应用机器学习构建及验证新生儿中等长度导管并发症风险预测模型的研究

doi:10.3761/j.issn.2096-7446.2026.04.002

摘要/Abstract

摘要：

目的应用机器学习法构建新生儿中等长度导管（midline catheter，MC）并发症风险预测模型，验证并对比各模型性能，为新生儿MC并发症预测及早期防控提供参考。方法回顾性收集2020年8月至2023年12月在安徽省蚌埠市某三级甲等综合医院新生儿重症监护室接受MC置入的243例患儿的临床资料，包括新生儿一般资料、基础疾病、置管相关信息等。将研究对象按照7∶3的比例随机分为训练集170例和测试集73例。借助递归特征消除算法实现特征的初步筛选，采用逻辑回归（logistic regression，LR）、随机森林（random forest，RF）、支持向量机（support vector machine，SVM）、最小绝对收缩与选择算子（LASSO）和极端梯度提升（extreme gradient boosting，XGBoost）5种机器学习算法构建预测模型，通过交叉验证优化模型参数，比较不同模型的性能，选择最佳模型进行内部验证，并构建列线图。结果 243例MC置入患儿中有37例（15.23%）发生并发症。在训练集中，5种模型均具有较好的拟合度，其中RF模型受试者操作特征曲线下面积（area under curve，AUC）最大（0.843），特异度（0.916）、阳性预测值（0.613）、准确率（0.882）、F1值（0.655）优于其他模型；在测试集中，RF模型在AUC（0.881）、特异度（0.937）、阳性预测值（0.636）、准确率（0.904）、F1值（0.666）方面也均优于其他模型。RF模型中，前5位最重要的临床特征依次是出生体重、导管尖端位置、胎龄、送管是否顺利、穿刺部位。结论基于机器学习构建的新生儿MC并发症风险预测RF模型性能较优，列线图有助于临床护理人员及早识别MC并发症高风险患儿，进而采取针对性的预防措施。

关键词: 新生儿, 中等长度导管, 并发症, 预测模型, 机器学习

Abstract:

Objective To construct a risk prediction model for complications of midline catheters（MC） in neonates using machine learning methods，verify and compare the performance of various models，and provide a reference for the prediction and early prevention of MC complications in neonates. Methods The clinical data of 243 neonates who received MC insertion in the neonatal intensive care unit of a provincial tertiary general hospital in Bengbu City，Anhui Province from August 2020 to December 2023 and met the inclusion and exclusion criteria were retrospectively collected，including general neonatal information，underlying diseases，catheterization-related information，etc. The research subjects were randomly divided into a training set of 170 cases and a testing set of 73 cases in a 7：3 ratio. Recursive feature elimination algorithm was used to initially screen the features，and five machine learning algorithms，including logistic regression，random forest，support vector machine，least absolute shrinkage and selection operator，and extreme gradient boosting，were used to construct the prediction models. Model parameters were optimized through cross-validation，and the performance of different models was compared to select the best model for internal validation and to construct a nomogram. Results Among the 243 neonates with MC insertion，37（15.23%） developed complications. In the training set，all five models had good fitting degrees，among which the RF model had the largest AUC（0.843），specificity（0.916），PPV（0.613），accuracy（0.882），and F1 value（0.655），outperforming the other models. In the test set，the RF model also outperformed the other models in terms of AUC（0.881），specificity（0.937），PPV（0.636），accuracy（0.904），and F1 value（0.666）. In the RF model，the top five most important clinical features were birth weight，catheter tip position，gestational age，smoothness of catheter insertion，and puncture site. Conclusion The RF model for predicting the risk of MC complications in neonates based on machine learning has certain accuracy and practicality. The nomogram is helpful for clinical medical staff to identify high-risk neonates in advance and take targeted preventive measures.

Key words: Newborn, Medium Catheter, Complications, Prediction Model, Machine Learning

杨丽娟, 徐兵, 王一新, 陶言菊, 黄辉芳, 陈颖, 孙彩霞. 应用机器学习构建及验证新生儿中等长度导管并发症风险预测模型的研究[J]. 中华急危重症护理杂志, 2026, 7(4): 396-403.

YANG Lijuan, XU Bing, WANG Yixin, TAO Yanju, HUANG Huifang, CHEN Ying, SUN Caixia. Research on the construction and validation of a risk prediction model for complications of midline catheters in neonates using machine learning methods[J]. Chinese Journal of Emergency and Critical Care Nursing, 2026, 7(4): 396-403.

