Survival analysis on prediction of waiting time for kidney transplantation / Nur Syafieka Salleh and Balkiah Moktar
Kidney transplantation (KT) describes implanting a healthy kidney from a deceased or living donor. In 2020, the statistics stated that the waiting list remains significantly longer than the supply of kidneys, posing a continuing challenge. Therefore, this research aims to estimate the likelihood of...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Book Section |
| Language: | English |
| Published: |
College of Computing, Informatics and Media, UiTM Perlis
2023
|
| Subjects: | |
| Online Access: | https://ir.uitm.edu.my/id/eprint/100821/1/100821.pdf https://ir.uitm.edu.my/id/eprint/100821/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Kidney transplantation (KT) describes implanting a healthy kidney from a deceased or living donor. In 2020, the statistics stated that the waiting list remains significantly longer than the supply of kidneys, posing a continuing challenge. Therefore, this research aims to estimate the likelihood of getting a KT, determine the key factors influencing KT waiting time and choose the best model within two distributions. This study includes 46,750 secondary data collected from the Kaggle website between January 1, 2000, and December 31, 2017, involving the deceased donor transplants. Potential predictors included in this study are gender, gestation, prior transplant, age, dialysis duration, underlying disease, blood group and calculated panel reactive antibody (c-PRA) level. Firstly, the Kaplan-Meier (KM) technique predicts the survival curves for each determinant group before being compared using the Log rank test. The KM plot for the age factor shows the greatest significant divergence compared to other factors indicating the remarkable impact on the KT waiting time. The Log-rank test shows that all variables are important and impact the KT waiting time. Next, the Cox proportional hazards (PH) regression model helps examine the influence of determinants on patients on the KT waiting list. Findings revealed that all determinants above significantly affect the waiting time. Lastly, the Exponential and Weibull models determine the best model for fitting the KT dataset. Both models generated similar significant variables, but the Weibull models apt the dataset better due to a lower Akaike Information Criterion (AIC) index. |
|---|