Comparison of Prediction Models for In-hospital Stroke Recurrence in Patients with Ischemic Stroke Based on
Logistic Regression and XGBoost Methods

doi:10.3969/j.issn.1673-5765.2020.06.003

Abstract

Abstract:

Objective To compare prediction models for in-hospital stroke recurrence in patients with ischemic stroke based on logistic regression and XGBoost methods. Methods Data of ischemic stroke inpatients discharged according to medical advice from China National Stroke Registry Ⅱ (CNSR Ⅱ) database were retrospectively analyzed. Logistic regression and XGBoost methods were used to develop a model for predicting in-hospital stroke recurrence. Candidate predictors included demographic characteristics, stroke severity, medical history, medication history, and clinical measure indicators. The performance measures of the predictive models included the area under the receiver operating characteristic curve (AUC), calibration intercept, calibration slope and Brier score. All statistical analysis was performed using R (version 3.6.2). Results A total of 17 227 eligible patients were included in this analysis. The mean age was 64.72±11.84 years, and 6317 (36.7%) cases were females. A total of 14 482 (84.1%) patients had a mRS score of 0 or 1 point before symptoms onset, and the NIHSS score at admission was 4 (2-6). A total of 444 (2.6%) patients had recurrent stroke during hospitalization. The three leading strong predictors were mRS score, atrial fibrillation and stroke history in logistic regression model, and mRS score, atrial fibrillation and total cholesterol in XGBoost model. No significant difference was found in AUC between logistic regression model and XGBoost model (0.63, 95%CI 0.58-0.68 vs 0.64, 95%CI 0.59-0.68, P =0.9229). The calibration intercept, calibration slope and Brier score in logistic regression model were -0.81, 0.76 and 0.03, respectively; and were -1.37, 1.20 and 0.38 in XGBoost model. Logistic regression model had better calibration than XGBoost model. Conclusions No significant difference was found in discrimination between logistic-based prediction model and XGBoost-based prediction model for in-hospital stroke recurrence constructed using data of CNSR II, while logistic-based prediction model had better calibration.

Key words: Ischemic stroke; In-hospital stroke recurrence; Prediction model

摘要：

目的基于Logistic回归和XGBoost方法构建缺血性卒中院内复发风险预测模型，并进行初步比较。方法利用中国国家卒中登记Ⅱ（China National Stoke Registry Ⅱ，CNSRⅡ）数据库中按医嘱离院的缺血性卒中患者数据，分别基于Logistic回归和XGBoost方法构建缺血性卒中院内复发风险预测模型。备选的预测因子包括人口学特征、卒中严重程度、既往病史、用药史以及临床测量指标。模型的评价指标包括ROC曲线下面积（area under the cure，AUC）、校准截距、校准斜率以及Brier得分。所有统计分析均在R（3.6.2版）中完成。结果最终纳入17 227例符合条件的患者，平均年龄64.72±11.84岁，女性6317例（36.7%），发病前 mRS评分为0或1分的病例14 482例（84.1%），入院NIHSS评分4（2～6）分，院内卒中复发444例（2.6%）。预测模型识别的前三位强预测因子，在Logistic回归中为发病前mRS评分、心房颤动及卒中史；在 XGBoost中为发病前mRS评分、心房颤动及总胆固醇。Logistic回归预测模型与XGBoost预测模型的AUC无显著差异（0.63，95%CI 0.58～0.68 vs 0.64，95%CI 0.59～0.68，P =0.9229）。Logistic预测模型校准截距、校准斜率以及Brier得分分别为-0.81、0.76和0.03；XGBoost预测模型的校准截距、校准斜率以及 Brier得分分别为-1.37、1.20和0.38。Logistic预测模型校准度更好。结论利用CNSRⅡ数据构建的缺血性卒中院内复发风险预测模型应用中，基于XGBoost方法构建的预测模型相比Logistic回归构建的预测模型的区分度没有显著差异，但校准度略低。

关键词: 缺血性卒中; 院内复发; 预测模型

GU Hong-Qiu, WANG Chun-Juan, LI Zi-Xiao, WANG Yi-Long,WANG Yong-Jun, JIANG Yong. Comparison of Prediction Models for In-hospital Stroke Recurrence in Patients with Ischemic Stroke Based on Logistic Regression and XGBoost Methods[J]. Chinese Journal of Stroke, 2020, 15(06): 587-594.

谷鸿秋，王春娟，李子孝，王伊龙，王拥军，姜勇. 基于Logistic回归与XGBoost构建缺血性卒中院内复发风险预测模型的初步比较研究[J]. 中国卒中杂志, 2020, 15(06): 587-594.

References

[1] GBD 2016 Lifetime Risk of Stroke Collaborators，

FEIGIN V L，NGUYEN G，et al. Global，regional，

and country-specific lifetime risks of stroke，1990

and 2016[J]. N Engl J Med，2018，379（25）：2429-

2437.

[2] 国家卫生计生委脑卒中防治工程委员会. 中国脑卒中

防治报告（2015）[M]. 北京：中国协和医科大学出版

社，2015：6-7.

[3] ZHOU M G，WANG H D，ZENG X Y，et al.

