Comparing the Sensitivity, Specificity, and Predictive ...

Archives of Psychiatric Nursing 30 (2016) 486–491

Contents lists available at ScienceDirect

Archives of Psychiatric Nursing journal homepage: www.elsevier.com/locate/apnu

Comparing the Sensitivity, Specificity, and Predictive Values of the Montreal Cognitive Assessment and Mini-Mental State Examination When Screening People for Mild Cognitive Impairment and Dementia in Chinese Population Jui-Chen Tsai a,b, Chia-Wei Chen b,f,1, Hsin Chu c,d, Hui-Ling Yang b, Min-Huey Chung b, Yuan-Mei Liao b, Kuei-Ru Chou a,b,e,⁎ a

Department of Nursing, Taipei Medical University—Shuang Ho Hospital, Taipei, Taiwan School of Nursing, College of Nursing, Taipei Medical University, Taipei, Taiwan Institute of Aerospace Medicine, School of Medicine, National Defense Medical Center, Taipei, Taiwan d Department of Neurology, Tri-Service General Hospital, Taipei, Taiwan e Psychiatric Research Center, Taipei Medical University Hospital, Taipei, Taiwan f Department of Nursing, Taipei Veterans General Hospital, Taipei, Taiwan b c

a b s t r a c t Background: The study compared the sensitivity, specificity, and diagnostic value of the Montreal Cognitive Assessment (MoCA) and Mini-Mental State Examination (MMSE) in screening for mild cognitive impairment (MCI) and dementia. Methods: A cross-sectional descriptive design was used, and 142 participants were screened for MCI and mild dementia by using the MoCA and MMSE. The receiver operating characteristic curves and the cutoff scores with the largest area under the curve (AUC) were determined and compared to calculate the sensitivity, specificity, and diagnostic value (positive predictive value [PPV] and negative predictive value [NPV]). Results: The optimal MoCA cutoff scores for MCI and dementia were 24 and 20, respectively. According to these scores, the sensitivities were 0.88 and 0.79, the specificities were 0.74 and 0.80, the AUCs were 0.91 and 0.87, the PPVs were 0.93 and 0.74, and the NPVs were 0.74 and 0.87, respectively. The optimal cutoff MMSE scores for MCI and dementia were 27 and 24, respectively. Hence, the sensitivities were 0.88 and 0.84, the specificities were 0.70 and 0.86, the AUCs were 0.88 and 0.89, the PPVs were 0.94 and 0.80, and the NPVs were 0.81 and 0.88, respectively. Conclusion: In the Chinese population, the MoCA is more efficient in screening for MCI than for dementia, whereas the MMSE is more efficient in screening for dementia than for MCI. The MoCA and MMSE can be used by clinical staffs for quick and accurate cognitive impairment screening, thus facilitating early and appropriate clinical intervention and treatment. © 2016 Elsevier Inc. All rights reserved.

rate and early diagnosis and proper intervention benefit both patients and family members. Although several screening instruments have been used to detect dementia, the Montreal Cognitive Assessment (MoCA) and the Mini-Mental State Examination (MMSE) are specifically widely used in the early screening for dementia in Chinese population. Dementia is most commonly diagnosed used by the MMSE and is acknowledged as the gold standard for cognitive screening. The major advantage of the MoCA is its sensitivity for both mild cognitive impairment (90% sensitivity) and mild dementia (100% sensitivity) (Nasreddine et al., 2005). Tables 1 and 2 list the findings of previous studies on the reliability, validity, sensitivity, specificity, and predictive value (positive predictive value [PPV] and negative predictive value [NPV]) of multilingual versions of the MoCA and MMSE. According to Table 1, the MMSE consists of 7 cognitive domains; orientation, registration, attention, calculation, memory, language, and visual construction. In addition to the aforementioned 7 cognitive domains, executive

The prevalence rate of dementia increases as the global population ages. According to the World Health Organization, 36 million people in 2012 had a diagnosis of dementia worldwide (prevalence rate, 4.7%). The prevalence rates in the Americas and Asia are 6.5–8.5% and 4.2–7.0%, respectively. For adults older than 65 years, the prevalence rate nearly doubles every 5 years (Alzheimer's Disease International, 2009). Dementia can cause elderly people to become disabled or institutionalized, thereby increasing burdens on caretakers. Therefore, accu-

