Tagged: Department of Clinical Pharmacy

In my role as Biostatistician, I support the Medication Outcomes Center staff, faculty, and partners in data cleaning, data visualization, study design, and statistical data modeling. This includes working on large electronic health record (EHR) data sets to derive clinically meaningful variables, merging multiple data sets by key identifiers, converting repeated measures data to independent patient-level measures, tabulating descriptive statistics including parametric or non-parametric statistical tests of independence, and creating exploratory plots for patient demographics and clinical outcomes. Some of the observational data modeling approaches I frequently use include logistic regression, mixed models, marginal models, and proportional hazards models. In addition to classic statistical methods, I am interested in applying machine learning algorithms to better understand EHR data with complex variables, heteroscedasticity, and non-linearity.