基于Lasso-logistic回归建立脓毒症相关肝损伤预测模型及防控策略

高婷; 付红; 邵敏; 宫娟

doi:10.16766/j.cnki.issn.1674-4152.004358

基于Lasso-logistic回归建立脓毒症相关肝损伤预测模型及防控策略

doi: 10.16766/j.cnki.issn.1674-4152.004358

高婷¹,
付红¹,
邵敏¹,
宫娟^2, ,

1.
安徽医科大学第一附属医院重症医学科，安徽合肥 230022
2.
安徽医科大学第一附属医院内分泌科

基金项目:

国家自然科学基金项目 82370605

详细信息

通讯作者:
宫娟，E-mail：yfy2651094@fy.ahmu.edu.cn

中图分类号: R631 R575
计量
- 文章访问数: 7
- HTML全文浏览量: 6
- PDF下载量: 3
- 被引次数: 0
出版历程
- 收稿日期: 2025-06-23

Establishment of a prediction model for sepsis-related liver injury and development of prevention and control strategies based on Lasso-logistic regression

GAO Ting¹,
FU Hong¹,
SHAO Min¹,
GONG Juan^{2
, ,}

1.
Department of Critical Care Medicine, the First Affiliated Hospital of Anhui Medical University, Hefei, Anhui 230022, China

摘要

摘要: 目的鉴于脓毒症相关肝损伤(SALI)危害严重，本研究旨在运用Lasso-logistic回归构建SALI预测模型，并制定针对性的防控策略。方法选取2024年4月1日—2025年4月1日安徽医科大学第一附属医院收治的192例脓毒症患者为研究对象，根据是否发生肝损伤分为SALI组(40例)和NSALI组(152例)。收集2组患者临床相关资料，采取单因素分析研究不同因素对脓毒症患者SALI发生的影响。采用Lasso-logistic回归分析研究SALI发生的危险因素，建立SALI风险预测模型并使用ROC曲线进行验证分析。结果 2组合并糖尿病、机械通气、合并休克、低氧血症、尿素、血小板计数、血乳酸(LAC)、序贯器官衰竭评分(SOFA评分)比较差异均有统计学意义(P < 0.05)。Lasso-logistic回归分析显示，合并糖尿病、合并休克、LAC、SOFA评分是发生SALI的独立危险因素(P < 0.05)。ROC曲线分析显示，合并糖尿病、合并休克、LAC、SOFA评分及列线图模型均可预测SALI，曲线下面积分别为0.615、0.620、0.843、0.837、0.948，其中列线图风险预测模型的AUC最高。结论合并糖尿病、合并休克、LAC、SOFA评分是SALI发生独立危险因素，基于此构建的列线图模型预测与实用价值高，利于为SALI患者制定精准防控策略。
- 脓毒症相关肝损伤 /
- Lasso回归 /
- Logistic回归 /
- 预测模型 /
- 策略
Abstract: Objective In view of the serious harm of sepsis-related liver injury (SALI), this study aims to use lasso logistic regression to build a SALI prediction model and to develop targeted prevention and control strategies. Methods A total of 192 sepsis patients admitted to the First Affiliated Hospital of Anhui Medical University from April 1, 2024 to April 1, 2025 were selected as the research subjects. They were divided into the SALI group (40 cases) and the non SALI group (152 cases) based on whether liver injury occurred. Clinical data from two groups of patients were collected, and univariate analysis was performed to identify factors associated with the occurrence of SALI in sepsis patients. The risk factors of SALI were analyzed by lasso logistic regression, and the SALI risk prediction model was established and verified by ROC curve. Results Significant differences were observed between the two groups in the scores of diabetes, mechanical ventilation, shock, hypoxemia, blood urea, platelet count, blood lactic acid, and sequential organ failure assessment (SOFA) scores (P < 0.05). Lasso-logistic regression analysis showed that diabetes mellitus, combined shock, blood lactic acid, and SOFA scores were independent risk factors for SALI (P < 0.05). ROC curve analysis results showed that SALI can be predicted by combining diabetes, combined shock, blood lactic acid, SOFA scores, and the nomogram model were predictive of SALI, with area under the curve values of 0.615, 0.620, 0.843, 0.937, 0.948, respectively. Among these, the nomogram-based prediction model exhibited the highest predictive performance. Conclusion Diabetes mellitus, combined shock, blood lactic acid levels, and SOFA score are independent risk factors for SALI. The nomogram model constructed based on these factors shows high predictive accuracy and clinical applicability, and may facilitate the development of accurate prevention and control strategies for SALI patients.
- Sepsis related liver injury /
- Lasso regression /
- Logistic regression /
- Predictive model /
- Response strategy
利益冲突 无