图/表 5

参考文献 24

[1]	Nickel B, Gorski L, Kleidon T, et al. Infusion therapy standards of practice,9th edition[J]. J Infus Nurs, 2024, 47(1S Suppl 1):S1-S285. doi: 10.1097/NAN.0000000000000532
[2]	徐兵, 杨丽娟, 戴明红, 等. 中等长度导管在危重新生儿静脉输液中的应用效果[J]. 循证护理, 2022, 8(9):1276-1278.
	Xu B, Yang LJ, Dai MH, et al. Effect of midline catheters in intravenous infusion of critically ill neonates[J]. Chin Evid Based Nurs, 2022, 8(9):1276-1278.
[3]	潘晋洲. 超早产儿PICC导管相关血流感染的影响因素分析[J]. 护理研究, 2024, 38(24):4460-4462.
	Pan JZ. Analysis of influencing factors of PICC catheter-related bloodstream infection in extremely preterm infants[J]. Chin Nurs Res, 2024, 38(24):4460-4462.
[4]	Blackwood BP, Farrow KN, Kim S, et al. Peripherally inserted central catheters complicated by vascular erosion in neonates[J]. JPEN J Parenter Enteral Nutr, 2016, 40(6):890-895. doi: 10.1177/0148607115574000 pmid: 25700180
[5]	王文歆, 赵艳, 徐梅, 等. 迷你中长导管与留置针用于老年患者输液的效果比较[J]. 护理学杂志, 2024, 39(21):59-61,84.
	Wang WX, Zhao Y, Xu M, et al. Comparison of the effectiveness of mini-midline catheter with peripheral venous catheter in older patients[J]. J Nurs Sci, 2024, 39(21):59-61,84.
[6]	Kleidon TM, Schults JA, Wainwright C, et al. Comparison of midline catheters and peripherally inserted central catheters to reduce the need for general anesthesia in children with respiratory disease:a feasibility randomized controlled trial[J]. Paediatr Anaesth, 2021, 31(9):985-995. doi: 10.1111/pan.v31.9
[7]	陶言菊, 杨丽娟, 戴明红, 等. 新生儿重症监护病房患儿中等长度导管相关并发症影响因素分析[J]. 蚌埠医学院学报, 2024, 49(10):1407-1411.
	Tao YJ, Yang LJ, Dai MH, et al. Analysis of influencing factors of complications related to medline catheter in NICU children[J]. J Bengbu Med Coll, 2024, 49(10):1407-1411.
[8]	Fläring U, Lundevall H, Norberg Å, et al. The success rate and complications of midline catheters in pediatric outpatient parenteral antibiotic therapy(OPAT)[J]. Eur J Pediatr, 2024, 183(4):1703-1709. doi: 10.1007/s00431-024-05432-7
[9]	Lai JY, Wu MJ, Gautama MN, et al. Comparison of complication rates between midline catheters and peripherally inserted central catheters:a systematic review and meta-analysis of randomized controlled trials[J]. 2024, 151:131-139.
[10]	Mayampurath A, Carey K, Palama B, et al. Machine learning-based pediatric early warning score:patient outcomes in a pre-versus post-implementation study,2019-2023[J]. Pediatr Crit Care Med, 2025, 26(2):e146-e154.
[11]	汪淑华. 基于机器学习的PICC相关性血栓风险预测模型构建[D]. 镇江: 江苏大学, 2022.
	Wang SH. Construction of PICC-related thrombosis risk pre-diction model based on machine learning[D]. Zhenjiang: Jiang- su University, 2022.
[12]	中华护理学会静脉输液治疗专业委员会, 孙红, 郭彩霞, 等. 静脉导管常见并发症临床护理实践指南[J]. 中华现代护理杂志, 2022, 28(18):2381-2395.
	Intravenous Infusion Therapy Committee of Chinese Nursing Association, Sun H, Guo CX, et al. Clinical nursing practice guidelines for common complications of intravenous catheters[J]. Chin J Mod Nurs, 2022, 28(18):2381-2395.
[13]	李祥云, 王洁, 刘畅, 等. 迷你中等长度导管在不同部位置入效果的比较研究[J]. 中华护理杂志, 2024, 59(20):2437-2443. doi: 10.3761/j.issn.0254-1769.2024.20.001
