Construction of a Nomogram Model of Tsutsugamushi Disease Complicated with Sepsis Based on Lasso Regression
-
摘要:
目的 基于Lasso回归建立恙虫病合并脓毒症的列线图(Nomogram)模型,为恙虫病合并脓毒症的诊治提供参考。 方法 选取2012年6月至2023年12月昆明市第三人民医院收治的恙虫病患者作为研究对象(n = 235),其中恙虫病合并脓毒症患者作为实验组(n = 138),未合并脓毒症的恙虫病患者作为对照组(n = 97),利用Lasso回归筛选恙虫病合并脓毒症的影响因素,并通过Logistic回归构建列线图模型,使用受试者工作特征(receiver operating characteristic,ROC)曲线和校准曲线进行模型的效能评价,并通过临床决策曲线(decision curve analysis,DCA)进行临床效用分析。 结果 通过Logistic回归的多因素分析结果显示:age(OR = 1.039,95%CI:1.017~1.061)、PLT(OR = 0.995,95%CI:0.990~1.000)、UA(OR = 1.004,95%CI:1.001~1.008)、IgA(OR = 0.680,95%CI:0.472~0.979)为恙虫病合并脓毒症的独立影响因素。ROC曲线下面积(area under the curve,AUC)为0.721(95%CI:0.656~0.786,P < 0.001),模型的灵敏度为69.1%,特异度为63.1%。校准曲线提示该模型具有良好的一致性,临床决策曲线提示该模型具有较高的净收益值。 结论 恙虫病患者伴随age和UA水平的增高及PLT和IgA水平的降低,继发脓毒症的风险越高。 Abstract:Objective To establish a nomogram model for tsutsugamushi disease complicated with sepsis based on Lasso regression, so as to provide a reference for the diagnosis and treatment of tsutsugamushi disease complicated with sepsis. Methods Selecting patients with scrub typhus patients admitted to Kunming Third People’ s Hospital from June 2012 to December 2023 as the research subjects (n = 235), with scrub typhus patients patients with sepsis as the experimental group (n = 138) and scrub typhus patients patients without sepsis as the control group (n = 97), Lasso regression was used to screen the influencing factors of scrub typhus combined with sepsis, and a nomogram model was constructed through logistic regression. The effectiveness of the model was evaluated using receiver operating characteristic(ROC) curves and calibration curves, and clinical utility analysis was performed using decision curve analysis(DCA). Results The results of multivariate analysis through logistic regression showed: that age (OR = 1.039, 95%CI: 1.017~1.061), PLT (OR = 0.995, 95%CI: 0.990~1.000), UA (OR = 1.004, 95%CI: 1.001~1.008) and IgA (OR = 0.680, 95%CI: 0.472~0.979) were independent influencing factors of scrub typhus combined with sepsis. The area under the ROC curve (AUC) was 0.721 (95%CI: 0.656~0.786, P < 0.001), the sensitivity of the model was 69.1%, and the specificity was 63.1%. The calibration curve indicates that the model had good consistency, and the DCA indicated that the model had a high net benefit value. Conclusion Patients with scrub typhus are associated with increased age and UA levels and decreased PLT and IgA levels, and have a higher risk of secondary sepsis. -
Key words:
- Tsutsugamushi disease /
- Sepsis /
- Nomogram model
-
表 1 2组患者的一般资料比较[n(%)/M(P25,P75)/$\bar x \pm s $]
Table 1. Comparison of general information between two groups of patients [n(%)/M(P25,P75)/$\bar x \pm s $]
项目 样本总量(n=235) 观察组(n=138) 对照组(n=97) t/χ2/Z P 年龄(岁) 48.7±14.8 50.9±14.7 46.5±14.6 t=2.352 0.019* 性别 男性 137(58.3) 75(54.7) 62(63.9) χ2=2.146 0.143 女性 98(41.7) 63(45.3) 35(36.1) 居住地 市区 100(42.6) 58(41.8) 42(43.1) χ2=0.037 0.846 郊县 135(57.4) 80(58.2) 55(56.9) 民族 汉族 214(91.1) 122(88.1) 92(95.2) χ2=2.903 0.088 其他族 21(8.9) 16(11.9) 5(4.8) 婚姻 已婚 217(92.3) 130(94.2) 87(89.7) χ2=1.639 0.200 未婚 18(7.7) 8(5.8) 10(10.3) 职业 农民 157(66.8) 90(65.2) 67(69.1) χ2=0.382 0.537 非农 78(33.2) 48(34.8) 30(30.9) 发病时间(d) 14(9.5,19.0) 14(10,19.0) 14(9.0,19.0) Z=−0.445 0.656 发热 229(97.4) 135(97.8) 94(96.9) χ2=0.193 0.660 头痛 138(58.7) 80(58.0) 58(59.8) χ2=0.078 0.780 畏寒 139(59.1) 81(58.7) 58(59.8) χ2=0.028 0.866 乏力 106(45.1) 62(44.9) 44(45.4) χ2=0.004 0.948 肌痛 96(40.9) 56(40.6) 40(41.2) χ2=0.010 0.919 咳嗽 80(34.0) 48(34.8) 32(33.0) χ2=0.082 0.775 K+(mmol/L) 3.52±0.44 3.46±0.44 3.58±0.43 t=1.940 0.041* ALT(U/L) 82.50(50.25,126.0) 70(49,112.5) 92(53,137) Z =−1.925 0.044* IgA(g/L) 2.15(1.89,2.17) 2.17(1.67,2.20) 2.12(1.98,2.37) Z=−1.852 0.046* IgG(g/L) 10.29(9.17,10.97) 9.11(8.54,11.23) 10.34(9.86,11.45) Z=−1.762 0.018* IL-1β(pg/mL) 24.49(17.54,24.51) 23.12(22.15,25.18) 25.42(18.88,27.91) Z=−1.281 0.020* IL-5(pg/mL) 2.31(2.11,2.81) 1.15(1.11,2.61) 2.40(1.61,2.87) Z=−1.695 0.009* *P < 0.05。 表 2 恙虫病合并脓毒症的多因素Logistic回归分析