Conflict of interest: None declared. ⁎ Corresponding Author: Kuei-Ru Chou, PhD, RN, Professor School of Nursing, College of Nursing, Taipei Medical University, Taipei, Taiwan, No.250, Wu-Hsing St., Taipei 110, Taiwan. E-mail address: [email protected] (K.-R. Chou). 1 Chia-Wei Chen contributed equally as first author of this manuscript. http://dx.doi.org/10.1016/j.apnu.2016.01.015 0883-9417/© 2016 Elsevier Inc. All rights reserved.

49

J.-C. Tsai et al. / Archives of Psychiatric Nursing 30 (2016) 486–491

487

Table 1 MoCA and MMSE for Dementia Assessment. Tool/author (year)

Items/ time

MoCA Nasreddine et al.(2005) MMSE Folstein et al.(1975) NOTE.“※”:

Method Areas assessed

Reliability

Memory Attention Orientation Calculation Language Registration Executive Conceptual Visuospatial Cronbach's Test– function thinking structure α retest

Interrater

12 items/ ※ 15 minutes

˅

˅

˅

˅

˅

0.87–0.95

11 items/ ※ 10 minutes

˅

˅

˅

˅

˅

˅

˅

˅

˅

0.72–0.87

0.86–0.96

˅

0.68–0.96

0.74–0.99 0.83

test and interview administered by clinicians or trained professionals; “˅”: areas covered; MoCA = Montreal Cognitive Assessment; MMSE = Mini-Mental State Examination.

functions and conceptual thinking domains are added in the MoCA. Both instruments demonstrated good internal consistency, test retest reliability, and inter-rater reliability. According to Table 2, the MoCA has a higher sensitivity, specificity, positive predictive value, and negative predictive value than the MMSE especially for screening people with MCI and mild AD (Nasreddine et al., 2005; Wong et al., 2009). However, because of cultural differences in Eastern and Western countries, it is critical for optimal usage to further examine the sensitivity, specificity, and psychometric properties of the two instruments across a broad range of cognitive dysfunctions in Chinese clinical settings.

MONTREAL COGNITIVE ASSESSMENT The MoCA is a psychiatric instrument developed by Nasreddine et al. (2005) for assessing mild cognitive impairment (MCI). The current version comprises eight cognitive functions: attention and concentration, executive functions, memory, language, visuoconstructional abilities, conceptual thinking, calculations, and orientation. The internal consistency based on Cronbach's α is 0.72–0.87, the test–retest reliability is 0.86–0.96, and the inter-rater reliability is 0.87–0.95 (Hu et al., 2013; Wong et al., 2009). The MoCA has been translated into several languages and demonstrates high sensitivity and specificity in assessing MCI (Tables 1 and 2). For assessing MCI, the cutoff score is 26, the sensitivity of the original MoCA version is 0.90, the PPV is 0.89, and the NPV is 0.91 (Nasreddine et al., 2005). For the multilingual MoCA versions, the cutoff score is

22–27, the sensitivity is 0.89–0.92, and the specificity is 0.78–0.85 (Hu et al., 2013; Lee et al., 2008; Tsai et al., 2012). For assessing dementia, the cutoff score is 21–26, the sensitivity is 0.92–1.00, the specificity is 0.87–0.96 (Hu et al., 2013; Nasreddine et al., 2005; Tsai et al., 2012), the PPV is 0.88–0.89, and the NPV is 1.00 (Nasreddine et al., 2005; Tsai et al., 2012) of the MoCA. Tsai et al. (2012) translated the MoCA into Chinese, and the translated version (MoCA-T) demonstrates high reliability and validity. For assessing MCI, the optimal cutoff score of the MoCA-T is 23–24, area under the curve (AUC) is 0.91 (95% confidence interval [CI]: 0.86–1.00), sensitivity is 0.92, specificity is 0.78, PPV is 0.88, and NPV is 0.94. The MoCA-T is more efficient than the MMSE for assessing MCI. For assessing dementia, the MoCA-T cutoff score is 21–22, AUC is 0.99 (95% CI: 0.98–1.00), sensitivity is 0.98, specificity is 0.95, PPV is 0.88, and NPV is 1.00.

