1
|
Qiu Y, Cheng S, Wu Y, Yan W, Hu S, Chen Y, Xu Y, Chen X, Yang J, Chen X, Zheng H. Development of rapid and effective risk prediction models for stroke in the Chinese population: a cross-sectional study. BMJ Open 2023; 13:e068045. [PMID: 36858471 PMCID: PMC9980356 DOI: 10.1136/bmjopen-2022-068045] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 03/03/2023] Open
Abstract
OBJECTIVES The purpose of this study was to use easily obtained and directly observable clinical features to establish predictive models to identify patients at increased risk of stroke. SETTING AND PARTICIPANTS A total of 46 240 valid records were obtained from 8 research centres and 14 communities in Jiangxi province, China, between February and September 2018. PRIMARY AND SECONDARY OUTCOME MEASURES The area under the receiver operating characteristic curve (AUC), sensitivity, specificity and accuracy were calculated to test the performance of the five models (logistic regression (LR), random forest (RF), decision tree (DT), extreme gradient boosting (XGBoost) and gradient boosting DT). The calibration curve was used to show calibration performance. RESULTS The results indicated that XGBoost (AUC: 0.924, accuracy: 0.873, sensitivity: 0.776, specificity: 0.916) and RF (AUC: 0.924, accuracy: 0.872, sensitivity: 0.778, specificity: 0.913) demonstrated excellent performance in predicting stroke. Physical inactivity, hypertension, meat-based diet and high salt intake were important prediction features of stroke. CONCLUSION The five machine learning models all had good predictive and discriminatory performance for stroke. The performance of RF and XGBoost was slightly better than that of LR, which was easier to interpret and less prone to overfitting. This work provides a rapid and accurate tool for stroke risk assessment, which can help to improve the efficiency of stroke screening medical services and the management of high-risk groups.
Collapse
Affiliation(s)
- Yuexin Qiu
- School of Public Health, Nanchang University, Nanchang, Jiangxi, China
- Key Laboratory of Preventive Medicine, Nanchang University, Nanchang, Jiangxi, China
| | - Shiqi Cheng
- Neurosurgery Department, Nanchang University Second Affiliated Hospital, Nanchang, Jiangxi, China
| | - Yuhang Wu
- Department of Epidemiology and Health Statistics, Central South University, Changsha, Hunan, China
| | - Wei Yan
- Institute of Chronic Non-communicable Diseases, Center for Disease Control and Prevention of Jiangxi Province, Nanchang, Jiangxi, China
| | - Songbo Hu
- School of Public Health, Nanchang University, Nanchang, Jiangxi, China
- Key Laboratory of Preventive Medicine, Nanchang University, Nanchang, Jiangxi, China
| | - Yiying Chen
- Institute of Chronic Non-communicable Diseases, Center for Disease Control and Prevention of Jiangxi Province, Nanchang, Jiangxi, China
| | - Yan Xu
- Institute of Chronic Non-communicable Diseases, Center for Disease Control and Prevention of Jiangxi Province, Nanchang, Jiangxi, China
| | - Xiaona Chen
- Institute of Chronic Non-communicable Diseases, Center for Disease Control and Prevention of Jiangxi Province, Nanchang, Jiangxi, China
| | - Junsai Yang
- School of Public Health, Nanchang University, Nanchang, Jiangxi, China
- Key Laboratory of Preventive Medicine, Nanchang University, Nanchang, Jiangxi, China
| | - Xiaoyun Chen
- School of Public Health, Nanchang University, Nanchang, Jiangxi, China
- Key Laboratory of Preventive Medicine, Nanchang University, Nanchang, Jiangxi, China
| | - Huilie Zheng
- School of Public Health, Nanchang University, Nanchang, Jiangxi, China
- Key Laboratory of Preventive Medicine, Nanchang University, Nanchang, Jiangxi, China
| |
Collapse
|
2
|
Binary cutpoint and the combined effect of systolic and diastolic blood pressure on cardiovascular disease mortality: A community-based cohort study. PLoS One 2022; 17:e0270510. [PMID: 35771898 PMCID: PMC9246156 DOI: 10.1371/journal.pone.0270510] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/08/2022] [Accepted: 06/12/2022] [Indexed: 11/24/2022] Open
Abstract
Objectives This study aimed to examine the risk of cardiovascular disease (CVD) death according to blood pressure levels and systolic and/or diastolic hypertension. Methods From 20,636 cohort participants, 14,375 patients were enrolled after patients with prior hypertension on antihypertensive drugs were excluded. For the combination analysis, participants were divided into four groups (systolic/diastolic hypertension, systolic hypertension only, diastolic hypertension only, and non-hypertension). The risk of CV death was calculated using the hazard ratio (HR) and 95% confidence intervals (95% CI) in a Cox regression model. Results The risk of CVD death increased in systolic hypertension (HR = 1.59, 95% CI 1.26–2.00) and systolic/diastolic hypertension (HR = 1.84, 95% CI 1.51–2.25). The highest risks of hemorrhagic and ischemic stroke were observed in the diastolic hypertension (HR = 4.11, 95% CI 1.40–12.06) and systolic/diastolic hypertension groups (HR = 2.59, 95% CI 1.92–3.50), respectively. The risk of CVD death was drastically increased in those with SBP≥120 mmHg/DBP≥80 mmHg. The highest risk was observed in those with SBP of 130–131 mmHg and 134–137 mmHg. Conclusion The combined analysis of systolic and/or diastolic hypertension appears to be a good predictor of CVD death. The risk of CVD death in the prehypertensive group could be carefully monitored as well as in the hypertensive group, presumably due to less attention and the lack of antihypertensive treatment.