Table 2. Multivariate logistic regression analysis of tsutsugamushi disease complicated with sepsis
指标 β S.E. Wald P OR 95%CI age 0.038 0.011 12.584 <0.001* 1.039 1.017~1.061 PLT −0.005 0.002 4.553 0.033* 0.995 0.990~1.000 UA 0.004 0.002 5.319 0.021* 1.004 1.001~1.008 IgA −0.386 0.186 4.313 0.038* 0.680 0.472~0.979 IL-6 −0.004 0.002 2.627 0.105 0.996 0.991~1.001 IFN-γ 0.009 0.005 2.983 0.084 1.009 0.999~1.018 常量 −0.722 0.803 0.808 0.369 0.486 *P < 0.05。 -
[1] 王子文,张林娜,徐猛,等. 肺部感染并发脓毒症患者入院28 d内死亡预测模型的构建[J]. 山东医药,2023,63(3):37-43. doi: 10.3969/j.issn.1002-266X.2023.03.008 [2] 李兰娟,任红. 传染病学 [M]. 9版. 北京: 人民卫生出版社,2018: 142-150. [3] Braun D. A retrospective review of the sepsis definition after publication of sepsis-3[J]. Am J Med,2019,132(3):382-384. doi: 10.1016/j.amjmed.2018.11.003 [4] 梁桐, 刘莹, 李幼霞, 等. 恙虫病临床特征及重症危险因素分析[J]. 中国热带医学,2023,23(9):961-965. [5] Luce-Fedrow A,Lehman M L,Kelly D J,et al. A review of scrub typhus (orientia tsutsugamushi and related organisms): Then,now,and tomorrow[J]. Trop Med Infect Dis,2018,17(1):8-11. [6] 谢晓菲,王高玉,黄艺,等. 中国恙虫病流行及临床研究进展(2010-2020)[J]. 海南医学院学报,2023,29(19):1505-1509. [7] Lee N,Ip M,Wong B,et al. Risk factors associated with life-threatening rickettsial infections[J]. Am J Trop Med Hyg,2008,78(6):973-978. doi: 10.4269/ajtmh.2008.78.973 [8] Bhargava A,Kaushik R,Kaushik R M,et al. Scrub typhus in uttarakhand & adjoining uttar pradesh: Seasonality,clinical presentations & predictors of mortality[J]. Indian J Med Res,2016,144(6):901-909. doi: 10.4103/ijmr.IJMR_1764_15 [9] Hou N,Li M,He L,et al. Predicting 30-days mortality for MIMIC-III patients with sepsis-3: A machine learning approach using XGboost[J]. J Transl Med,2020,18(1):462-471. doi: 10.1186/s12967-020-02620-5 [10] Sharma R,Mahajan S K,Singh B,et al. Predictors of severity in scrub typhus[J]. J Assoc Physicians India,2019,67(4):35-38. [11] Zhang C,Shang X,Yuan Y,et al. Platelet-related parameters as potential biomarkers for the prognosis of sepsis[J]. Exp Ther Med,2023,25(3):133-140. doi: 10.3892/etm.2023.11832 [12] 路伟民,杨小涛,朱瑛,等. 儿童恙虫病175例的临床特征及重症恙虫病危险因素[J]. 昆明医科大学学报,2022,43(8):72-80. doi: 10.12259/j.issn.2095-610X.S20220814 [13] Verma S K,Gupta K K,Arya R K,et al. Clinical and biochemical profile of scrub typhus patients at a tertiary care hospital in Northern India[J]. J Family Med Prim Care,2021,10(3):1459-1465. doi: 10.4103/jfmpc.jfmpc_1162_20 [14] Mahajan S K,Sharma R,Singh B. Is hyperuricemia a marker of severity of disease in Scrub typhus[J]. J Assoc Physicians India,2022,69(12):11-12. [15] 王英,李梅,许汪斌,等. 云南省2017至2018年重症恙虫病临床特征分析[J]. 中华危重病急救医学,2019,31(8):1018-1023. doi: 10.3760/cma.j.issn.2095-4352.2019.08.021 [16] 文芳,司少魁. 醒脑静注射液联合乌司他丁、连续性肾脏替代治疗对脓毒症合并急性肾损伤患者的疗效、免疫功能及血流动力学的影响[J]. 临床肾脏病杂志,2024,24(1):25-32. doi: 10.3969/j.issn.1671-2390.2024.01.005 [17] Corona A,Richini G,Simoncini S,et al. Treating Critically Ⅲpatients experiencing SARS-CoV-2 severe infection with IgM and IgA enriched IgG infusion[J]. Antibiotics (Basel),2021,10(8):930-941. doi: 10.3390/antibiotics10080930 [18] Bhattacharya P K,Murti V S,Jamil M,et al. Clinical profile and determinants of scrub typhus presenting with sepsis based on sepsis-3 criteria[J]. J Vector Borne Dis,2020,57(4):307-313. doi: 10.4103/0972-9062.313963 -