MINI-MENTAL STATE EXAMINATION The MMSE, the most commonly used instrument for assessing cognitive function, includes orientation, memory, calculation and attention, registration, language, and visuospatial function (Folstein, Folstein, & McHugh, 1975). For assessing MCI, the optimal MMSE cutoff scores ranged 23–28. Some differences have been observed in the sensitivity, specificity, and predictive value. For assessing dementia, the MMSE cutoff score is 24–26, sensitivity is 0.89–0.95, and specificity is 0.87–0.98 (Blesa et al., 2001; Dong et al., 2012; Tsai et al., 2012).

Table 2 Sensitivity and Specificity Analysis of the MoCA and MMSE. n MoCA Hu et al. (2013) bChinese versionN Tsai et al. (2012) bTaiwanese versionN Lee et al. (2008) bKorean versionN Nasreddine et al. (2005)

MMSE Dong et al. (2012)

Tsai et al. (2012) bTaiwanese versionN Blesa et al. (2001)

Normal MCI Dementia Normal MCI Dementia Normal MCI Dementia Normal MCI Mild AD

146 84 72 38 71 98 115 37 44 90 94 93

Normal MCI Dementia Normal MCI Dementia Normal MCI Dementia

33 61 136 38 71 98 253 86 111

Cut-off point

Sensitivity

Specificity

AUC

PPV

NPV

26/27 25/26

0.92 0.92

0.85 0.96

– –

– –

– –

23/24 21/22

0.92 0.98

0.78 0.95

0.91 0.99

0.88 0.88

0.94 1.00

22/23

0.89

0.84

–

–

–

26 26

0.90 1.00

– 0.87

– –

0.89 0.89

0.91 1.00

23/24

0.72

0.83

0.84

0.72

0.83

27/28 25/26

– 0.95

0.63 0.98

– –

– –

– –

24/25

0.89

0.87

–

–

–

MoCA = the Montreal Cognitive Assessment; MMSE = the Mini-Mental State examination; MCI = mild cognitive impairment; AD = Alzheimer's disease; AUC = area under curve; PPV = positive predictive value; NPV = negative predictive value.

50

488


The MMSE has been translated into several languages and is widely applied clinically. Tombaugh and McIntyre (1992) reported that the MMSE demonstrated high sensitivity for assessing moderate-to-severe cognitive impairment, whereas it exhibited significantly low sensitivity for assessing MCI. However, the assessment can be affected by age, education, language, and cultural background and can easily result in the ceiling effect. Moreover, age, quality of education, and the differences in the MMSE scores between African Americans and Caucasian Americans can be attributed to illness severity (Pedraza et al., 2012). Overall, the MoCA emphasizes more on the frontal executive function and attention tasks than does the more commonly used MMSE (Wong et al., 2013). It covers critical cognitive areas and is therefore a viable replacement for the MMSE for assessing dementia. Hoops et al. (2009) reported that the MoCA was more efficient than the MMSE in screening people with MCI and had no ceiling effect in younger, welleducated individuals (Table 2). Although the two assessment instruments have been tested in the Chinese population, the sensitivity, specificity, and predictive value of these instruments have been rarely compared, thereby limiting their utility for additional cognitive assessment research, particularly in Chinese people with MCI or dementia. Therefore, the purpose of this study is to examine the sensitivity, specificity, and predictive value of the MoCA and MMSE in screening for MCI and dementia in Chinese population.

possibility of cognitive impairment; however, currently, respondents with b 12 years of education earn 1 extra point (Nasreddine et al., 2005). The cutoff score for the Chinese version is 23 (Nasreddine et al., 2005; Tsai et al., 2012). The MoCA is easy to use and demonstrates high reliability and validity. The internal consistency based on Cronbach's α is 0.83, and the test–retest reliability is 0.92.

