ABSTRACT
ABSTRACT Introduction: Anemia is frequent in patients undergoing replacement therapy for kidney failure. Anemia in the pre- and post-transplantation period might be related to kidney transplant outcomes. The current study therefore sought to assess the relationship between anemia, delayed allograft function (DGF), chronic kidney allograft dysfunction (CAD), and death from any cause following kidney transplantation from a deceased donor. Methods: This was a retrospective study with 206 kidney transplant patients of deceased donors. We analyzed deceased donors' and kidney transplant patients' demographic data. Moreover, we compared biochemical parameters, anemia status, and medicines between DGF and non-DGF groups. Afterward, we performed a multivariate analysis. We also evaluated outcomes, such as CAD within one year and death in ten years. Results: We observed a lower frequency of pre-transplant hemoglobin concentration (Hb) but higher frequency of donor-serum creatinine and red blood transfusion within one week after transplantation in the group with DGF. In addition, there was an independent association between Hb concentration before transplantation and DGF [OR 0.252, 95%CI: 0.159-0.401; p < 0.001]. There was also an association between Hb concentration after six months of kidney transplantation and both CAD [OR 0.798, 95% CI: 0.687-0.926; p = 0.003] and death from any cause. Conclusion: An association was found between pre-transplantation anemia and DGF and between anemia six months after transplantation and both CAD and death by any cause. Thus, anemia before or after transplantation affects the outcomes for patients who have undergone kidney transplantation from a deceased donor.
RESUMO Introdução: A anemia é frequente em pacientes submetidos à terapia substitutiva para insuficiência renal. A anemia nos períodos pré e pós-transplante pode estar relacionada aos desfechos do transplante renal. Portanto, o presente estudo buscou avaliar a relação entre anemia, função retardada do enxerto (FRE), disfunção crônica do enxerto renal (DCE) e óbito por qualquer causa após transplante renal de doador falecido. Métodos: Este foi um estudo retrospectivo com 206 pacientes transplantados renais de doadores falecidos. Analisamos dados demográficos de doadores falecidos e pacientes transplantados renais. Além disso, comparamos parâmetros bioquímicos, status de anemia e medicamentos entre os grupos FRE e não-FRE. Posteriormente, realizamos uma análise multivariada. Também avaliamos desfechos, como DCE em um ano e óbito em dez anos. Resultados: Observamos menor frequência de concentração de hemoglobina (Hb) pré-transplante, mas maior frequência de creatinina sérica do doador e transfusão de hemácias no período de uma semana após o transplante no grupo FRE. Além disso, houve associação independente entre a concentração de Hb antes do transplante e a FRE [OR 0,252; IC 95%: 0,159-0,401; p < 0,001]. Houve também associação entre a concentração de Hb após seis meses de transplante renal e ambos, DCE [OR 0,798; IC95%: 0,687-0,926; p = 0,003] e óbito por qualquer causa. Conclusão: Encontrou-se uma associação entre anemia pré-transplante e FRE e entre anemia seis meses após o transplante e ambos, DCE e óbito por qualquer causa. Assim, a anemia antes ou após o transplante afeta os desfechos de pacientes que foram submetidos a transplante renal de doador falecido.
ABSTRACT
Supervised machine learning (ML) is a class of algorithms that "learn" from existing input-output pairs, which is gaining popularity in pattern recognition for classification and prediction problems. In this scoping review, we examined the use of supervised ML algorithms for the prediction of long-term allograft survival in kidney transplant recipients. Data sources included PubMed, the Cumulative Index to Nursing and Allied Health Literature, and the Institute for Electrical and Electronics Engineers (IEEE) Xplore libraries from inception to November 2019. We screened titles and abstracts and potentially eligible full-text reports to select studies and subsequently abstracted the data. Eleven studies were identified. Decision trees were the most commonly used method (n = 8), followed by artificial neural networks (ANN) (n = 4) and Bayesian belief networks (n = 2). The area under receiver operating curve (AUC) was the most common measure of discrimination (n = 7), followed by sensitivity (n = 5) and specificity (n = 4). Model calibration examining the reliability in risk prediction was performed using either the Pearson r or the Hosmer-Lemeshow test in four studies. One study showed that logistic regression had comparable performance to ANN, while another study demonstrated that ANN performed better in terms of sensitivity, specificity, and accuracy, as compared with a Cox proportional hazards model. We synthesized the evidence related to the comparison of ML techniques with traditional statistical approaches for prediction of long-term allograft survival in patients with a kidney transplant. The methodological and reporting quality of included studies was poor. Our study also demonstrated mixed results in terms of the predictive potential of the models.