Mortality，morbidity，and risk factors in China and

its provinces，1990-2017：a systematic analysis for

the Global Burden of Disease Study 2017[J]. Lancet，

2019，394（10204）：1145-1158.

[4] WANG Y J，LI Z X，WANG Y L，et al. Chinese

Stroke Center Alliance：a national effort to improve

healthcare quality for acute stroke and transient

ischaemic attack：rationale，design and preliminary

findings[J]. Stroke Vasc Neurol，2018，3（4）：256-

262.

[5] PAN Y S，JING J，CHEN W Q，et al. Risks and

benefits of clopidogrel-aspirin in minor stroke or TIA：

time course analysis of chance[J]. Neurology，2017，

88（20）：1906-1911.

[6] CHEN T Q，GUESTRIN C. Xgboost：a scalable

tree boosting system[C]//Association for Computing

Machinery Special Interest Group on management

of data，ACM Special Interest Group on Knowledge

Discovery in Data. Proceedings of the 22nd ACM

SIGKDD International Conference on Knowledge

Discovery and Data Mining. New York：NY ACM，

2016：785-794.

[7] LI Z X，WANG C J，ZHAO X Q，et al. Substantial

progress yet significant opportunity for improvement

in stroke care in China[J]. Stroke，2016，47（11）：

2843-2849.

[8] POWERS W J，RABINSTEIN A A，ACKERSON

T，et al. Guidelines for the early management of

patients with acute ischemic stroke：2019 update

to the 2018 guidelines for the early management of

acute ischemic stroke：a guideline for healthcare

professionals from the American Heart Association/

American Stroke Association[J/OL]. Stroke，2019，

50（12）：e344-e418[2020-02-05]. https：//doi.org/10.

1161/STR.0000000000000211.

[9] CHAUDHARY D，ABEDI V，LI J，et al. Clinical

risk score for predicting recurrence following a

cerebral ischemic event[J]. Front Neurol，2019，10：

1106.

[10] LUNDBERG S M，LEE S I. A unified approach

to interpreting model predictions[C]//Advances in

Neural Information Processing Systems. California：

NIPS，2017：4765-4774.

[11] BRIER G W. Verification of forecasts expressed in

terms of probability[J]. Mon Weather Rev，1950，78：

1-3.

[12] AL'AREF S J，MALIAKAL G，SINGH G，et

al. Machine learning of clinical variables and

coronary artery calcium scoring for the prediction

of obstructive coronary artery disease on coronary

computed tomography angiography：analysis from

the CONFIRM registry[J]. Eur Heart J，2020，41（3）：

359-367.

[13] GE Y Q，WANG Q H，WANG L，et al. Predicting

post-stroke pneumonia using deep neural network

approaches[J]. Int J Med Inform，2019，132：103986.

[14] HONG W S，HAIMOVICH A D，TAYLOR R

A. Predicting hospital admission at emergency

department triage using machine learning[J/OL].

PLoS One，2018，13（7）：e0201016[2020-02-05].

https：//doi.org/10.1371/journal.pone.0201016.

[15] ELFIKY A A，PANY M J，PARIKH R B，et al.

Development and application of a machine learning approach to assess short-term mortality risk among

patients with cancer starting chemotherapy[J/OL].

JAMA Netw Open，2018，1（3）：e180926[2020-02-

05]. https：//doi.org/10.1001/jamanetworkopen.2018.

0926.

[16] WANG A X，XU J，CHEN G J，et al. Oxidized

low-density lipoprotein predicts recurrent stroke in

patients with minor stroke or TIA[J/OL]. Neurology，

2018，91（10）：e947-e955[2020-02-05]. https：//doi.

org/10.1212/WNL.0000000000006118.

[17] ZHU B H，PAN Y S，JING J，et al. Neutrophil

counts，neutrophil ratio，and new stroke in minor

ischemic stroke or TIA[J/OL]. Neurology，2018，90

（21）：e1870-e1878[2020-02-05]. https：//doi.org/10.

1212/WNL.0000000000005554.

[18] ZHU B H，LIU H H，PAN Y S，et al. Elevated

neutrophil and presence of intracranial artery

stenosis increase the risk of recurrent stroke[J].

Stroke，2018，49（10）：2294-2300.

[19] LI J J，ZHAO X Q，MENG X，et al. High-sensitive

C-reactive protein predicts recurrent stroke and poor

functional outcome：subanalysis of the Clopidogrel

in High-Risk Patients With Acute Nondisabling

Cerebrovascular Events Trial[J]. Stroke，2016，47

（8）：2025-2030.

[20] PAN Y S，CAI X L，JING J，et al. Stress

hyperglycemia and prognosis of minor ischemic

stroke and transient ischemic attack：the CHANCE

study（Clopidogrel in High-Risk Patients With Acute

Nondisabling Cerebrovascular Events）[J]. Stroke，

2017，48（11）：3006-3011.

[21] 王拥军. 脑卒中诊疗王拥军2019观点[M]. 北京：科

学技术文献出版社，2019：137-152.

[22] BONNETT L J，SNELL K I E，COLLINS G S，et

al. Guide to presenting clinical prediction models for

use in clinical settings[J]. BMJ，2019，365：kl737.