	Li XY, Wang J, Liu C, et al. Comparison of the effects of different insertion sites for mini-midline catheters[J]. Chin J Nurs, 2024, 59(20):2437-2443. doi: 10.3761/j.issn.0254-1769.2024.20.001
[14]	施卸丽, 祝柯鑫, 胡小黎, 等. 应用机器学习法构建及验证新生儿压力性损伤风险预测模型的研究[J]. 中华急危重症护理杂志, 2025, 6(5):536-542. doi: 10.3761/j.issn.2096-7446.2025.05.004
	Shi XL, Zhu KX, Hu XL, et al. Research on the construction and validation of a risk prediction model for neonatal pressure injury usingmachine learning methods[J]. Chin J Emerg Crit Care Nurs, 2025, 6(5):536-542.
[15]	Guan CJ, Gong AW, Zhao Y, et al. Interpretable machine learn-ing model for new-onset atrial fibrillation prediction in critically ill patients:a multi-center study[J]. Crit Care, 2024, 28(1):349. doi: 10.1186/s13054-024-05138-0
[16]	李艳敏, 平莉莉, 马晓蕾, 等. 不同胎龄不同体重新生儿医院感染的流行病学及临床特点[J]. 中国感染控制杂志, 2024, 23(9):1119-1125.
	Li YM, Ping LL, Ma XL, et al. Epidemiological and clinical characteristics of healthcare-associated infection in neonates with different gestational ages and birth weights[J]. Chin J Infect Contr, 2024, 23(9):1119-1125.
[17]	鲁华鹏, 马梅, 何晴, 等. 中等长度导管及经外周静脉穿刺中心静脉置管导管相关血流感染发生率的Meta分析[J]. 中国感染控制杂志, 2021, 20(4):309-319.
	Lu HP, Ma M, He Q, et al. Incidence of catheter-related bloodstream infection in patients with midline catheters and peripherally inserted central catheters:a Meta-analysis[J]. Chin J Infect Contr, 2021, 20(4):309-319.
[18]	赵林芳, 蔡志云, 樊小朋, 等. 中等长度静脉导管置管不同尖端位置的效果比较[J]. 中华护理杂志, 2022, 57(5):517-524. doi: 10.3761/j.issn.0254-1769.2022.05.001
	Zhao LF, Cai ZY, Fan XP, et al. Application comparison of different tip positions of midline catheters[J]. Chin J Nurs, 2022, 57(5):517-524. doi: 10.3761/j.issn.0254-1769.2022.05.001
[19]	Adams DZ, Little A, Vinsant C, et al. The midline catheter:a clinical review[J]. 2016, 51(3):252-258.
[20]	陈玉姬, 黄兰, 黄静雅, 等. 头静脉汇入腋静脉或锁骨下静脉夹角对经头静脉入路PICC置管术的影响[J]. 中国临床解剖学杂志, 2021, 39(6):638-641.
	Chen YJ, Huang L, Huang JY, et al. Effect of the angle between cephalic vein and axillary vein or subclavian vein on PICC catheterization through cephalic vein approach[J]. Chin J Clin Anat, 2021, 39(6):638-641. doi: 10.13418/j.issn.1001-165x.2021.06.003
[21]	Czarnik T, Czuczwar M, Borys M, et al. Ultrasound-guided infraclavicular axillary vein versus internal jugular vein cannulation in critically ill mechanically ventilated patients:a randomized trial[J]. Crit Care Med, 2023, 51(2):e37-e44. doi: 10.1097/CCM.0000000000005740
[22]	漆永林. 温度干预对早产儿PICC置管疼痛及置管效果的影响研究[D]. 大理: 大理大学, 2024.
	Qi YL. Study on the influence of temperature intervention on the pain and effect of PICC catheterization in premature infants[D]. Dali: Dali University, 2024.
[23]	杨丽娟, 徐兵, 戴明红, 等. 危险因素评估下针对性护理干预对新生儿PICC静脉炎预防效果分析[J]. 蚌埠医学院学报, 2023, 48(9):1278-1281.
	Yang LJ, Xu B, Dai MH, et al. Effect of targeted nursing intervention on PICC-associated phlebitis prevention in neonates under risk factor assessment[J]. J Bengbu Med Coll, 2023, 48(9):1278-1281.
[24]	张欢, 甘秀妮, 周雯, 等. ICU患者活动减少型谵妄风险动态列线图预测模型的构建与验证[J]. 中华急危重症护理杂志, 2024, 5(6):485-492.
	Zhang H, Gan XN, Zhou W, et al. Development and validation of a dynamic nomogram prediction model for hypoactive delirium risk in ICU patients[J]. Chin J Emerg Crit Care Nurs, 2024, 5(6):485-492.