A Study On The Reliability And Validity Of E-NIHSS In Chinese Posterior Circulation Infarct Patients

Objective: To evaluate the reliability and validity of expanded National Institutes of Health Stroke Scale(e-NIHSS)in Chinese posterior circulation infarct patients.Methods: A total of 152 consecutive patients with posterior circulation infarct hospitalized in the Affiliated Hospital of Qingdao University were recruited from February,2017 to November,2018.Each patient was tested by two raters using the National Institutes of Health Stroke Scale(NIHSS)and e-NIHSS on the first day of hospitalization,the 7th day of hospitalization and 90 days after stroke.All patients were tested again 2 hours after the first evaluation on the first day of hospitalization,and on the 90 th day of stroke,modified Rankin Scale(m RS)was assessed.Intrarater reliability was assessed by the scores of the same patient by the same rater at different times on the first day of hospitalization,and inter-rater reliability was assessed by the scores of the same patient by two different raters at the time of admission.Intrarater reliability and inter-rater reliability were measured using κ coefficient.The split-half reliability and internal consistency were assessed by the scores of one rater at the time of admission,which were evaluated using Guttman Split-half coefficient and Cronbach’s α coefficient,respectively.Concurrent validity was assessed by the correlation between the scores of e-NIHSS and NIHSS by the same rater at the time of admission,the 7th day of hospitalization and 90 days after stroke.Predictive validity was assessed by the correlation between the scores of e-NIHSS at the time of admission,the 7th day of hospitalization,90 days after stroke and the scores of NIHSS,e-NIHSS,m RS 90 days after stroke,respectively.Construct validity was assessed by the scores of one rater at the time of admission,which was explored using factor analysis.Results:(1)Intrarater reliability: the κ coefficients of all items of e-NIHSS ranged from 0.664 to 0.887(P=0.000),seven of which were above 0.8.Left motor arm received the highest κ value,whereas ataxia received the lowest κ value.(2)Inter-rater reliability: the κ coefficients of all items of e-NIHSS ranged from 0.627 to 0.964(P=0.000),ten of which were above 0.8.LOC questions received the highest κ value,whereas ataxia received the lowest κ value.(3)Split-half reliability: Guttman Split-half coefficients were 0.898,0.507 and 0.801 respectively when divided into the anterior and posterior part,divided using an even-odd approach,and ordered by the total score of each item and then divided using an even-odd approach.(4)Internal consistency reliability: Cronbach α was 0.740,which could be higher when deleting item 5,6 or 11.(5)Concurrent validity: the correlation between the total e-NIHSS score and the total NIHSS score at the same time was statistically significant(P=0.000).(6)Predictive validity: the correlation between the total e-NIHSS score at baseline,the 7th day of hospitalization,90 days after stroke and the NIHSS,e-NIHSS,m RS scores 90 days after stroke was statistically significant(P=0.000).(7)Construct validity: the KMO test(KMO value 0.742)and Bartlett’s test for sphericity(P=0.000)showed that it was suitable for factor analysis.Using principal component analysis,we extracted four factors with eigenvalues >1,which explained 79.773% of the variance in the data,and all items exhibited good loading values on one of the factors.Language,level of consciousness(LOC),LOC commands,LOC questions,extinction/inattention and visual fields loaded on factor 1,representing the functions of left and right cerebral hemispheres.Left motor arm,left motor leg,dysarthria,facial palsy,hypoglossal and glossopharyngeal palsy,and sensory loaded on factor 2,representing the motor and sensory functions of the right part of brainstem.Right motor arm and right motor leg loaded on factor 3,representing the motor function of the left part of brainstem.Ataxia,eye movements,nystagmus and Horner’s syndrome loaded on factor 4,representing cerebellar and some brainstem function.Conclusions: e-NIHSS is a new useful version of NIHSS with good reliability and validity,which can improve the sensibility of NIHSS in posterior circulation stroke,evaluate the condition of stroke patients better,have an impact on therapeutic strategy,and to some extent,predict outcomes.