Department of Biostatistics

The School of Medicine at Vanderbilt University created the Department of Biostatistics in September 2003. In doing so, Vanderbilt made a major new funding commitment to build a world-class department that will include a doctoral level Ph.D. program. Chaired by Frank E. Harrell, Jr., the new department has exceptional institutional support from a medical center that is ranked 7th in the Faculty Scholarly Productivity Index (January 2007, based on purely objective measures) and is ranked in the ISI top five for research impact in clinical medicine and pharmacology. The School of Medicine received $278M in total NIH funding in 2007, representing an increase from 2006 of 1,366 to 1,463 awarded grants and national ranking of 18th out of 123 medical schools. From 2000-2005, the school has the fastest growth in NIH funding in the US, with a compound annual growth of 17.8%. The Vanderbilt School of Medicine has strong basic, clinical, health services research, and informatics programs. In addition to providing a highly favorable academic environment in general, the School of Medicine provides biostatisticians with the opportunity to make important contributions to medical research.

The Dean and other senior medical school faculty are committed to providing outstanding collaborative support in biostatistics to clinical and basic scientists and to developing a graduate program in biostatistics that will train outstanding collaborative scientists and will focus on the methods of modern applied statistics. The biostatistics faculty sincerely thank the following leaders of the medical center who had the vision to create the new department: Dean Steven Gabbe, Vice-Chancellor Harry Jacobson, Health Services Research Director Robert Dittus, and former Cancer Center Director Harold Moses.

Click here for an article about the new department that was published in the Vanderbilt Medical Center's weekly news magazine The Reporter. The Reporter also ran a story about the fifth anniversary of the department.

Department Vision and Goals | Organizational Chart

deptposter.jpg

Biostatistics is a basic science of biomedical research. Our vision for biostatisticians in the Department is that they become statistical scientists who are on the forefront of biomedical research and who contribute to the body of knowledge in medicine. In the role of statistical scientist, collaboration with medical researchers is of major importance, as is curiosity about diseases, treatments, patient outcomes, and biology. Our faculty are also independent researchers in the methods of biostatistics. For faculty to be effective in academic medicine and to have fruitful academic careers, it is important that they be well rounded. Though biostatisticians at Vanderbilt occasionally provide short-term consultation with physician scientists, we emphasize thematic long-term collaborations that advance medical science.

We believe in multidisciplinary collaboration but also in having a strong methodologic identity. In the words of Phyllis M Wise PhD, Provost, U. Washington, "Interdisciplinary only works when there are disciplines to inter with." (AAMC Oct 05). Methodologists need depth in their discipline.

Specific Goals

  • Department members should play key roles in making important medical discoveries
  • Department members should be innovators, scientists, collaborators, and teachers
  • Methodologic research done by department faculty should be state of the art while solving real world problems
  • Biostatisticians in the department should develop or choose new flexible experimental designs and routinely use them in biomedical research, aiming to enhance the ethical conduct of research, to do research in a more cost-effective way, and to result in evidence that is stated directly in terms of biological or medical relevance
  • Modern state of the art biostatistical methodology should be put to everyday use when its benefits have been demonstrated
  • Biostatisticians at Vanderbilt should be able to apply modern statistical methods to everyday problems as quickly as statisticians at other medical centers apply older methodologies
  • Biostatisticians should be fully integrated into the research program of other departments in the School of Medicine. Among other things, this integration involves participating in journal clubs, research conferences, and teaching.
  • Faculty and staff statisticians should be able to respond to a wide variety of funding opportunities
  • A first-class graduate program will be developed that is based on modern methods of applied statistics and that emphasizes new flexible and robust analytic methods. The program will train students in a variety of methods for dealing with messy data due to noncompliance, missing data, treatment selection bias in observational studies, and strange distributions. The program will also emphasize highly flexible, sequential, adaptive experimental designs.
  • We live by John Tukey's saying "We need to find approximate solutions for the right problem, which may be vaguely stated, rather than exact solutions for the wrong problem, which can always be made precise." Development of statistical methods should be driven by problem solving rather than mathematical elegance.

Approaches

  • Don't limit ourselves by the statistical computing tools we choose. The most modern analytical methods are implemented only in the most modern software.
  • Practice reproducible research
  • Emphasize the use of open source software, and give back to the computing and scientific open communities in recognition for what those communities have given to us.

What We Offer

Motivated by improving public health and enhancing understanding of biology, we
  • Optimize biological and medical measurements
  • Design experiments, in a highly interactive fashion with investigators
  • Enhance data acquisition
  • Account for as many sources of variation as possible
  • Normalize and adjust for background information, building these adjustments into the primary analysis when possible
  • Rigorously analyze associations
  • Rigorously validate predictions and models
  • Interpret results and help make them influence clinical practice and biologic research
  • Assist in making experimentation on humans and animals as ethical as possible
  • Assist in the formation and operation of collaborative groups to help sustain flexibility and discipline of thought and to overcome confirmation bias

According to Inouye SK, Fiellin DA An evidence-based guide to writing grant proposals for clinical research, 42% of NIH grants have inadequate sample size justifications and 66% of NIH grants have inadequate statistical analysis plans. Grant proposals with insufficiently thought-out study and experiment design do not fare well during peer review. Biostatistician involvement has a significant benefit to grant development.

-- FrankHarrell - 01 Feb, 02 Mar, 14 Aug 2004, 12 Jan, 19 Apr, 23 Oct 2005, 27 Feb 2006, 13 Jan 2007, 24 Jun 2008, 4 Feb 2012, 17 Feb 2015, with input from DeanBillheimer 02 Mar 2004

Topic revision: r28 - 17 Feb 2015, FrankHarrell
 

This site is powered by FoswikiCopyright © 2013 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