Inteligência artificial para avaliação dos fatores que influenciam o óbito em pacientes com insuficiência renal

Abstract

Chronic renal failure represents an impact at the individual and collective level that translates into the suffering that the disease brings to patients and the gradual increase in expenses both with dialysis methods and with the diseases associated with this public. Some technologies can be extremely useful in decision-making in healthcare. One of these technologies is the Bayesian network, which has, as an example, the ability to help, even during screening, in the best choice of vascular access for patients with severe renal failure, among other situations. The objective of this study was to analyze the possible causal relationships, using probabilistic inferences, between the related factors to stipulate the main causes of death in patients with severe renal failure. The methodology was based on the analysis of medical records of 121 patients using artificial intelligence, Bayesian networks, in order to establish a conditional probability relationship between the variables of patients with severe renal failure. Through the study, it was possible to identify that the choice of venous access type arteriovenous fistula (AVF), whenever possible, should be prioritized, as it proves to be a fundamental strategy for maintaining the number of deaths in hemodialysis centers, as well as maintaining extra care when to smoking among patients, as it was a major risk factor, both individually and in addition to other variables, in the group of patients in question. The analysis, using Bayesian networks, is of great value for the group belonging to the study, given the opportunity to identify the evolution of renal failure and consequently promote a reduction in the number of deaths. And it is also of great benefit to health professionals, mainly assisting in Clinical Decision Support (CDS).