Construction of Risk Prediction Model for Poor Prognosis among Hemorrhagic Stroke Patients:A Cross-sectional Study
10.3870/j.issn.1672-0741.22.11.021
- VernacularTitle:构建出血性卒中不良预后的风险预测模型:一项横断面研究
- Author:
Xinsheng LIU
1
;
Wangxiang JIANG
Author Information
1. 武汉市急救中心,武汉 430022
- Keywords:
hemorrhagic stroke;
prognosis;
predictive model;
cross-sectional study
- From:
Acta Medicinae Universitatis Scientiae et Technologiae Huazhong
2024;53(3):368-374
- CountryChina
- Language:Chinese
-
Abstract:
Objective To study the predictive factors for a poor prognosis of hemorrhagic stroke,build a prediction model,and improve the current evaluation system.Methods A cross-sectional survey was conducted from the perspective of pre-hospi-tal and intra-hospital integration.After single factor comparison,indicators with significant differences were sequentially includ-ed in univariant and multivariant Poisson regression analysis.The selected independent risk factors were constructed as predic-tive models.The prediction model was converted into a visual optimization rating scale in the form of a column chart to obtain the corresponding prediction probability for the corresponding rating.The ROC curve was used to test the effectiveness of opti-mizing scores and ICH-CT scores in predicting poor prognosis.Relevant data of patients in the validation group were extracted based on the two scoring information,and the final score of each validation group patient was obtained.The scoring and prog-nostic results were substituted into the ROC curve to evaluate the predictive ability of different prediction models,and the pre-diction models were converted into visual optimization scoring scales to quantify the probability of adverse prognosis outcomes.The ICH-CT model was used as a reference to explore the effectiveness of optimizing scoring in evaluating poor prognosis out-comes in patients with hemorrhagic stroke.Results After sample size calculation,273 patients were ultimately selected as the model group for this cross-sectional study:a total of 110 hemorrhagic stroke patients had poor prognosis,163 patients had a good prognosis,and the time span was from January 2021 to September 2021.Another 81 patients with acute hemorrhagic stroke between September 2021 and January 2022 were collected as the validation group:21 patients with poor prognosis and 60 patients with acceptable prognosis were included in the hemorrhagic stroke group.The demographic characteristics of the valida-tion group were compared,and significant statistical differences were observed in the proportion of men,age,and history of dia-betes(P＜0.05),Comparison of other clinical data between groups showed significant statistical differences(P＜0.05)in GCS score,visit time,amount of bleeding and hematoma,mixed sign,cerebral hernia,and intraventricular hemorrhage.The cut-off values for the optimization score and ICH-CT score were 186 and 128,respectively.The AUC and Youden indices of the optimi-zation score were both higher than those of the ICH-CT score.The evaluation efficiency of the optimization score for adverse prognosis was better than that of the ICH-CT score(Z=2.369,P＜0.05).Conclusion A predictive model for poor prognosis in patients with hemorrhagic stroke was constructed in this study,and it was converted it into an optimized score that has strong feasibility in clinical practice.