Comparison of Prediction Models for In-hospital Stroke Recurrence in Patients with Ischemic Stroke Based on Logistic Regression and XGBoost Methods

基于Logistic回归与XGBoost构建缺血性卒中院内复发风险预测模型的初步比较研究

PDF

Knowledge

Abstract

Cite this article

share this article

References

Related Articles 15

Recommended Articles

Metrics

[1]	WU Yuqian, ZHANG Yumei, FAN Xiaowei, WANG Anxin, PANG Wenbin. The Efficacy of Kinesthetic Motor Imagery Therapy for Hand Function Rehabilitation of Stroke Patients [J]. Chinese Journal of Stroke, 2022, 17(12): 1366-1371.
[2]	YI Luo, ZHOU Hongyu, QIU Xin, LI Zixiao, WANG Yongjun. DNA Methylation Changes in Atherosclerotic Diseases [J]. Chinese Journal of Stroke, 2022, 17(11): 1163-1170.
[3]	WANG Yubo, LI Zixiao, WANG Yongjun. Progress of the Relationship between Epigenetics and Oxidative Stress after Ischemic Stroke [J]. Chinese Journal of Stroke, 2022, 17(11): 1171-1177.
[4]	CUI Lingyun, QIU Xin, ZHOU Hongyu, LI Zixiao, WANG Yongjun. Progress of the Relationship between DNA Methylation and Type 2 Diabetes [J]. Chinese Journal of Stroke, 2022, 17(11): 1183-1188.
[5]	LIU Guangliang, HAN Kaiyue, SU Wenlong, ZHANG Hao. Effects of Aerobic Cycling Training Intensity on Motor and Cardiopulmonary Fitness in Patients with Ischemic Stroke [J]. Chinese Journal of Stroke, 2022, 17(11): 1203-1208.
[6]	XU Yingxin, ZHANG Shuang, ZHOU Xiaomei, YUAN Jinglin, ZHAO Liuzhuang, YANG Haihua. Correlation of Non-Stenotic Carotid Plaque with Anterior Circulation Embolic Stroke of Undetermined Source [J]. Chinese Journal of Stroke, 2022, 17(11): 1233-1237.
[7]	CHENG Shi, LI Jing, GUO Junping, HU Aixiang, YU Xinwei, HAN Wei, LIANG Ying, ZHANG Yuewei, JI Ruijun. Clinical Characteristics and Risk Factors of Infection in Patients with Acute Stroke during Hospitalization [J]. Chinese Journal of Stroke, 2022, 17(10): 1065-1070.
[8]	XIANG Wei, XU Luyao, ZHANG Manman, WEI Hongchun, LIANG Zhigang. The Frequency and Influencing Factors of Asymmetrical Prominent Veins Sign on Susceptibility-Weighted Imaging in Acute Ischemic Stroke [J]. Chinese Journal of Stroke, 2022, 17(10): 1097-1102.
[9]	ZHAO Yongrui, LIU Zhanhui. Progress of Diffusion Tensor Imaging in Evaluating Motor Function Prognosis of Patients with Ischemic Stroke [J]. Chinese Journal of Stroke, 2022, 17(10): 1114-1119.
[10]	ZHANG Boyu, GUO Yucheng, CUI Yuxia, LI Yingli, LYU Yumei. Seeking Medical Services Decision-Making Status and Relevant Influencing Factors in Patients with Acute Ischemic Stroke: A Review [J]. Chinese Journal of Stroke, 2022, 17(10): 1127-1132.
[11]	WU Tingting, LIU Yafei, YAN Shixi, WU Yiping, ZHANG Zhongbo, ZAHNG Hanfei, FANG Pinping. Application of “Internet Plus” in Emergency Care of Acute Ischemic Stroke [J]. Chinese Journal of Stroke, 2022, 17(10): 1133-1138.
[12]	ZHANG Yige, ZHANG Wan, ZHANG Luyang, WANG Xin, DING Lan, WANG Yuying, PEI Lulu, FANG Hui, XU Yuming, SONG Bo. Association of Artery Occlusion and Recurrent Ischemic Stroke Related with Cerebral Artery Dissection [J]. Chinese Journal of Stroke, 2022, 17(09): 932-936.
[13]	LYU Wei, WANG Anxin, TIAN Xue, ZHANG Xiaoli, XU Yuyuan, MENG Xia. Effect of Xinxuening Capsule on Short-term Functional Outcome in Patients with Ischemic Stroke and High Homocysteine Level [J]. Chinese Journal of Stroke, 2022, 17(09): 972-976.
[14]	ZHANG Xinmiao, PAN Yuesong, XIANG Xianglong, WANG Yongjun, MENG Xia. Early Clinical Pathway Intervention to Improve the Key Performance Indicators of Medical Care for Ischemic Stroke [J]. Chinese Journal of Stroke, 2022, 17(09): 977-984.
[15]	GONG Yutian, LI Jiejie. Advances in the Relationship between IL-6 and Atherosclerotic Ischemic Stroke [J]. Chinese Journal of Stroke, 2022, 17(09): 1022-1026.