Application Of Cognitive Impairment Rapid Screening System For Screening Mild Cognitive Impairment Among The Elderly In Community

Objective:This study aimed to optimize Cognitive Impairment Rapid Screening System(CIRSS)through the diagnostic study design,establish a diagnosis model and develop a score table,and examine screening effect and community applicability of CIRSS on Mild Cognitive Impairment(MCI)among the elderly in community,so as to provide scientific evidence for CIRSS to effectively carry out the rapid screening of MCI among the elderly in community.Methods:In Fuzhou City,954 subjects were recruited,including 638 subjects in cognition normal(NC)group and 316 subjects in MCI group.The basic demographic information included: age,gender,height,weight,education level,types of combined chronic diseases and intelligent products proficiency,etc.The Cognitive assessment included: CIRSS assessment and neuropsychological assessments(Ascertain Dementia-8,Montreal cognitive assessment of Fuzhou version,Lawton Instrumental Activities of Daily Living Scale and Geriatric Depression Scale-15).According to the diagnostic criteria of MCI,neurologists diagnosed the elderly in combination with above neuropsychological assessments.The CIRSS diagnostic model was developed and validated using internal validation methods.Through SPSS 25.0 statistical software,70% of the total sample size was randomly selected as the development set,and the remaining 30% was used as the validation set.In the development set,the differences of each variable in CIRSS between NC and MCI groups were compared.Binary logistic regression model was used to construct the diagnostic model.ROC curve was used to evaluate diagnostic performance of the model.Hosmer-lemeshow test was used to evaluate calibration ability of the model.Variables with P value less than 0.1were selected to establish the CIRSS score table.The CIRSS score table was applied to the validation set.ROC curve was used to evaluate the CIRSS score table,and the best cut-off value was explored to examine the screening effect of CIRSS on MCI among the elderly in community.Results:In the development set,there was no statistical differences in age,gender,BMI,smoking history,drinking history and Lawton-IADL scores between NC and MCI groups(P>0.05);there were statistical differences in types of chronic diseases,intelligent products proficiency,Mo CA,AD8 and GDS-15 scores between NC and MCI groups(P<0.05).In the validation set,there was no statistical differences in age,gender,BMI,smoking history,types of chronic diseases,drinking history and Lawton-IADL scores between NC and MCI groups(P>0.05);there were statistical differences in intelligent products proficiency,Mo CA,AD8 and GDS-15 scores between NC and MCI groups(P<0.05).There was no statistically significant difference between NC and MCI groups of subjects for the satisfaction with CIRSS operation in the development and validation sets(P>0.05).All subjects gave "very satisfaction" or "satisfaction" to CIRSS operation accounting for 91.9%.The median time to complete CIRSS was 396.0 seconds in NC group and 471.0 seconds in MCI group.Computer version of GDS-15 was significantly positively correlated with the consistency analysis of paper GDS-15 score(r=0.705,P<0.001).There were significant statistical differences in cube copying test,clock drawing test,trail making test,memory pattern test and simple reaction time test between NC and MCI groups(P<0.05).Two diagnostic models were constructed.Considering CIRSS’s binary classification and time variables,result showed that diagnostic model 1 of the AUC value = 0.912,and the Hosmer-Lemeshow test P value = 0.005;Diagnosis model 2 introducing education level,intelligent products proficiency and GDS-15 on basis of diagnosis model 1,showed that it’s AUC value=0.917,Hosmer-Lemeshow test P value=0.101.Finally,meaningful variables in diagnosis model 2 were selected to develop the CIRSS score table.The variables with statistical differences in diagnosis model 2 included cube copying test(P<0.001),minute hand of drawing clock test(P<0.001),the second connection of trail making test(P=0.066),memory pattern test’s fruit(P=0.011),simple reaction time test(P=0.076),trail making test’s time(P=0.088)and memory pattern test’s selection time(P=0.013).The β coefficients of above variables were converted into integers to develop a score table,and applied it to the validation set for examination.The results showed that AUC value of the CIRSS score table was 0.878(95% CI: 0.836,0.921;P<0.001),and the best cutoff value was 6 points.The sensitivity,specificity,Youden index,positive predictive value,negative predictive value,false positive rate and false negative rate were 83.5%,82.6%,0.661,69.1%,91.5%,17.4% and 16.5%,respectively.Conclusion:There is good screening performance for CIRSS to screening MCI among the elderly in community.The time efficiency and operation design of CIRSS can meet the requirements of community cognitive screening.