Faculty Profile

Brenda  Gillespie, PhD

Brenda Gillespie, PhD

  • Research Associate Professor of Biostatistics
  • Associate Director, CSCAR
  • 3550 Rackham Bldg.
  • 915 E. Washington
  • Ann Arbor, Michigan 48109-1070

Brenda Gillespie is Research Associate Professor of Biostatistics and Associate Director of UM's Consulting for Statistics, Computing and Analytics Research (CSCAR) . She provides statistical collaboration and support for numerous research projects at the University of Michigan. She teaches Biostatistics courses as well as CSCAR short courses in survival analysis, regression analysis, sample size calculation, generalized linear models, meta-analysis, and statistical ethics. Her major areas of expertise are clinical trials and survival analysis.

  • PhD, Statistics, Temple University, 1989
  • M.S., Statistics, Ohio State University, 1975
  • B.A., Mathematics, Earlham College, 1972

  • My research interests are in the area of censored data and clinical trials.

    One research interest concerns the application of categorical regression models to the case of censored survival data. This technique is useful in modeling the hazard function (instead of treating it as a nuisance parameter, as in Cox proportional hazards regression), or in the situation where time-related interactions (i.e., non-proportional hazards) are present. An investigation comparing various categorical modeling strategies is currently in progress.

    Another area of interest is the analysis of cross-over trials with censored data. I have developed (with M. Feingold) a set of nonparametric methods for testing and estimation in this setting. Our methods out-perform previous methods in most cases.

Applications of survival analysis in novel settings:  (a) the use of survival analysis to improve methodology for the analysis of age at first partner abuse in women by censoring at age of interview for those who had not experienced abuse. This method improved estimates of the cumulative probability of abuse at each age over previous crude methods.  (b,c) survival methods for left-censored data, i.e., values below a limit of detection, (d) use of left truncation of survival times when time-dependent covariate data were missing or unavailable prior to a known time, and (e) comparing the bias in four ways to ‘complete’ the Kaplan-Meier estimator.

Rigorous standards for study conduct and statistical methods for clinical trials:  (a) the Collaborative Initial Glaucoma Treatment Study (CIGTS), (b) a trial of a powerful new antiviral treatment for hepatitis C virus as part of the Adult-to-adult Living Donor Liver Transplantation Cohort Study (A2ALL), (c) a trial of alternative medicine treatments for heart failure, and (d) a trial of acupressure for classroom alertness carried out by a graduate course in clinical trials.

Methods of statistical analysis in research involving kidney disease (a-d) and liver transplantation (e).  (a) investigated the relationships between biomarkers, cardiovascular risk factors and patient outcomes, and (b) tested serum sodium levels as a predictor of patient outcomes using survival analysis with time-dependent covariates.  (c) investigated the relationships of heart rate variability and pulse wave velocity as predictors of outcomes including mortality. (d) presents a comparison of kidney transplant outcomes across 3 continents. (e) presents a unique comparison of data elements collected twice on the same patients, made possible by two studies that independently collected much of the same data from the same patients; one, an unfunded federal mandate to submit data, and the other, a study funded by the National Institutes of Health. 

Complete List of Published Work (over 300 papers) in MyBibliography