Skip to page navigation menu Skip entire header
Brown University
Skip 13 subheader links

Research on the development of microsimulation models for medical decision making; a study on models for bladder cancer

Description

Abstract:
Bladder Cancer is a complex disease, with high incidence, the number of projected new bladder cancer cases in the United States for 2022 was 81,180. The Cancer Intervention and Surveillance Monitoring Program (CISNET) Bladder Cancer Incubator modeling group at Brown University is in the process of developing KYSTIS, an open-source microsimulation model for bladder cancer. This thesis is conducted in association with this modeling group and has two objectives; a PRISMA compliant systematic review of complex predictive models pertaining to bladder cancer and the study of variance reduction techniques (VRTs) in simulation studies. This work has found 100 manuscripts with relevant complex predictive models for bladder cancer, that are largely compartmental models (40) and microsimulations (33), with some decision tree (16) and cohort-based models (12). These models are heterogeneous in their goals and nomenclature, with different reporting practices and unfortunately only 11 of these 100 manuscripts are excellent in their reporting status. Through simulation studies the utilization of variance reduction techniques showed a significant reduction in the variance of the estimates from 300 replications of different simulation experiments. Showing a promising result for the implementation of these techniques in more complex simulations, such as the KYSTIS model.
Notes:
Thesis (Sc. M.)--Brown University, 2023

Citation

Hedspeth, Timothy Hogan, "Research on the development of microsimulation models for medical decision making; a study on models for bladder cancer" (2023). Biostatistics Theses and Dissertations. Brown Digital Repository. Brown University Library. https://repository.library.brown.edu/studio/item/bdr:k7c5s58t/

Relations

Collection: