Jeffrey D. Blume, PhD

Professor of Biostatistics

Professor of Biomedical Informatics

Director of Graduate Education, Data Science Institute

Vice-Chair for Education, Biostatistics

E-mail: j 'dot' blume 'at' vanderbilt 'dot' edu
Phone: (615) 343-9267
Fax: (615) 343-4924
Data Science Program Manager: Dana Daniele
Vanderbilt University School of Medicine
Department of Biostatistics
2525 West End, Ste. 11000
Nashville, TN 37203
I am an Professor in the Department of Biostatistics at Vanderbilt University School of Medicine. I am the Director of Graduate Education at the Data Science Institute where I have recently developed a new MS in Data Science. I served as the Director of Graduate Studies in Biostatistics at Vanderbilt for over 10 years. I received my Ph.D. in Biostatistics from the John Hopkins School of Public Health, where I was the recipient of a National Eye Institute Traineeship in Clinical Trials. I am the leading expert in likelihood methods for measuring statistical inference and I publish on the foundations of statistical inference. Recent work has led to the development of a “second-generation p-value” that has significantly improved statistical and scientific properties. Throughout my career I have published broadly on the foundations of statistical inference, methodology for analyzing and interpreting receiver operating characteristic curves, clinical trials design and analysis, and neuroimaging. My collaborative experience is quite diverse, with a continuing emphasis on diagnostic and cancer trials, large-scale trials in emergency medicine, translational biomedicine, and radiologic and fMRI studies.

I also have extensive experience in the development, operation, analysis, and methodological aspects of Clinical Trials. Before coming to Vanderbilt, I served as Deputy Director of the Biostatistics and Data Management Center for the American College of Radiology Imaging Network. ACRIN is a NCI funded cooperative group that specializes in clinical trials evaluating new medical imaging technology. I have been the lead statistician for numerous large-scale multicenter trials and regularly chair and serve on data safety monitoring boards. I regularly teach in the Radiological Society of North America’s exclusive Clinical Trial Methodology Workshop for junior faculty.

My current research program is focused on three areas: missing data and prediction, mediation analysis, and empirical Bayes methods for the evaluation of surrogate outcomes and for recovering information from published prediction models. Three PhD students work with me in the Statistical Evidence in Data Science (SEDS) lab. Recent work is as follows: we have discovered how to construct a prediction model that minimizes the average out-of-sample prediction error without imputing any missing data; we have developed a new framework for estimating mediation effects that posits just one regression model instead of three; we have extended the Prentice criteria for surrogate outcomes to partial surrogates; and we have discovered how to extract the statistical information behind a published black-box prediction model and use it to improve other prediction models.

Professor Blume's Curriculum Vitae

Webpage on Statistical Evidence

Biostatistics Graduate Program

Topic revision: r17 - 13 Nov 2020, JeffreyBlume
This site is powered by FoswikiCopyright &© 2013-2022 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Vanderbilt Biostatistics Wiki? Send feedback