Using machine learning to estimate survival curves for transplarnt patients receiving an increased risk for disease transmission donor organ versus waiting for a standard organ

Survival analysis;Increased risk for disease transmission donors;Organ transplant survival;Risk analysis;Statistics;Machine learning

Mark, Ethan Joshua ; Goldsman, David Keskinocak, Pinar Sokol, Joel Gurbaxani, Brian Patzer, Rachel Industrial and Systems Engineering ; Goldsman, David

University:Georgia Institute of Technology

Department:Industrial and Systems Engineering

关键词: Survival analysis;    Increased risk for disease transmission donors;    Organ transplant survival;    Risk analysis;    Statistics;    Machine learning;   

Others  :  https://smartech.gatech.edu/bitstream/1853/62673/1/MARK-DISSERTATION-2019.pdf

美国|英语

来源: SMARTech Repository

【 摘 要 】

In 1994, the Centers for Disease Control and Prevention (CDC) and the Public Health Service (PHS) released guidelines classifying donors at risk of transmitting human immunodeficiency virus (HIV) through organ transplantation. In 2013, the guidelines were updated to include donors at risk of transmitting hepatitis B (HBV) and hepatitis C (HCV). These donors are known as increased risk for disease transmission donors (IRD). Even though donors are now universally screened for HIV, HBV, and HCV by nucleic acid testing (NAT), NAT can be negative during the eclipse phase, when the virus is not detectable in blood. In part due to the opioid epidemic, over 19% of organ donors were classified as IRD in 2014. Despite the risks of disease transmission and associated mortality from accepting an IRD organ offer, patients also face mortality risks if they decline the organ and wait for a non-IRD organ. The main theme of this thesis is to build organ transplant and waitlist survival models and to help patients decide between accepting an IRD organ offer or remaining on the waitlist for a non-IRD organ. In chapter one, we introduced background information and the outline of the thesis. In chapter two, we used machine learning to build an organ transplant survival model for the kidney that achieves greater performance than the model currently being used in the U.S. kidney allocation system. In chapter three, we used similar modeling techniques and simulation to compare the survival for patients accepting IRD kidney offers vs. waiting for non-IRD kidneys. We then extend our IRD vs. non-IRD survival comparisons to the liver, heart and lung in chapter four, using different models and parameters. In chapter five, we built a model that predicts how the health of a patient changes from waitlist registration to transplantation. In chapter six, we utilized the transplant and waitlist survival models built in chapters three and four to create an interactive tool that displays the survival curves for a patient receiving an IRD organ or waiting for a non-IRD organ. The tool can also show the survival curve if a patient chooses to receive a non-IRD organ immediately. We then concluded with a discussion and major takeaways in chapter seven.