Date on Master's Thesis/Doctoral Dissertation

1-2021

Document Type

Master's Thesis

Degree Name

M.S.

Department

Bioinformatics and Biostatistics

Degree Program

Biostatistics, MS

Committee Chair

Zheng, Qi

Committee Co-Chair (if applicable)

Kulasekera, Karunarathna

Committee Member

Kulasekera, Karunarathna

Committee Member

Kong, Maiying

Committee Member

Bhatnagar, Aruni

Author's Keywords

Simulation; group testing; causal inference; inverse probability weighting; estimating equations

Abstract

The use of group testing to identify individuals with targeted outcomes in a population can greatly improve the efficiency, speed, and cost effectiveness of testing a population for an outcome, or at least for identifying the prevalence of an outcome in a population. The implementation of causal inference techniques can provide the basis for an observational study that would allow an investigator to gather estimates for treatment effectiveness if group testing was conducted on the population in a certain way. This thesis examines a simulation of the above outlined principles in order to demonstrate a potential application for determining treatment efficacy from observational data obtained via testing for disease outcome in a partially treated population. It is made evident that it is reasonable to make conclusions about treatment conditional incidence of an outcome in a sample of tested individuals based on outcome tests conducted on groups of individuals. Examining group study observations in the manner described in this thesis will allow researchers to estimate treatment effectiveness from partial data in situations where outcome testing may have been limited or where quick results are required from limited data.

Included in

Biostatistics Commons

Share

COinS