Construction and validation of a risk prediction model for poor rehabilitation outcome in ischemic stroke patients with vestibular symptoms
Objective To explore the influencing factors of poor rehabilitation outcome in ischemic stroke patients with vestibular symptoms,and to construct and validate a risk prediction model based on independent risk factors.Methods 108 patients with ischemic stroke with vestibular symptoms attending the hospital from August 2020 to October 2022 were retrospectively selected and divided into modeling set(n=76)and validation set(n=32)according to the ratio of 7:3,patients with different rehabilitation effects according to the modeling set were divided into the group with good rehabilitation(n=40)and the group with poor rehabilitation(n=36),and the risk warning model was constructed and validated based on independent risk factors through single-factor analysis and multifactorial Logistic analysis.The independent risk factors were derived from single-factor analysis and multifactor Logistic regression analysis,and the risk warning model was constructed based on the regression analysis method,and the corresponding column line graphs were plotted using the R language software,and the predictive efficacy was tested by using the patient's work characteristics(ROC)curves and the calibration curves.The data from the validation set were used to plot two curves for external validation of the model's early warning efficacy.Results The results of multifactorial logistic regression analysis showed that age,National Institutes of Health Stroke Scale(NIHSS)score,Vestibular Symptom Index Scale(VSI)score,degree of revascularization,comorbid hypertension,comorbid depression,and tumor necrosis factor-alpha(TNF-alpha)level were all independent risk factors for poor rehabilitation outcome in patients with ischemic stroke with vestibular symptoms,the difference was statistically significant(P<0.05),and the area under curve(AUC)of the risk prediction model constructed based on the above seven independent risk factors was 0.899,and the sensitivity and specificity corresponding to the optimal cutoffvalue of 0.654 were 0.722 and 0.900,respectively,and the results of calibration curves showed that the mean absolute error(MAE)was 0.047,and the difference between the predicted probability of the model and the actual probability of occurrence is not significant,with good prediction accuracy.The AUC of the validation set was 0.857,and the optimal cutoffvalue of 0.286 corresponded to a sensitivity and specificity of 0.929 and 0.722,respectively,and the MAE of the calibration curve was 0.070 with an average external calibration performance.Conclusion Poor rehabilitation outcome in patients with ischemic stroke with vestibular symptoms is affected by age,NIHSS score,VIS score,degree of revascularization,etc.The predictive efficacy of the risk prediction model constructed on the basis of seven independent risk factors is good,and it can be used as a reference for the allocation of resources for the rehabilitation treatment of patients with ischemic stroke with vestibular symptoms.
Ischemic strokeVestibular symptomsRehabilitationInfluential factorsEarly warning model
NETL
NSTL
万方数据
