Development of a prognostic scoring system for HIV-associated talaromycosis
- HIV/AIDs mentor: Thuy Le (email@example.com)
The objective is to develop and validate a prognostic scoring system to predict survival in HIV-associated talaromycosis utilizing data from multiple retrospective and prospective talaromycosis cohorts in Vietnam and China. This will involve collating data from multiple data sets, data cleaning (in particular of retrospective cohorts), and applying statistical modeling skills including using logistic regression for model derivation and bootstrapping for internal model validation. The eventual goal is to generate a simple clinical algorithm that can be built as an App on a smart phone to assist clinicians at bedside in the prognosis and management of talaromycosis patients.
Timeline and desired outcomes
Week 1-2: Data merging and cleaning. Pre-defining model variables and outcomes. Week 3-4: Model derivation Week 5-6: Model validation Week 7-10: Preparing manuscript for publication
Expectations for intern
Experience in data cleaning and analysis of large data sets is required or desirable. The intern will be working under the supervision of Dr. Thuy Le and a senior biostatistician at Duke and will acquire statistical modeling skills, such as logistic regression and bootstrapping method for model validation during the internship.