DESIGN AND METHOD

Sample size was calculated according to the Hanley and McNeil (1983) by using the 95% CI, estimated correlation coefficient, and the AUC. The estimated correlation coefficient between the MoCA and the MMSE was 0.8. Accordingly, at least 140 participants were required for achieving a ± 10% 95% CI. A second sample size was calculated using a Web-based calculator (Anaesthetist.com, 2011). Considering a similar AUC of 0.8 in both the MoCA and MMSE, a sample size of 140 participants was considered adequate for this study.

Participants This study was approved by the Institutional Review Board for the Protection of Human Subjects (reference number: VGHIRB No.201205-004A and TSGHIRB No.1-101-5-016). The participants were recruited from the geriatric psychiatry, neurology, and geriatric departments of two medical centers in Taiwan. The inclusion criteria were (1) age N 65 years; (2) deteriorated memory and reduced daily activities, or complaints from family members about such problems; and (3) ability to communicate normally, understand the questionnaire content, and follow instructions. After enrollment, the participants' clinical diagnoses regarding MCI and dementia were based on the National Institute on Aging, the Alzheimer Association, and the Diagnostic and Statistical Manual, Fourth Edition, Text Revision (DSM-IV-TR) guidelines, of which DSMIV-TR was the reference standard; the MoCA and MMSE were administered to each participant by a well-trained, on-site neurologist to obtain a comprehensive diagnostic validation. The clinical criteria for MCI, as proposed by the National Institute on Aging and the Alzheimer Association, comprise the following: (1) concern regarding a change in cognition, wherein the participant reported a change compared with the previous level; (2) one or more impaired cognitive domains (memory, executive function, attention, language, and visuospatial function on the MMSE); (3) preserved independence in functional abilities, in which the participant reported independent functioning in all instrumental daily living activities; and (4) absence of dementia (score of N 24/30 on the MMSE and score of b 1 on the Clinical Dementia Rating Scale) (Albert et al., 2011). The clinical criteria proposed by the DSMIV-TR comprise impaired memory and deficits in other cognitive domains (aphasia, apraxia, agnosia, or executive dysfunction) (American Psychiatric Association, 2000). Instruments Montreal Cognitive Assessment The MoCA is used for assessing cognitive functions, because it covers various critical cognitive areas and must be administered by a physician or trained professional. The instrument has 12 items and takes approximately 15 minutes to complete, and the score for each item ranges from 0 to 3. The highest attainable score is 30, and a low score indicates poor cognitive function. In the original version, a score of b 26 indicates the

Chinese Version of the Mini-Mental State Examination The MMSE is used to measure cognitive function and must be administered by a physician or other trained professional. The instrument, which has 11 items, takes 5–10 minutes to complete, and the score for each item is 1 point (Folstein et al., 1975). The total possible score is 30, and for a respondent with more than a high school education, a score of ≤23 indicates possible cognitive impairment. For respondents with less than high school education, a score of ≤18 indicates possible cognitive impairment. A low score indicates poor cognitive function (Guo et al., 1988). Research has shown that the MMSE is easy to use and demonstrates high reliability and validity. The internal consistency based on Cronbach's α is 0.68–0.96 (Foreman, 1987; Kay et al., 1985), the test–retest reliability is 0.74–0.99 (Dick et al., 1984; Folstein et al., 1975), and the inter-rater reliability is 0.83 (Folstein et al., 1975). Sample Size Calculation

Statistical Analyses Statistical analyses were performed using SPSS 21.0. The MoCA and MMSE scores were plotted as receiver operating characteristic (ROC) curves, and the cutoff scores with the largest AUC were determined for both instruments. The corresponding sensitivity, specificity, and predictive value were subsequently calculated and compared. The ROC curve was analyzed and plotted using STAT 9.1 (StataCorp LP, College Station, TX, USA), and a high cutoff score indicated low sensitivity and high specificity. The PPV was the percentage of people who tested positive for the disease (true positive or true positive + false positive), and the NPV was the percentage of people who tested negative for the disease (true negative or true negative + false negative). The PPV and NPV were used to determine the accuracy of the screening instruments. The chisquared test was performed using the PROC LOGISTIC procedure of the SAS v. 9.3 (SAS Institute Inc., Cary, NC, USA) for comparing the ROC curves between MoCA and MMSE in diagnosing MCI or dementia. Ethical Considerations The study protocol was approved by the Institutional Review Board of Tri-Service General Hospital (approval number: 1-101-05-016) and Taipei Veterans General Hospital (approval number: 2012-05-004A). All participants were informed of the study purpose, process, potential risks, and the right to withdraw from the study at any time. Informed consent was obtained from all participants. RESULTS We assessed 142 participants, of whom 26 demonstrated normal cognitive function, 59 had MCI, and 57 had dementia. The average age was 76.2 (SD = 8.5) years, and 73 men (51.4%) and 69 women (49.6%) participated in the study. Regarding the educational level,