Collapse
|
3
|
Varying combination of feature extraction and modified support vector machines based prediction of myocardial infarction. EVOLVING SYSTEMS 2022. [DOI: 10.1007/s12530-021-09410-4] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/19/2022]
|
4
|
Park EH, Gil YJ, Kim C, Kim BJ, Hwang SS. Presence of Thrombectomy-capable Stroke Centers Within Hospital Service Areas Explains Regional Variation in the Case Fatality Rate of Acute Ischemic Stroke in Korea. J Prev Med Public Health 2021; 54:385-394. [PMID: 34875821 PMCID: PMC8655371 DOI: 10.3961/jpmph.21.329] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/14/2021] [Accepted: 08/12/2021] [Indexed: 11/09/2022] Open
Abstract
OBJECTIVES This study aimed to explore the status of regional variations in acute ischemic stroke (AIS) treatment and investigate the association between the presence of a thrombectomy-capable stroke center (TSC) and the case fatality rate (CFR) of AIS within hospital service areas (HSAs). METHODS This observational cross-sectional study analyzed acute stroke quality assessment program data from 262 hospitals between 2013 and 2016. TSCs were defined according to the criteria of the Joint Commission. In total, 64 HSAs were identified based on the addresses of hospitals. We analyzed the effects of structure factors, process factors, and the presence of a TSC on the CFR of AIS using multivariate logistic regression. RESULTS Among 262 hospitals, 31 hospitals met the definition of a TSC. Of the 64 HSAs, only 20 had a TSC. At hospitals, the presence of a stroke unit, the presence of stroke specialists, and the rate of endovascular thrombectomy (EVT) treatment were associated with reductions in the CFR. In HSAs, the rate of EVT treatment (odds ratio [OR], 0.98; 95% confidence interval [CI], 0.97 to 0.99) and the presence of a TSC (OR, 0.93; 95% CI, 0.88 to 0.99) significantly reduced the CFR of AIS. CONCLUSIONS The presence of a TSC within an HSA, corresponding to structure and process factors related to the quality of care, contributed significantly to lowering the CFR of AIS. The CFR also declined as the rate of treatment increased. This study highlights the importance of TSCs in the development of an acute stroke care system in Korea.
Collapse
Affiliation(s)
- Eun Hye Park
- Department of Preventive Medicine, Graduate School of Public Health, Seoul National University, Seoul, Korea
| | - Yong Jin Gil
- Department of Preventive Medicine, Graduate School of Public Health, Seoul National University, Seoul, Korea
| | - Chanki Kim
- Department of Preventive Medicine, Graduate School of Public Health, Seoul National University, Seoul, Korea
| | - Beom Joon Kim
- Department of Neurology, Seoul National University Bundang Hospital, Seongnam, Korea
| | - Seung-Sik Hwang
- Department of Preventive Medicine, Graduate School of Public Health, Seoul National University, Seoul, Korea
| |
Collapse
|
5
|
Kim HC. Epidemiology of cardiovascular disease and its risk factors in Korea. Glob Health Med 2021; 3:134-141. [PMID: 34250288 DOI: 10.35772/ghm.2021.01008] [Citation(s) in RCA: 29] [Impact Index Per Article: 9.7] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/27/2021] [Revised: 03/27/2021] [Accepted: 04/02/2021] [Indexed: 01/04/2023]
Abstract
Cardiovascular disease (CVD) is the leading cause of death and a major contributor to disability worldwide. Currently, Korea is among countries with the lowest CVD mortality rates, and the age-adjusted CVD mortality rate is still decreasing. However, depending on the CVD type, the mortality and incidence trends vary. Without age-standardization, cerebrovascular disease mortality peaked in 1994 (82.1 per 100K) and continued to decline until 2018 (44.7 per 100K), while heart disease mortality recorded the lowest level in 2001 (44.9 per 100K) then increased again until 2018 (74.5 per 100K). Age-standardized mortality rates showed different trends: both cerebrovascular disease and heart disease mortality rates have declined over the past few decades, although the rate of decline varies. Based on the National Health Insurance claim database, the numbers of hospitalization for cerebrovascular disease and ischemic heart disease are increasing, but the age-standardized hospitalization rates are decreasing. Unlike other types of CVDs, heart failure is rapidly increasing in both mortality and hospitalization rates regardless of age-standardization. Seventy percent of Korean adults have at least one risk factor, 41% have ≥ 2 risk factors, and 19% have ≥ 3 risk factors including hypertension, diabetes, hypercholesterolemia, obesity, and smoking. Exposure to multiple risk factors increases with age, with 65% of senior citizens over 70 having ≥ 2 risk factors and 34% having ≥ 3 risk factors. As the elderly population, especially those with multiple risk factors and chronic disorders, is increasing, the management of this high-risk group will be an important challenge to prevent CVD in Korea.
Collapse
Affiliation(s)
- Hyeon Chang Kim
- Department of Preventive Medicine, Integrative Research Center for Cerebrovascular and Cardiovascular Diseases, Yonsei University College of Medicine, Seoul, Korea
| |
Collapse
|