphasizes both foundations of and new developments in biostatistical theory and methodology and key studies and concepts in CVD research. Areas of emphasis include randomized CVD clinical trials and associated biostatistical methodology for design and analysis; CVD research in health outcomes, health policy, quality-of-care, and medical economics and relevant biostatistical methods; and biostatistical methods for observational database analysis applied in the CVD context. Specific areas of emphasis in biostatistics include survival analysis; longitudinal data analysis; methods for clinical trial design and analysis, including sequential monitoring; general statistical modeling, including nonparametric smoothing methods and model selection; methods for missing and mismeasured data; and methods for analysis of observational data and causal inference.

Type of Training: Pre-doctoral and Post-doctoral

Key Faculty Available as Preceptors

Andrew S. Allen, Ph.D., is Assistant Professor of Biostatistics and Bioinformatics at Duke University. Dr. Allen's biostatistical research focuses on developing new statistical methodology for mapping complex disease genes using linkage disequilibrium methods. This work is related to and motivated by DCRI projects on genomic and proteomic issues and the genetic substudies of several DCRI CVD clinical trials in which Dr. Allen is involved

Huiman X. Barnhart, Ph.D., is Associate Professor of Biostatistics and Bioinformatics at Duke University. Dr. Barnhart joined the DCRI in 2003 after 11 years on the faculty at the Department of Biostatistics in The Rollins School of Public Health of Emory University, where she directed 4 PhD students. Her extensive experience in longitudinal studies and randomized clinical trials in CVD inspired her research on methods for multivariate random length data such as number of multiple stenosis and percents of the multiple stenosis in a trial assessing impact of a cholesterol lowering drug on progression of coronary artery disease. Her current research focuses on methods for assessing agreement among methods, devices, or observers in medical studies.

Dennis D. Boos, Ph.D., Professor of Statistics at NCSU. Dr. Boos is a noted statistical methodologist who has made significant contributions to foundations of statistical inference, to the theory and practice of bootstrap and permutation methods, and to methods for model selection. Much of his research has focused on important problems arising in the analysis of multi-center trials. He is a Fellow of the American Statistical Association (ASA) and has directed the doctoral dissertation research of 18 students.

Robert M. Califf, M.D., is Vice Chancellor for Clinical Research at Duke University Medical Center, Director of DCRI, Donald Fortin Professor of Cardiology, and Professor of Medicine, all at Duke University. Dr. Califf is an internationally-known leader in clinical and CVD research with diverse interests in clinical trials, health outcomes research, health policy, and medical economics, and is a recognized authority on evidence-based medicine. His primary CVD research interests are in chronic coronary artery disease, heart failure, and acute coronary syndromes. Dr. Califf is Editor-in-Chief of the American Heart Journal, has led the Coordinating Center effort for some of the most important clinical trials in CVD medicine, including the GUSTO Trials, PURSUIT, VALIANT, CARS and the EPIC, EPILOG, and EPISTENT Trials. Dr. Califf has served on the Cardio-renal Advisory Panel of the FDA, the Pharmaceutical Roundtable and Clinical Research Roundtable of the Institute of Medicine (IOM), and the IOM Committee that successfully recommended Medicare coverage of clinical trials.

Marie Davidian, Ph.D., is William Neal Reynolds Distinguished Professor of Statistics at NCSU and Adjunct Professor of Biostatistics and Bioinformatics at Duke University. Dr. Davidian is internationally-recognized for her research in longitudinal data analysis, pharmacokinetic analysis, and clinical trials. She has served as Coordinating Editor and Executive Editor of the journal Biometrics, a primary outlet for biostatistical methodological research; in leadership roles in the ASA and the International Biometric Society; and on advisory committees for the Food and Drug Administration (FDA) and the National Institutes of Health. She has served as doctoral dissertation advisor to 20 students and is a Fellow of the ASA. Dr. Davidian serves as Director of the training program, overseeing academic aspects at NCSU and coordination of all joint NCSU-DCRI activities.

Elizabeth DeLong, Ph.D., is Associate Professor of Biostatistics and Bioinformatics at Duke University, and Co-Director of the Outcomes Research and Assessment Group at DCRI. She is a leader in the field of outcomes and quality-of-care research, bringing more than 20 years of biostatistical, clinical research, and bioinformatics experience to the DCRI. Dr. DeLong is the lead statistician for in the DCRI Outcomes Group and has served as mentor for numerous DCRI biostatisticians and physician-scientists, the latter through her role in the Duke Clinical Research Training Program. Dr.DeLong serves as Co-Director of the training program, overseeing all DCRI activities.

Sujit K. Ghosh, Ph.D., is Associate Professor at NCSU. Dr. Ghosh's research focuses on Bayesian statistical inference and its application to biomedical and environmental research using modern computational tools such as Markov chain Monte Carlo (MCMC). He has directed the dissertation research of 13 PhD students.

Christopher Granger, M.D., is Associate Professor of Medicine at Duke University and Director of the Cardiac Care Unit at Duke University Medical Center. Dr. Granger's research interest is in large randomized clinical trials focusing on acute ischemic heart disease and heart failure. He has served in leadership roles on numerous high-profile CVD trials, and serves on the steering committees of several important CVD registries.