HTML全文

图 1 脓毒症患者发生SALI的Lasso回归模型构建

Figure 1. Construction of lasso regression model for SALI in patients with sepsis

下载: 全尺寸图片幻灯片

图 2 脓毒症患者发生SALI的Lasso回归模型筛选

Figure 2. Screening of lasso regression model for SALI in patients with sepsis

下载: 全尺寸图片幻灯片

图 3 脓毒症患者发生SALI的列线图预测模型

Figure 3. Nomogram prediction model of SALI in patients with sepsis

下载: 全尺寸图片幻灯片

图 4 列线图模型预测脓毒症患者SALI发生风险的ROC曲线

Figure 4. ROC curve of the line graph model for predicting the risk of SALI in patients with sepsis

下载: 全尺寸图片幻灯片

图 5 SALI风险预测列线图模型的校准曲线

Figure 5. Calibration curve of sali risk prediction nomogram model

下载: 全尺寸图片幻灯片

图 6 SALI风险预测列线图模型的决策曲线

Figure 6. Decision curve of sali risk prediction nomogram model

下载: 全尺寸图片幻灯片

表 1 SALI组和NSALI组脓毒症患者临床指标比较

Table 1. Comparison of clinical indicators between SALI group and NSALI group of sepsis patients

项目	NSALI组(n=152)	SALI组(n=40)	统计量	P值
性别[例(%)]			0.376^a	0.540
男性	88(57.89)	21(52.50)
女性	64(42.11)	19(47.50)
年龄(x±s，岁)	54.18±6.39	53.05±5.47	1.027^b	0.306
BMI(x±s)	23.54±2.26	23.75±2.39	0.533^b	0.594
合并糖尿病[例(%)]			10.792^a	0.001
是	22(14.47)	15(37.50)
否	130(85.53)	25(62.50)
合并高血压[例(%)]			1.190^a	0.275
是	20(13.16)	8(20.00)
否	132(86.84)	32(80.00)
合并冠心病[例(%)]			0.068^a	0.794
是	24(15.79)	7(17.50)
否	128(84.21)	33(82.50)
饮酒史[例(%)]			0.935^a	0.333
是	31(20.39)	11(27.50)
否	121(79.61)	29(72.50)
吸烟史[例(%)]			1.362^a	0.243
是	29(19.08)	11(27.50)
否	123(80.92)	29(72.50)
机械通气[例(%)]			7.091^a	0.008
是	91(59.87)	33(82.50)
否	61(40.13)	7(17.50)
原发感染部位[例(%)]			0.258^a	0.879
腹腔感染	90(59.21)	22(55.00)
肺部感染	46(30.26)	13(32.50)
其他	16(10.53)	5(12.50)
合并休克[例(%)]			9.429^a	0.002
是	32(21.05)	18(45.00)
否	120(78.95)	22(55.00)
低氧血症[例(%)]			10.232^a	0.001
是	28(18.42)	17(42.50)
否	124(81.58)	23(57.50)
WBC(x±s, ×10⁹/L)	15.60±3.48	14.24±4.52	1.767^b	0.083
尿素(x±s, mmol/L)	9.94±3.09	12.16±4.14	3.169^b	0.003
PLT(x±s, ×10⁹/L)	95.23±38.40	60.45±26.81	5.389^b	< 0.001
PCT(x±s, μg/L)	11.57±2.32	12.43±3.26	1.564^b	0.124
LAC(x±s, mmol/L)	2.06±0.38	3.12±0.99	6.630^b	< 0.001
SOFA评分(x±s，分)	4.78±1.47	8.53±3.49	6.648^b	< 0.001
APACHE Ⅱ评分(x±s，分)	18.61±4.70	20.35±6.09	1.684^b	0.098
注：^a为χ²值，^b为t值。

下载: 导出CSV

表 2 脓毒症患者发生SALI的多因素分析

Table 2. Multivariate analysis of SALI in patients with sepsis

变量	B	SE	Waldχ²	P值	OR(95% CI)
年龄	-0.074	0.063	1.374	0.241	0.929(0.821~1.051)
BMI	0.213	0.171	1.557	0.212	1.238(0.885~1.731)
合并糖尿病	1.742	0.846	4.243	0.039	5.710(1.088~29.959)
吸烟史	0.597	0.844	0.501	0.479	1.817(0.348~9.494)
机械通气	1.941	1.105	3.086	0.079	6.964(0.799~60.691)
合并休克	2.532	1.039	5.937	0.015	12.577(1.641~96.402)
低氧血症	1.507	0.888	2.880	0.090	4.513(0.792~25.716)
尿素	0.034	0.111	0.095	0.758	1.035(0.832~1.287)
PLT	-0.015	0.011	1.899	0.168	0.985(0.964~1.006)
LAC	2.657	0.777	11.688	0.001	14.259(3.108~65.419)
SOFA评分	0.954	0.284	11.297	0.001	2.595(1.488~4.526)
APACHEⅡ评分	0.042	0.083	0.252	0.616	1.043(0.886~1.227)

