Scholarlink

Objectives <p>This study seeks to construct and internally validate a clinical prediction model for predicting new-onset dialysis-requiring acute kidney injury (AKI) following heart transplantation (HT).</p> Methods <p>The Kaplan-Meier survival analysis and log-rank test were utilized for conducting the survival analysis. A clinical prediction model was developed to predict postoperative dialysis-requiring AKI, based on a logistic regression model and likelihood ratio test with Akaike Information Criterion. The performance of the prediction model was assessed using C-index, receiver operating characteristic curves, calibration curves, Brier score, and the Spiegelhalter Z-test. Clinical utility was evaluated using decision curve analysis and clinical impact curves.</p> Results <p>This study included a total of 525 patients who underwent orthotopic HT in the single center located in Wuhan, China between January 2015 and December 2021, with 16.57% developing postoperative dialysis-requiring AKI. Patients who experienced postoperative dialysis-requiring AKI exhibited a lower overall survival rate. All enrolled participants were randomly allocated into derivation (<i>n</i> = 350) and validation (<i>n</i> = 175) cohorts at a ratio of 2:1. The final prediction model comprised six indicators: diabetes, stroke, gout, prognostic nutritional index, estimated glomerular filtration rate, and cardiopulmonary bypass duration. The prediction model demonstrated outstanding discrimination (C-index of 0.792 in the derivation cohort and 0.834 in the validation cohort) as well as calibration performance, indicating strong concordance between observed and nomogram-predicted probabilities. Subgroup analysis based on age, preoperative serum creatine levels, and year of surgery also exhibited robust discrimination and calibration capabilities.</p> Conclusions <p>Dialysis-requiring AKI following HT is associated with poor clinical prognosis. The prediction model, comprising six indicators, is capable of predicting dialysis-requiring AKI following HT. This prediction model holds promise in assisting both patients and clinicians in forecasting postoperative renal failure, thereby improving clinical management.</p> Clinical trial number <p>Not applicable.</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

Development and validation of a clinical prediction model for dialysis-requiring acute kidney injury following heart transplantation: a single-center study from China

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年3月3日

查看全文
  • Shirui Qian,
  • Bingxin Cao,
  • Ping Li,
  • Nianguo Dong

摘要

Objectives

This study seeks to construct and internally validate a clinical prediction model for predicting new-onset dialysis-requiring acute kidney injury (AKI) following heart transplantation (HT).

Methods

The Kaplan-Meier survival analysis and log-rank test were utilized for conducting the survival analysis. A clinical prediction model was developed to predict postoperative dialysis-requiring AKI, based on a logistic regression model and likelihood ratio test with Akaike Information Criterion. The performance of the prediction model was assessed using C-index, receiver operating characteristic curves, calibration curves, Brier score, and the Spiegelhalter Z-test. Clinical utility was evaluated using decision curve analysis and clinical impact curves.

Results

This study included a total of 525 patients who underwent orthotopic HT in the single center located in Wuhan, China between January 2015 and December 2021, with 16.57% developing postoperative dialysis-requiring AKI. Patients who experienced postoperative dialysis-requiring AKI exhibited a lower overall survival rate. All enrolled participants were randomly allocated into derivation (n = 350) and validation (n = 175) cohorts at a ratio of 2:1. The final prediction model comprised six indicators: diabetes, stroke, gout, prognostic nutritional index, estimated glomerular filtration rate, and cardiopulmonary bypass duration. The prediction model demonstrated outstanding discrimination (C-index of 0.792 in the derivation cohort and 0.834 in the validation cohort) as well as calibration performance, indicating strong concordance between observed and nomogram-predicted probabilities. Subgroup analysis based on age, preoperative serum creatine levels, and year of surgery also exhibited robust discrimination and calibration capabilities.

Conclusions

Dialysis-requiring AKI following HT is associated with poor clinical prognosis. The prediction model, comprising six indicators, is capable of predicting dialysis-requiring AKI following HT. This prediction model holds promise in assisting both patients and clinicians in forecasting postoperative renal failure, thereby improving clinical management.

Clinical trial number

Not applicable.