Robert A. Harrington, M.D., is Professor of Medicine and Director of Cardiovascular Clinical Trials, DCRI. Dr. Harrington's main research interests are in evaluating antithrombotic therapies to treat acute ischemic heart disease and to minimize the acute complications of percutaneous coronary procedures. He is actively involved in studying the mechanism of disease in ACS, in understanding issues of risk stratification in the care of patients with acute is chemic coronary syndromes, and in improving the methodology of large clinical trials. He has served as the lead investigator on several of the largest randomized clinical trials in ACS patients, including PURSUIT, PARAGON B, and SYMPHONY, and has been investigator on many others. Dr. Harrington has served as primary research mentor for 18 cardiovascular fellows and 2 Duke medical students, more that half of whom have gone on to academic appointments and investigative careers.

Jacqueline M. Hughes-Oliver, Ph.D., is Professor of Statistics at NCSU. Dr. Hughes-Oliver has made important contributions to statistical methods for drug discovery and environmental health and actively collaborates with scientists at GlaxoSmithKline and several biotechnology companies. She is the director of a large multidisciplinary project to develop statistical and computational methods to identify compounds that may be developed into medicines. Dr.Hughes-Oliver has directed 12 doctoral dissertations.

Andrzej Kosinski, Ph.D., is Associate Professor of Biostatistics and Bioinformatics at Duke University. Dr. Kosinski joined DCRI after over a decade at the Department of Biostatistics in The Rollins School of Public Health of Emory University, where he supervised 2 doctoral students. He has extensive experience in clinical trials, having directed the data coordinating centers for several important studies. His statistical research is in evaluation of diagnostic tests, adjustment for misclassification of exposure, and outlier detection in multivariate data.

Wenbin Lu, Ph.D., is Assistant Professor of Statistics at NCSU. Dr. Lu's research focuses on new methods for regression analysis of multivariate time-to-event data, including recurrent events, and on methods for model selection in the survival analysis. He is directing 1 doctoral dissertation.

L. Kristin Newby, M.D., is Associate Professor of Medicine at Duke University and Associate Director of the Cardiac Care Unit at Duke University Medical Center. She is an expert on the role of hormone replacement therapy for CVD prevention, and her research interests include risk stratification of ACS patients and use of protein biomarkers to stratify risk and guide therapy. She has also served as the coordinating center lead investigator for numerous DCRI studies.

Eric Peterson, M.D., is Associate Professor of Medicine at Duke University , Director of Cardiovascular Outcomes Research and Quality at DCRI, and Senior Fellow in the Division of Aging and Human Development at Duke University Medical Center. Dr. Peterson is a clinical cardiologist with formal quantitative training including a MPH from Harvard School of Public Health with emphasis in epidemiology and biostatistics. He has considerable health services research experience, including a Paul Beeson Faculty Scholar Research Award to study revascularization in the elderly, and he is a lead investigator on a number of registry projects as well as member of numerous government and professional advisory committees.

Leonard A. Stefanski, Ph.D., is Professor of Statistics, Co-Director of Statistics Graduate Programs, and Assistant Head of the Department of Statistics, all at NCSU. Dr. Stefanski is a world authority on statistical inference in the presence of measurement error, and is the originator of the popular simulation-extrapolation (SIMEX) method, which has seen widespread application. He is a Fellow of the ASA and has extensive experience collaborating with scientists at the Environmental Protection Agency's research laboratories in Research Triangle Park, NC, on problems related to exposure to environmental pollutants and contaminants. He has directed 21 doctoral dissertations.

Anastasios A. Tsiatis, Ph.D., is Drexel Professor of Statistics at NCSU and Adjunct Professor of Biostatistics and Bioinformatics, Duke University. Dr. Tsiatis is recognized worldwide for fundamental work in survival analysis, group sequential clinical trials, causal inference,and joint modeling of survival and longitudinal data. He is a recipient of the Mortimer Spiegelman Award from the American Public Health Association and is a Fellow of the ASA and the Institute of Mathematical Statistics. He is the recipient of a Method to Extend Research in Time (MERIT) Award from the National Institute of Allergy and Infectious Diseases. Dr. Tsiatis has served on numerous Data Safety and Monitoring Boards as well as on FDA advisory committees. He has served as doctoral dissertation advisor to 32 students.

Jung-Ying Tzeng, Ph.D., Assistant Professor of Statistics, NCSU. Dr. Tzeng's research focuses on haplotype-based association analysis and on how optimally to incorporate abundant DNA sequence information to improve and generalize existing association analysis tools, and generalization of association tests based on haplotype similarity that will allow elucidation of gene-disease association on the basis of complex models. She is currently directing 1 doctoral dissertation.

Daowen Zhang, Ph.D., Associate Professor of Statistics, NCSU. Dr. Zhang is a worldwide authority on methods for the analysis of longitudinal data using smoothing techniques. He has developed important new statistical methods for the analysis of hormone data that have uncovered important features that could not be identified using standard techniques. Dr. Zhang has served as doctoral dissertation advisor to 6 students.

Hao Helen Zhang, Ph.D., Assistant Professor of Statistics, NCSU. Dr. Zhang's research focuses on the development of novel nonparametric variable selection methods such as support vector machines (SVMs). She is currently directing the doctoral dissertation research for 2 students.


Last updated: January, 2007

Back to the top

Skip footer links and go to content
Twitter iconTwitterExternal link Disclaimer         Facebook iconFacebookimage of external link icon         YouTube iconYouTubeimage of external link icon         Google+ iconGoogle+image of external link icon