Scholarlink

<p>Acute kidney injury is a frequent complication after pediatric heart transplant and is associated with long-term renal dysfunction and mortality. Factors associated with acute kidney injury have been studied; however, the role of intraoperative physiologic monitoring, particularly renal near infrared spectroscopy, remains incompletely understood. This study aimed to evaluate the association between intraoperative hemodynamic markers with the development of postoperative acute kidney injury in pediatric heart transplant recipients. This was a single-center retrospective study of pediatric patients undergoing orthotopic heart transplantation between 2021 and 2024. Percent change from baseline for various intraoperative hemodynamic indices were collected. Discriminative ability of individual predictors was assessed using receiver operating curve analysis. A multivariable random forest classifier incorporating patient characteristics, operative variables, and intraoperative physiologic changes was developed using fivefold cross-validation, and partial dependence plots were generated for all independent variables. Eighty-four patients were included; 65 (77%) developed postoperative acute kidney injury. Receiver operating curve analysis demonstrated that percent change in renal near infrared spectroscopy was significantly associated with acute kidney injury (area under the curve 0.67, p &lt; 0.01), with an optimal cutoff at a 29% decrease from baseline. Percent change in mean arterial pressure showed modest discrimination (area under the curve 0.61, p = 0.08), while central venous pressure, cardiopulmonary bypass time, and ischemic time were not independently discriminative. The random forest model demonstrated good performance (area under the curve 0.67; Brier score 0.33). Renal near infrared spectroscopy change was the most important predictor, followed by mean arterial pressure change, and weight. Partial dependence analysis identified inflection points at approximately 30% decrease in renal near infrared spectroscopy associated with increased probability of acute kidney injury. In pediatric heart transplant recipients, renal near infrared spectroscopy demonstrated the strongest discriminative ability for postoperative acute kidney injury and outperforms traditional hemodynamic and operative variables. Renal near infrared spectroscopy may serve as a high-fidelity, real-time physiologic marker to identify patients at elevated risk for renal injury and represents a potential target for intraoperative management strategies.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Acute Kidney Injury After Pediatric Heart Transplant: The Impact of Intraoperative Hemodynamics Using High-Fidelity Data

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月16日

查看全文
  • Matthew Cummins,
  • Brian Madden,
  • Philip T. Thrush,
  • Rajit Basu,
  • Rohit S. Loomba

摘要

Acute kidney injury is a frequent complication after pediatric heart transplant and is associated with long-term renal dysfunction and mortality. Factors associated with acute kidney injury have been studied; however, the role of intraoperative physiologic monitoring, particularly renal near infrared spectroscopy, remains incompletely understood. This study aimed to evaluate the association between intraoperative hemodynamic markers with the development of postoperative acute kidney injury in pediatric heart transplant recipients. This was a single-center retrospective study of pediatric patients undergoing orthotopic heart transplantation between 2021 and 2024. Percent change from baseline for various intraoperative hemodynamic indices were collected. Discriminative ability of individual predictors was assessed using receiver operating curve analysis. A multivariable random forest classifier incorporating patient characteristics, operative variables, and intraoperative physiologic changes was developed using fivefold cross-validation, and partial dependence plots were generated for all independent variables. Eighty-four patients were included; 65 (77%) developed postoperative acute kidney injury. Receiver operating curve analysis demonstrated that percent change in renal near infrared spectroscopy was significantly associated with acute kidney injury (area under the curve 0.67, p < 0.01), with an optimal cutoff at a 29% decrease from baseline. Percent change in mean arterial pressure showed modest discrimination (area under the curve 0.61, p = 0.08), while central venous pressure, cardiopulmonary bypass time, and ischemic time were not independently discriminative. The random forest model demonstrated good performance (area under the curve 0.67; Brier score 0.33). Renal near infrared spectroscopy change was the most important predictor, followed by mean arterial pressure change, and weight. Partial dependence analysis identified inflection points at approximately 30% decrease in renal near infrared spectroscopy associated with increased probability of acute kidney injury. In pediatric heart transplant recipients, renal near infrared spectroscopy demonstrated the strongest discriminative ability for postoperative acute kidney injury and outperforms traditional hemodynamic and operative variables. Renal near infrared spectroscopy may serve as a high-fidelity, real-time physiologic marker to identify patients at elevated risk for renal injury and represents a potential target for intraoperative management strategies.