下载: 导出CSV

表 3 ROC曲线分析各指标对脓毒症患者发生SALI的预测价值

Table 3. Predictive value of ROC curve analysis of each indicator for SALI in patients with sepsis

项目	SE	P值	AUC	95% CI	cut-off值	约登指数	灵敏度	特异度
合并糖尿病	0.053	0.02	0.616	0.511~0.720	1.500	0.230	0.375	0.855
合并休克	0.052	0.020	0.620	0.517~0.722	1.500	0.239	0.450	0.789
LAC	0.049	< 0.001	0.843	0.747~0.940	2.610	0.708	0.800	0.908
SOFA评分	0.043	< 0.001	0.837	0.753~0.921	6.500	0.595	0.700	0.895
列线图模型	0.024	< 0.001	0.948	0.902~0.995	0.520	0.812	0.825	0.987

下载: 导出CSV

参考文献(21)

[1]	罗微, 张祖隆, 雷毅. 维持性血液透析的终末期肾病患者并发营养不良的影响因素[J]. 贵州医科大学学报, 2025, 50(6): 909-916. LUO W, ZHANG Z L, LEI Y. Influencing factors of malnutrition in maintenance hemodialysis patients with end-stage renal disease[J]. Journal of Guizhou Medical University, 2025, 50(6): 909-916.
[2]	HUANG W, CHEN H, HE Q, et al. Nobiletin protects against ferroptosis to alleviate sepsis-associated acute liver injury by modulating the gut microbiota[J]. Food Funct, 2023, 14(16): 7692-7704. doi: 10.1039/D3FO01684F
[3]	SUN S J, WANG L, WANG J M, et al. Maresin1 prevents sepsis-induced acute liver injury by suppressing NF-κB/Stat3/MAPK pathways, mitigating inflammation[J]. Heliyon, 2023, 9(11): e21883. DOI: 10.1016/j.heliyon.2023.e21883.
[4]	蔡稳, 刘通, 马广宇, 等. APRI联合血小板计数短期变化率对脓毒症相关肝损伤的预测价值[J]. 医学研究杂志, 2022, 51(6): 67-71, 77. CAI W, LIU T, MA G Y, et al. Predictive value of Apri combined with short-term change rate of platelet count for sepsis related liver injury[J]. Journal of Medical Research, 2022, 51(6): 67-71, 77.
[5]	LIANG H Y, SONG H, ZHANG X J, et al. Metformin attenuated sepsis-related liver injury by modulating gut microbiota[J]. Emerg Microbes Infect, 2022, 11(1): 815-828. doi: 10.1080/22221751.2022.2045876
[6]	LI Q L, ZHANG F P, WANG H, et al. NEDD4 lactylation promotes APAP induced liver injury through Caspase11 dependent non-canonical pyroptosis[J]. Int J Biol Sci, 2024, 20(4): 1413-1435. doi: 10.7150/ijbs.91284
[7]	SEYMOUR C W, LIU V X, IWASHYNA T J, et al. Assessment of clinical criteria for sepsis: for the third international consensus definitions for sepsis and septic shock (sepsis-3)[J]. JAMA, 2016, 315(8): 762-774. doi: 10.1001/jama.2016.0288
[8]	中国医药生物技术协会药物性肝损伤防治技术专业委员会, 中华医学会肝病学分会药物性肝病学组, 茅益民. 中国药物性肝损伤诊治指南(2023年版)[J]. 胃肠病学, 2023, 28(7): 397-431. The Professional Committee on Drug-Induced Liver Injury Prevention and Treatment of the China Association for Biomedical Technology, the Drug-Induced Liver Disease Group of the Hepatology Branch of the Chinese Medical Association, MAO Y M. Chinese guidelines for diagnosis and treatment of drug-induced liver injury (2023 edition)[J]. Gastroenterology, 2023, 28(7): 397-431.
[9]	郭莉, 颜南光, 刘锦明, 等. ICU脓毒症相关性肝损伤的临床特点及危险因素分析[J]. 海南医学, 2022, 33(12): 1528-1530. GUO L, YAN N G, LIU J M, et al. Clinical characteristics and risk factors of sepsis related liver injury in ICU[J]. Hainan Medical Journal, 2022, 33(12): 1528-1530.
[10]	ZHANG Z, TAN X J, SHI H Q, et al. Bibliometric study of sepsis-associated liver injury from 2000 to 2023[J]. World J Gastroenterol, 2024, 30(30): 3609-3624. doi: 10.3748/wjg.v30.i30.3609
