Core B, Data Management and Biostatistics Core will be responsible for study design, data collection, database management, as well as power and statistical analysis for the Projects. This Core will be headed up by Keith E. Muller, PhD, Professor in the Department of Health Outcomes Policy. He has over 30 years experience leading the biostatistics cores of clinical investigations. Of his 100+ publications, 50% are in sample size research and physiological modeling.
Aim 1. Data Management. The data management activities of this core will be led by Xuerong Wen, a certified Oracle Database Developer and a full time biostatistician at the University of Florida. Data management tasks will include designing the data collection form, data entry and quality control. Data intregity will be ensured through the use of range and logic checking of data at the time of entry. Throughout the entire process of data entiy, Xuerong will work weekly in the Analytical core to oversee data entry, screen validation, producing quality assurance checks, and suggesting any protocol changes necessary to improve data quality. She will work closely with the biostatisticians managing the core to ensure the study designs and objectives are met. We will utilize the University of Florida's Clinical and Translational Research Informatics Program because it has a powerful, reliable and secure computer environment for storing large sets of data.
Aim 2. Biostatistics. In addition to Dr. Muller, additional support will be provided by Wei Hou, PhD, a biostatistician in the Division. The biostatistical team will create operational definitions for outcome variables and other covariates, analyze pilot/exploratory studies to obtain data useful for planning future studies, perform power analyses and calculate appropriate sample sizes for comparing primary outcome measures, monitor and measure protocol deviations and assess the potential impact of any deviations on the studies, conduct statistical analysis for conclusions of research hypotheses, develop new methods when traditional methods fail to address the research questions, and support manuscript writing and statistical section writing for future grant proposals.
Showing the most recent 10 out of 14 publications
