Graduate STEM Fellow Profile
Thesis: Comparison of Methods to Test Intermediate Endpoint Effects in Logistic Regression
College/University: University of Colorado Denver
Research Advisor: Lorri Ogden
Degree Sought: Ph.D. Biostatistics
Department: Department of Biostatistics and Informatics
Research Focus: Statistical methods used to design and evaluate variables called intermediate endpoints, which are used in clinical trials.
Description of Research
An intermediate endpoint (IE) is defined as a variable that is on the causal pathway showing how an independent treatment/exposure variable affects an outcome, and the use of these variables in clinical and evaluation research is becoming increasingly popular. One such example would be in a study to determine whether a new cancer treatment is effective in decreasing the odds of death. If an IE variable is identified and is both affected by the treatment and related to the odds of death, that IE might be used to help determine if the new treatment was effective. This can result in reduced time and cost for a clinical trial, allowing the release of an effective treatment sooner, or discontinuing the use of an ineffective one. Despite the growing desire to use IE variables, there statistical methods used to analyze results are not fully developed. The purpose of my research is to thoroughly evaluate common statistical tests for IEs under a wide range of possible trial conditions, and determine the best method to be applied to future studies. In addition to data analysis, it is important to understand the methods which will be used to analyze data prior to designing a study and this research will also help determine better ways to prospectively design clinical trials and other evaluation studies.
Example of how my research is integrated into my GK-12 experience
In our GK-12 project, each middle school has one math and one science fellow. One example of bringing my research into the classroom will be through an integrated math-science unit. In these lessons, we will introduce the concept of a clinical trial to the students by performing a mock study. During this unit, we will develop an understanding of a clinical trial or other medical study through the following steps: 1) walk through the process of designing a study by using simple statistics of expected mean and variability, 2) have the students be our “subjects” assigned mock treatments and outcomes to gather data, and 3) analyze the data and draw conclusions. At the end of the unit, the students will have gained a better understanding of the process of a medical study, as well as have a better understanding and appreciation for the need for statistics in designing an experiment.