[11]	KENIG A, KEIDAR-HARAN T, AZMANOV H, et al. Low-dose colchicine attenuates sepsis-induced liver injury: a novel method for alleviating systemic inflammation[J]. Inflammation, 2023, 46(3): 963-974. doi: 10.1007/s10753-023-01783-9
[12]	陈文胜, 杨巧云, 俞建峰, 等. 早期乳酸/前白蛋白比值对脓毒症相关性肝损伤的预测价值分析[J]. 中华急诊医学杂志, 2024, 33(11): 1559-1565. CHENG W S, YANG Q Y, YU J F, et al. Predictive value of early lactate/prealbumin ratio for sepsis related liver injury[J]. Chinese Journal of emergency medicine, 2024, 33(11): 1559-1565.
[13]	陈闪闪, 嵇金陵, 潘胜男, 等. 脓毒症患者肝损伤风险预测模型的构建和验证[J]. 江苏大学学报(医学版), 2025, 35(1): 62-69. CHEN S S, JI J L, PAN S N, et al. Construction and validation of liver injury risk prediction model in patients with sepsis[J]. Journal of Jiangsu University(Medical Edition), 2025, 35(1): 62-69.
[14]	YANG Y P, WANG A X, ZHOU J, et al. LncRNA SNHG11 induces ferroptosis in liver injury cells through miR-324-3p/GPX4 axis-mediated sepsis[J]. Cell Mol Biol, 2023, 69(12): 163-169. doi: 10.14715/cmb/2023.69.12.26
[15]	刘志红, 覃亚勤, 苏明华. 脓毒症相关肝损伤的研究进展[J]. 中国热带医学, 2023, 23(12): 1358-1364. LIU Z H, QIN Y Q, SU M H. Research progress of sepsis related liver injury[J]. China Tropical Medicine, 2023, 23(12): 1358-1364.
[16]	钟朕, 欧阳训彦, 郭起财, 等. 核因子E2相关因子2蛋白通过调控铁死亡介导脓毒症相关性肝损伤[J]. 中华危重病急救医学, 2024, 36(5): 491-495. ZHONG Z, OUYANG X Y, GUO Q C, et al. Nuclear factor E2-related factor 2 protein mediates sepsis-associated liver injury by regulating ferroptosis[J]. Chinese Critical Care Medicine, 2024, 36(5): 491-495.
[17]	朱玉菡, 王旭升, 宋麦芬, 等. 重症监护室老年脓毒症患者合并急性肝损伤的特征及影响因素[J]. 中华老年多器官疾病杂志, 2023, 22(5): 362-366. ZHU Y H, WANG X S, SONG M F, et al. Characteristics and influencing factors of acute liver injury in elderly patients with sepsis in intensive care unit[J]. Chinese Journal of Multiple Organ Diseases in the Elderly, 2023, 22(5): 362-366.
[18]	符艺影, 冯小伟, 林坚, 等. 血清miR-122a、MIF和HMGB1对儿童脓毒症并发肝损伤的预测价值[J]. 中华医院感染学杂志, 2024, 34(9): 1428-1432. FU Y Y, FENG X W, LIN J, et al. Predictive value of serum miR-122a, MIF and HMGB1 for liver injury in children with sepsis[J]. Chinese Journal of Nosocomial Infection, 2024, 34(9): 1428-1432.
[19]	何翠霞, 朱敏慧, 徐媛媛, 等. 跨膜蛋白TMEM88在LPS/D-GaIN诱导的急性肝损伤中的作用[J]. 中华全科医学, 2024, 22(5): 786-790. doi: 10.16766/j.cnki.issn.1674-4152.003503 HE C X, ZHU M H, XU Y Y, et al. Role of transmembrane protein tmem88 in LPS/D-GaIN induced acute liver injury[J]. Chinese Journal of General Practice, 2024, 22(5): 786-790. doi: 10.16766/j.cnki.issn.1674-4152.003503
[20]	SONG Y, MAGED ABDULSALAM MOHAMMED ALI A M, YANG W Y, et al. Clinical characteristics and prognosis of patients with early sepsis-related liver injury in Northeast China[J]. J Intensive Care Med, 2025, 40(3): 253-262. doi: 10.1177/08850666241277512
[21]	申变红, 汪程鹏, 高可润, 等. 精神分裂症患者长期服用第二代抗精神病药物导致肝损伤的影响因素分析[J]. 中华全科医学, 2023, 21(10): 1693-1697. doi: 10.16766/j.cnki.issn.1674-4152.2016.04.014 SHEN B H, WANG C P, GAO K R, et al. Analysis of influencing factors of liver injury caused by long-term use of second-generation antipsychotics in patients with schizophrenia[J]. Chinese Journal of General Practice, 2023, 21(10): 1693-1697. doi: 10.16766/j.cnki.issn.1674-4152.2016.04.014