Patient Treatment Trajectory Modeling With Markov Chains
Name
Markus Haug
Abstract
The vast amount of electronically stored medical data offers a possibility to investigate patients’ treatment trajectories. These trajectories give us a foundation to find out the best healthcare practices, evaluate the economics of treatment patterns and model the treatment paths. The aim of this work is to implement a technical solution for the aforementioned tasks. In the course of this thesis two R-packages will be developed. The first package uses electronical health data to construct treatment trajectories. The second package implements these trajectories and composes Markov models. The achieved tools are distributable to any database system operating on the open source Observational Health Data Sciences and Informatics Observational Medical Outcomes Partnership Common Data Model. Our universal solutions will be demonstrated on heart failure data supplied by Estonian Health Insurance Fund.
Graduation Thesis language
Estonian
Graduation Thesis type
Master - Data Science
Supervisor(s)
Raivo Kolde
Defence year
2022