51


489

Fig. 1. ROC curves of the MoCA and MMSE (normal vs. MCI).

Fig. 2. ROC curves for the MoCA and MMSE (normal vs. dementia).

35.9, 16.2, 19.0, and 28.9% of the participants had received elementary school, junior high school, senior high school, and university levels of education, respectively. In total, 62 (43.7%), 50 (35.2%), and 30 (21.1%) participants were recruited from the neurology, geriatric psychiatry, and geriatric departments, respectively. The sensitivity, specificity, PPV, and NPV for both instruments in screening MCI and dementia were addressed as below.

difference between the ROC curves for these two instruments in MCI diagnosis (chi-square: 2.0291; P = 0.1543). MoCA and MMSE for Screening Dementia The patient diagnosis was used as the reference standard, and the MoCA and MMSE efficiencies in dementia screening were compared (Fig. 2 and Table 4). The AUC for MMSE was 0.89 (95% CI: 0.83–0.95). According to the ROC curve, the cutoff score was 24, and the sensitivity, specificity, PPV, and NPV were 84, 86, 80, and 88%, respectively. The AUC for MoCA was 0.87 (95% CI: 0.81–0.94). According to the ROC curve, the cutoff score was 20, and the sensitivity, specificity, PPV, and NPV were 79, 80, 74, and 87%, respectively. When the MoCA and MMSE cutoff scores were set at 20 and 24, respectively, the sensitivity, specificity, and PPV of the MMSE were higher than those of the MoCA in dementia screening. Similarly, the chi-squared test results showed no significant difference between the ROC curves of the two instruments in dementia diagnosis (chi-square: 0.1829; P = 0.6689).

MoCA and MMSE for Screening for Mild Cognitive Impairment The patient diagnosis was used as the reference standard, and the MoCA and MMSE efficiencies for assessing MCI were compared (Fig. 1 and Table 3). The AUC for MMSE was 0.88 (95% CI: 0.80–0.95). According to the ROC curve, the cutoff score was 27, and the sensitivity, specificity, PPV, and NPV were 88, 70, 94, and 81%, respectively. The AUC for MoCA was 0.91 (95% CI: 0.85–0.96). According to the ROC curve, the cutoff score was 24, and the sensitivity, specificity, PPV, and NPV were 88, 74, 93, and 74%, respectively. Table 3 shows the MoCA and MMSE ROC curves for assessing MCI. The AUC of the MoCA was similar to that of the MMSE (cutoff score: 24 [0.91] and 27 [0.88], respectively). The sensitivity and PPV of the MoCA were similar to those of the MMSE. The specificity of the MoCA was 4% higher than that of the MMSE. The NPV of the MMSE was 81%, whereas that of the MoCA was 74%. However, the chi-squared test results showed no significant

DISCUSSION The MoCA and MMSE demonstrate appropriate sensitivity and specificity for screening patients with cognitive impairment. Screening for Mild Cognitive Impairment

Table 3 ROC Curve Analysis of the MoCA and MMSE (Normal vs. MCI).

