Construction and validation of risk prediction model of dysarthria in patients with stroke

KANG Ying; LIANG Yanchao; LIU Yujing; CHENG Jie; YU Ying; HOU Fang; LIANG Yajing; KANG Xinli

doi:10.12102/j.issn.1009-6493.2024.14.010

Abstract：Objective:To explore the risk prediction model of dysarthria in patients with stroke,construct risk prediction model and validate their effectiveness.Methods:Convenience sampling was used to select stroke patients from 4 tertiary hospitals in Hebei province from July to November 2022.Patients in case group were 173 cases and control group were 316 cases.A general information questionnaire,oral feature questionnaire,Chinese version of Brief Oral Health Status Examination,National Institutes of Health Stroke Scale,and the Chinese version of Montreal Cognitive Assessment Scale were used to survey patients,and Lovett muscle strength grading and water swallow test were used to evaluate patients.Results:Age≥60 years old,brainstem stroke,history of stroke,poor tongue and body movement,moderate and severe neurological dysfunction,poor oral health,cognitive and swallowing dysfunction were risk factors of dysarthria in patients with stroke(P<0.05).The area under the curve of receiver operator characteristic of predictive model based on risk factors was 0.837[95%CI(0.798,0.876)],the sensitivity was 0.905,the specificity was 0.642,Brier score calculated by Bootstrap internal validation was 0.148,H-L test showed P=0.177,the decision curve showed high risk probability ranged from 0.08 to 0.10,and the C-index of the Nomogram was 0.837,P=0.743.Conclusions:Medical staff should focus on the age,stroke history,stroke location,degree of neurological dysfunction,tongue and oral health,cognitive and swallowing functions of stroke patients.The predictive model constructed in this study can be used to evaluate patients and develop targeted measures early.