Method for predicting the probability of kidney transplantation for patients on the waiting list in Colombia

Abstract

Introduction: The criteria for organ distribution in Colombia establish an initial local distribution, then regional and finally national. In December 2019, 2,822 people were on the kidney transplant waiting list in Colombia, assigned mostly to the Bogotá and IPS regions with the largest waiting lists. This high concentration of patients could be generating unwanted effects on the opportunity that patients have to receive a kidney transplant. Objectives: In this paper we seek to study, based on synthetic data generated with the information available from the INS, the probability of organ allocation, identifying the most informative variables and proposing a method to calculate the probability of allocation for a given patient on the waiting list of organs. Material and methods: The adjustment of a model based on decision trees is presented, which showed a high precision and allows the prediction of the probability of obtaining an organ. Results: Time, transplant IPS and blood group were identified as the most informative variables. Likewise, there are differences in the time it takes to obtain kidney transplants between regions and between transplanting IPS due to the effect of the size of their waiting list. Conclusions: The proposed method allows us to identify the importance of the variables that define obtaining an organ. Finally, for a given patient, it is possible to estimate the probability of being classified in one of the outcome categories.