MoCA

MMSE

Cut-off point

Sensitivity

Specificity

AUC

PPV

NPV

22 23 24a 25 26 25 26 27a 28 29

0.72 0.79 0.88a 0.94 0.98 0.58 0.72 0.88a 0.98 0.99

0.96 0.74 0.74a 0.57 0.44 0.96 0.81 0.70a 0.46 0.12

0.84 0.77 0.81a 0.75 0.71 0.77 0.77 0.79a 0.55 0.50

1 0.99 0.93a 0.93 0.90 0.98 0.98 0.94a 0.92 0.89

1 0.96 0.74a 0.74 0.57 0.96 0.96 0.81a 0.68 0.42

In this study, the optimal cutoff score, sensitivity, and specificity of the Chinese MMSE were 27, 88, 70%, respectively. These scores were similar to those of Tsai et al. (2012), in which the optimal cutoff score and specificity were 27 and 63%, respectively. Therefore, a cutoff score of 27 can be used in screening for MCI. The optimal cutoff score and sensitivity of the MoCA in MCI screening were 26 and 90%, respectively, according to Nasreddine et al. (2005), whereas the optimal cutoff score, sensitivity, and specificity were 23 or 24, 92, and 78%, respectively, according to Tsai et al. (2012). In the present study, the optimal cutoff score was 24, which was consistent with that of Tsai et al. (2012). Therefore, the optimal cutoff score for the MoCA is 24; thus, the MOCA can be effectively used for assessing MCI in the Chinese population. In the present study, the Chinese versions of both the MoCA and MMSE had appropriate sensitivities in screening for MCI; however, the MMSE had lower specificity than that of the MoCA. Both the MoCA

NOTE. ROC = the receiver operating characteristic; MCI = mild cognitive impairment; MoCA = Montreal Cognitive Assessment; MMSE = Mini-Mental State Examination; AUC = area under curve; NPV = negative predictive value; PPV = positive predictive value. a Optimal cut-off point and its sensitivity, specificity, AUC, PPV and NPV.

52

490


Table 4 ROC Curve Analysis of the MoCA and MMSE (Normal vs. Dementia). Cut-off point

Sensitivity

Specificity

AUC

PPV

NPV

MoCA

18 19 20a 21 22

0.58 0.67 0.79a 0.84 0.93

0.93 0.87 0.80a 0.71 0.61

0.76 0.77 0.80a 0.77 0.77

0.87 0.83 0.74a 0.69 0.62

0.96 0.93 0.87a 0.80 0.71

MMSE

22 23 24a 25 26

0.61 0.72 0.84a 0.88 0.93

0.91 0.88 0.86a 0.79 0.58

0.76 0.80 0.85a 0.83 0.75

0.89 0.81 0.80a 0.80 0.74

0.96 0.90 0.88a 0.85 0.78

the sensitivity, specificity, PPV, and NPV of the MMSE were 84, 86, 0.80, and 0.88%, respectively, which were higher than those of the MoCA. Overall, both the MMSE and MoCA were easy to use, and the MoCA was more efficient for MCI screening than for dementia screening (Tsai et al., 2012). Consistent with previous study, the Chinese MMSE was more efficient for dementia screening than for MCI screening because of the high sensitivity and PPV (Xu et al., 2003). CONCLUSIONS The MoCA and MMSE demonstrate appropriate sensitivity and specificity for screening patients with cognitive impairment. The MoCA is more sensitive in screening for MMSE than for MCI, whereas the MMSE is more sensitive than the MoCA in screening patients with dementia. These instruments can be used by clinical staffs for quick and accurate cognitive impairment screening, thus facilitating early and appropriate clinical intervention.

ROC = the receiver operating characteristic; MoCA = Montreal Cognitive Assessment; MMSE = Mini-Mental State Examination; AUC = area under curve; PPV = positive predictive value; NPV = negative predictive value. a Optimal cut-off point, sensitivity, specificity, AUC, PPV and NPV.

NOTE.

and MMSE cover multiple cognitive functions. Compared with the MoCA, the MMSE comprises visuospatial structure, attention and calculation, language, orientation, memory, conceptual thinking, and execution function for assessing cognitive function. As dementia progresses, conceptual thinking, attention, language, executive function, and orientation of the patient may deteriorate, particularly at an early stage of the illness. The results emphasize the importance of assessing execution function and conceptual thinking in differentiating normal cognition from MCI. The results are consistent with those of previous studies, which have concluded that early and mild symptoms can be effectively identified by evaluating the cognitive functions (including orientation, attention, conceptual thinking, executive functions, and language) in people with MCI (Kavcic & Duffy, 2003; O'Brien et al., 2001; Quental, Brucki, & Bueno, 2009). Comparing the MoCA and MMSE efficiencies in screening for MCI revealed that the MoCA was more suitable for MCI screening than for dementia screening in a Chinese population (Nasreddine et al., 2005; Tsai et al., 2012).

