医学
接收机工作特性
多元统计
多元分析
预测值
冲程(发动机)
物理疗法
巴氏指数
内科学
日常生活活动
统计
数学
机械工程
工程类
作者
Lili Chen,Wenting Wang,Sai Zhang,Huimiao Liu,Xiaoyang Yuan,Xu Yang,Ping Gu
标识
DOI:10.1016/j.jocn.2021.05.049
摘要
Objective This study aimed to investigate the value of Barthel, PLAN, and NIHSS scores for predicting death in the 5-year follow-up after patients with AIS are discharged and find a simple and convenient predictive scale. Methods Data were prospectively collected from 678 patients with AIS. Patients’ death after 5 years of follow-up was considered the final event. The predictors of death were examined through single-factor and multivariate analysis. The receiver operating characteristic curve (ROC) of the patients’ Barthel, PLAN, and NIHSS scores was drawn, and the area under the curve (AUC) was calculated. Differences in the predictive power of the three scales were compared using MedCalc. The goodness of fit between predictive and actual models was evaluated with the Hosmer–Lemeshow method. Results Multivariate analysis suggested that the BI score was an independent predictor of death from AIS in the 5-year follow-up. The Barthel, PLAN, and NIHSS scale scores predicted the 5-year mortality AUC values of AIS were 0.687 [95% CI, (0.649–0.722)], 0.621 [95% CI, (0.583–0.659)], 0.637 [95% CI, (0.599–0.674)], the Hosmer–Lemeshow test revealed P > 0.05, indicating that the three models had a good fit. In pairwise comparison, the AUC value of the BI score was significantly greater than that of the NIHSS scores (Pc = 0.0189). BI and PLAN scores were very close to statistical significance (Pc = 0.0513). However, PLAN and NIHSS scores had no significant differences (Pc = 1.7493). Conclusion Simple BI scores had a high predictive value for the death of Chinese patients with AIS within 5 years.
科研通智能强力驱动
Strongly Powered by AbleSci AI