Acute Kidney Injury Prediction Model Using Cystatin-C, Beta-2 Microglobulin, and Neutrophil Gelatinase-Associated Lipocalin Biomarker in Sepsis Patients.

IF 2.1 Q2 UROLOGY & NEPHROLOGY

International Journal of Nephrology and Renovascular Disease Pub Date : 2024-03-28 eCollection Date: 2024-01-01 DOI:10.2147/IJNRD.S450901

Hani Susianti, Aswoco Andyk Asmoro, Sujarwoto, Wiwi Jaya, Heri Sutanto, Amanda Yuanita Kusdijanto, Kevin Putro Kuwoyo, Kristian Hananto, Matthew Brian Khrisna

{"title":"Acute Kidney Injury Prediction Model Using Cystatin-C, Beta-2 Microglobulin, and Neutrophil Gelatinase-Associated Lipocalin Biomarker in Sepsis Patients.","authors":"Hani Susianti, Aswoco Andyk Asmoro, Sujarwoto, Wiwi Jaya, Heri Sutanto, Amanda Yuanita Kusdijanto, Kevin Putro Kuwoyo, Kristian Hananto, Matthew Brian Khrisna","doi":"10.2147/IJNRD.S450901","DOIUrl":null,"url":null,"abstract":"Introduction: AKI is a frequent complication in sepsis patients and is estimated to occur in almost half of patients with severe sepsis. However, there is currently no effective therapy for AKI in sepsis. Therefore, the therapeutic approach is focused on prevention. Based on this, there is an opportunity to examine a panel of biomarker models for predicting AKI.Material and methods: This prospective cohort study analysed the differences in Cystatin C, Beta-2 Microglobulin, and NGAL levels in sepsis patients with AKI and sepsis patients without AKI. The biomarker modelling of AKI prediction was done using machine learning, namely Orange Data Mining. In this study, 130 samples were analysed by machine learning. The parameters used to obtain the biomarker panel were 23 laboratory examination parameters.Results: This study used SVM and the Naïve Bayes model of machine learning. The SVM model's sensitivity, specificity, NPV, and PPV were 50%, 94.4%, 71.4%, and 87.5%, respectively. For the Naïve Bayes model, the sensitivity, specificity, NPV, and PPV were 83.3%, 77.8%, 87.5%, and 71.4%, respectively.Discussion: This study's SVM machine learning model has higher AUC and specificity but lower sensitivity. The Naïve Bayes model had better sensitivity; it can be used to predict AKI in sepsis patients.Conclusion: The Naïve Bayes machine learning model in this study is useful for predicting AKI in sepsis patients.","PeriodicalId":14181,"journal":{"name":"International Journal of Nephrology and Renovascular Disease","volume":"17 ","pages":"105-112"},"PeriodicalIF":2.1000,"publicationDate":"2024-03-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10984190/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"International Journal of Nephrology and Renovascular Disease","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.2147/IJNRD.S450901","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2024/1/1 0:00:00","PubModel":"eCollection","JCR":"Q2","JCRName":"UROLOGY & NEPHROLOGY","Score":null,"Total":0}

引用次数: 0

Abstract

Introduction: AKI is a frequent complication in sepsis patients and is estimated to occur in almost half of patients with severe sepsis. However, there is currently no effective therapy for AKI in sepsis. Therefore, the therapeutic approach is focused on prevention. Based on this, there is an opportunity to examine a panel of biomarker models for predicting AKI.

Material and methods: This prospective cohort study analysed the differences in Cystatin C, Beta-2 Microglobulin, and NGAL levels in sepsis patients with AKI and sepsis patients without AKI. The biomarker modelling of AKI prediction was done using machine learning, namely Orange Data Mining. In this study, 130 samples were analysed by machine learning. The parameters used to obtain the biomarker panel were 23 laboratory examination parameters.

Results: This study used SVM and the Naïve Bayes model of machine learning. The SVM model's sensitivity, specificity, NPV, and PPV were 50%, 94.4%, 71.4%, and 87.5%, respectively. For the Naïve Bayes model, the sensitivity, specificity, NPV, and PPV were 83.3%, 77.8%, 87.5%, and 71.4%, respectively.

Discussion: This study's SVM machine learning model has higher AUC and specificity but lower sensitivity. The Naïve Bayes model had better sensitivity; it can be used to predict AKI in sepsis patients.

Conclusion: The Naïve Bayes machine learning model in this study is useful for predicting AKI in sepsis patients.

查看原文本刊更多论文

利用胱抑素-C、β-2 微球蛋白和中性粒细胞明胶酶相关脂质体生物标记物的脓毒症患者急性肾损伤预测模型

导言：AKI 是败血症患者经常出现的一种并发症，据估计，几乎一半的严重败血症患者都会出现 AKI。然而，目前还没有治疗脓毒症急性肾损伤的有效方法。因此，治疗方法的重点在于预防。在此基础上，我们有机会研究一组用于预测 AKI 的生物标志物模型：这项前瞻性队列研究分析了有 AKI 的脓毒症患者和无 AKI 的脓毒症患者胱抑素 C、β-2 微球蛋白和 NGAL 水平的差异。预测 AKI 的生物标志物建模是通过机器学习（即 Orange 数据挖掘）完成的。在这项研究中，机器学习分析了 130 个样本。用于获得生物标志物面板的参数是 23 项实验室检查参数：本研究使用了 SVM 和奈夫贝叶斯机器学习模型。SVM 模型的灵敏度、特异性、NPV 和 PPV 分别为 50%、94.4%、71.4% 和 87.5%。Naïve Bayes 模型的灵敏度、特异性、NPV 和 PPV 分别为 83.3%、77.8%、87.5% 和 71.4%：本研究的 SVM 机器学习模型具有较高的 AUC 和特异性，但灵敏度较低。讨论：本研究的 SVM 机器学习模型具有较高的 AUC 和特异性，但灵敏度较低，而 Naïve Bayes 模型的灵敏度较高，可用于预测脓毒症患者的 AKI：结论：本研究中的奈伊夫贝叶斯机器学习模型可用于预测脓毒症患者的 AKI。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

International Journal of Nephrology and Renovascular Disease UROLOGY & NEPHROLOGY-

CiteScore

3.90

自引率

5.00%

发文量

审稿时长

16 weeks

期刊介绍： International Journal of Nephrology and Renovascular Disease is an international, peer-reviewed, open-access journal focusing on the pathophysiology of the kidney and vascular supply. Epidemiology, screening, diagnosis, and treatment interventions are covered as well as basic science, biochemical and immunological studies. In particular, emphasis will be given to: -Chronic kidney disease- Complications of renovascular disease- Imaging techniques- Renal hypertension- Renal cancer- Treatment including pharmacological and transplantation- Dialysis and treatment of complications of dialysis and renal disease- Quality of Life- Patient satisfaction and preference- Health economic evaluations. The journal welcomes submitted papers covering original research, basic science, clinical studies, reviews & evaluations, guidelines, expert opinion and commentary, case reports and extended reports. The main focus of the journal will be to publish research and clinical results in humans but preclinical, animal and in vitro studies will be published where they shed light on disease processes and potential new therapies and interventions.