Funding: This study was supported by Taipei Medical University—Shuang Ho Hospital (104TMU-SHH-16). References Albert, M. S., DeKosky, S. T., Dickson, D., Dubois, B., Feldman, H. H., Fox, N. C., et al. (2011). The diagnosis of mild cognitive impairment due to Alzheimer's disease: Recommendations from the National Institute on Aging–Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimer's & Dementia, 7(3), 270–279. Alzheimer's Disease International (2009). World Alzheimer Report 2009 http://www.alz. co.uk/research/files/WorldAlzheimerReport.pdf (Retrieved January 26, 2014, from) American Psychiatric Association (2000). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author (Text Revision). Anaesthetist.com (2011). The magnificent ROC. Retrieved June 3, 2015, from http:// www.anaesthetist.com/mnm/stats/roc/Findex.htm Blesa, R., Pujol, M., Aguilar, M., Santacruz, P., Bertran-Serra, I., Hernández, G., et al. (2001). Clinical validity of the ‘mini-mental state’ for Spanish speaking communities. Neuropsychologia, 39(11), 1150–1157. Dick, J., Guiloff, R., Stewart, A., Blackstock, J., Bielawska, C., Paul, E., et al. (1984). Minimental state examination in neurological patients. Journal of Neurology, Neurosurgery & Psychiatry, 47(5), 496–499. Dong, Y., Lee, W. Y., Basri, N. A., Collinson, S. L., Merchant, R. A., Venketasubramanian, N., et al. (2012). The Montreal Cognitive Assessment is superior to the Mini-Mental State Examination in detecting patients at higher risk of dementia. International Psychogeriatrics, 24(11), 1749–1755. Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975). “Mini-mental state”: A practical method for grading the cognitive state of patients for the clinician. Journal of psychiatric research, 12(3), 189–198. Foreman, M. D. (1987). Reliability and validity of mental status questionnaires in elderly hospitalized patients. Nursing Research, 36(4), 216–220. Guo, N., Liu, H., Wong, P., Liao, K., Yan, S., Lin, K., et al. (1988). Chinese version and norms of the Mini-Mental State Examination. Journal of Rehabilitation Medicine Association, 16, 52–59. Hanley, J. A., & McNeil, B. J. (1983). A method of comparing the areas under receiver operating characteristic curves derived from the same cases. Radiology, 148(3), 839–843. Hoops, S., Nazem, S., Siderowf, A., Duda, J., Xie, S., Stern, M., et al. (2009). Validity of the MoCA and MMSE in the detection of MCI and dementia in Parkinson disease. Neurology, 73(21), 1738–1745. Hu, J. B., Zhou, W. H., Hu, S. H., Huang, M. L., Wei, N., Qi, H. L., et al. (2013). Cross-cultural difference and validation of the Chinese version of Montreal Cognitive Assessment in older adults residing in Eastern China: Preliminary findings. Archives of gerontology and geriatrics, 56(1), 38–43. Kavcic, V., & Duffy, C. J. (2003). Attentional dynamics and visual perception: Mechanisms of spatial disorientation in Alzheimer's disease. Brain, 126(5), 1173–1181. Kay, D., Henderson, A., Scott, R., Wilson, J., Rickwood, D., & Grayson, D. (1985). Dementia and depression among the elderly living in the Hobart community: The effect of the diagnostic criteria on the prevalence rates. Psychological Medicine, 15(4), 771–788. Lee, J. Y., Lee, D. W., Cho, S. J., Na, D. L., Jeon, H. J., Kim, S. K., et al. (2008). Brief screening for mild cognitive impairment in elderly outpatient clinic: Validation of the Korean version of the Montreal Cognitive Assessment. Journal of Geriatric Psychiatry and Neurology, 21(2), 104–110. Nasreddine, Z. S., Phillips, N. A., Bédirian, V., Charbonneau, S., Whitehead, V., Collin, I., et al. (2005). The Montreal Cognitive Assessment, MoCA: A brief screening tool for mild cognitive impairment. Journal of the American Geriatrics Society, 53(4), 695–699. O'Brien, H. L., Tetewsky, S. J., Avery, L. M., Cushman, L. A., Makous, W., & Duffy, C. J. (2001). Visual mechanisms of spatial disorientation in Alzheimer's disease. Cerebral Cortex, 11(11), 1083–1092. Pedraza, O., Clark, J. H., O'Bryant, S. E., Smith, G. E., Ivnik, R. J., Graff‐Radford, N. R., et al. (2012). Diagnostic validity of age and education corrections for the Mini‐Mental State Examination in older African Americans. Journal of the American Geriatrics Society, 60(2), 328–331.

Screening for Dementia Although a cutoff score analysis in the original version is lacking, the MMSE cutoff scores set by researchers in other countries have ranged from 24 to 26. Sensitivity and specificity were in the range of 0.72– 0.95 and 0.83–0.98, respectively. The optimal cutoff score applied in the present study was consistent with that in Tsai et al. (2012), and the sensitivity and specificity were 95 and 98%, respectively. Therefore, the optimal cutoff score for the MMSE is 24; thus, dementia can be effectively detected in the Chinese population according to this score. Nasreddine et al. (2005) reported that when the MoCA with a cutoff score of 26 was used for assessing dementia, the sensitivity and specificity were 100 and 87%, respectively. Researchers from other countries set their cutoff scores between 21 and 26 for assessing dementia (Table 2). The optimal cutoff score in this study was 20 or 21, which was similar to that in the study by Tsai et al. (2012). The sensitivity and specificity obtained in these studies were considered more appropriate for the Chinese population. The ROC curve analysis revealed that the MMSE demonstrated appropriate sensitivity and PPV but low specificity in assessment for MCI. The sensitivity, specificity, and predictive value were all appropriate in assessment for dementia. Therefore, the MMSE is more suitable for assessing dementia than for assessing MCI. The MoCA sensitivity and PPV were higher for assessing MCI than for assessing dementia. This finding is consistent with that of previous studies, indicating that the MoCA was a more effective screening instrument than the MMSE for assessing MCI. When patients were screened for MCI, the MMSE sensitivity (88%) was similar to that of the MoCA (88%) for a cutoff score of 25; however, both instruments demonstrated low specificity. In dementia screening,

53


491

cognitive domain deficit following aneurysmal subarachnoid haemorrhage. PLoS ONE, 8(4), e59946. Wong, A., Xiong, Y. Y., Kwan, P. W., Chan, A. Y., Lam, W. W., Wang, K., et al. (2009). The validity, reliability and clinical utility of the Hong Kong Montreal Cognitive Assessment (HK-MoCA) in patients with cerebral small vessel disease. Dementia and geriatric cognitive disorders, 28(1), 81–87. Xu, G., Meyer, J. S., Huang, Y., Du, F., Chowdhury, M., & Quach, M. (2003). Adapting Mini‐Mental State Examination for dementia screening among illiterate or minimally educated elderly Chinese. International journal of geriatric psychiatry, 18(7), 609–616.

Quental, N. B. M., Brucki, S. M. D., & Bueno, O. F. A. (2009). Visuospatial function in early alzheimer's disease: Preliminary study. Dementia e Neuropsychologia, 3(3), 234–240. Tombaugh, T. N., & McIntyre, N. J. (1992). The mini-mental state examination: A comprehensive review. Journal of the American Geriatrics Society, 40(9), 922–935. Tsai, C. F., Lee, W. J., Wang, S. J., Shia, B. C., Nasreddine, Z., & Fuh, J. L. (2012). Psychometrics of the Montreal Cognitive Assessment (MoCA) and its subscales: Validation of the Taiwanese version of the MoCA and an item response theory analysis. International Psychogeriatrics, 24(4), 651–658. Wong, G. K. C., Lam, S. W., Wong, A., Ngai, K., Poon, W. S., & Mok, V. (2013). Comparison of Montreal Cognitive Assessment and Mini-Mental State Examination in evaluating

54