Ïã¸ÛÁùºÏ²Ê¼´Ê±¿ª½±

Cut-off Scores

MCI and Dementia (Goldstein et al, 2014; N=81 African Americans; Normal Cognition (NC) n=38, mean age 65.8 (7.7), mean 13.4 (3.1) years of education 13.4; Mild Cognitive Impairment (MCI) n=38, mean age 71.9 (8.9), mean 10.9 (2.4) years of education; Dementia n=27, mean age 70.5 (11.4), mean 11.7 (2.2) years of education)

• When using the recommended MoCA cutoff score for impairment (¡Ü 26), the MoCA showed excellent sensitivity (100%) for identifying MCI or dementia. The cutoff score had low specificity (31%).
• The cutoff score of ¡Ü 24 points on the MoCA showed good sensitivity for detecting MCI (95%); specificity improved slightly (63%).
• A cutoff score of ¡Ü 22 points showed good sensitivity (96%) and specificity (88%).

Cut-off Scores:

Alzheimer¡¯s Disease and Mild Cognitive Impairment (Lam et al, 2013; N=107; mean age 72 (10.3); mean 14.5 (3.8) years of education 14.5; Alzheimer¡¯s Disease (AD) n=75; Mild Cognitive Impairment (MCI) n=32; mean MMSE score 24.9 (2.7); mean DRS score 124 (11.2))

• Based on the results, the recommended Cut-off subscores for the MoCA included:
  • Memory: 7 or 8
  • Visuospatial: 3
  • Language: 4
  • Attention: 3
  • Executive: 4
•  The memory (0.86, P<0.001), visuospatial ( 0.79, P<0.001), and executive (0.79, P<0.001) MoCA subscores showed good accuracy for identifying impairment. The language (0.70, P=0.001) and attention (0.72, P=0.001) MoCA subscores showed fair accuracy for impairment.

 A cutoff score of 7 for the memory domain shows good specificity (0.93), but lower sensitivity (0.59). A cutoff score of 8 shows lower specificity (0.43), but good sensitivity (0.91). Adjusting the cut-off by just 1 affects the sensitivity and specificity of the measure. The ideal cut-off will depend on the purpose of using the MoCA

Cognitive Impairment: 

(Nasreddine et al, 2005) 

• A score of 26 or above is considered normal

• For individuals with 12 years or fewer of formal education, one point is added to the score as a correction

Cognitive Impairment: 

(Smith et al, 2007; From a population in a memory clinic: 32 subjects diagnosed with dementia; 23 subjects with mild cognitive impairment; 12 comparison subjects; mean age 73.6 (10) years; mean MMSE score at baseline 27.4 (1.6);mean MoCA score 22.3 (3.6)) 

• The MoCA had better sensitivity (100%) identifying subjects with MCI who were diagnosed with dementia at a six month followup, than the MMSE with sensitivity of 25%. 

Cognitive Impairment: 

(Nasreddine et al, 2005 and mocatest.org website)

Older adults with MCI and AD:

(Nasreddine et al, 2005; n = 94 with MCI, mean age 75.2 (6.3) years; mean 12.28 (4.3) years of education; n = 93 patients with AD, mean age = 76.7 (8.8) years; mean 10.03 (3.8) years of education; n = 90 healthy controls; mean age 72.8 (7.0) years; mean 13.3 (3.4) years of education)

• Excellent test retest reliability (r = 0.92) with subgroup of 26 patients with cognitive impairment tested on average 35.0 (17.6) days apart

• Mean change in scores =  0.9 + 2.5 points

Older adults with MCI and AD:

(Nasreddine et al, 2005)

•  Excellent internal consistency (a = 0.83)

Concurrent Validity:

Alzheimer¡¯s Disease and Mild Cognitive Impairment (Raolf et al, 2013)

• Cut-off scores on the MMSE can be reliably compared with scores on the MoCA.
• Researchers used equipercentile equating to find comparable scores between the MoCA and the MMSE, resulting in conversion scores.
• Lower MoCA scores were associated with higher MMSE scores. For example, 20-21 on the MoCA is equivalent to 26 on the MMSE.

Concurrent Validity:

Alzheimer¡¯s Disease and Mild Cognitive Impairment (Lam et al, 2013)

• Excellent correlations between the MoCA total scores and the Dementia Rating Scale (correlation coefficient (r)=0.77, P<0.001).
• Excellent correlations between the MoCA and the memory criterion measures (r=0.66, P<0.001).
• There was a stronger association between the Dementia Rating Scale and the MoCA in comparison to the association between the Dementia Rating Scale and the Mini-Mental State Examination, which showed adequate correlations (r=0.57, P<0.001).
• Results showed a moderate to strong correlations between the MoCA subscores and analogous domain-specific criterion measures:
  • Excellent correlations with the memory domain ( r=0.73 , P <.001).
  • Adequate correlations with the visuospatial domain (r=0.56, P<.001).
  • Adequate correlations with the language domain (r=0.46, P<.001).
  • Adequate correlations with the attention domain (r=0.51, P<.001).
  • Excellent correlations with the executive domain (r=0.60, P<.001).

Convergent Validity:

Alzheimer¡¯s Disease and Mild Cognitive Impairment (Lam et al, 2013)

• Excellent convergent validity between the English MoCA and MMSE (r=0.66, P<0.001)

Older adults with MCI and AD:

(Nasreddine et al, 2005) 

• Total scores and majority of items differentiated between known groups of healthy controls, individuals with MCI and AD 
• All items differentiated between at least two of the groups

• Was developed based on clinical intuition of first author and clinical testing over a five year period prior to validation study in 2005 (Nasreddine et al, 2005)

?Parkinson's Disease:

(Dalyrimple-Alford et al, 2010; N = 114 PD and 47 controls; median duration PD motor symptoms 12.5, (1-30) years. 3 groups identified as PD-N (normal cognition, n = 72, disease duration 4.6 (3.9), HY stage 1.9 (0.9)], PD-MCI (mild cognitive impairment, n = 21, disease duration 7.3 (5.2), Mean HY 2.6 (0.9)]and PD-D [dementia, n = 21, disease duration 12.6 (8.1); H&Y stage mean = 3.4 (0.8)]. 

• MoCA screening cutoff for PD-MCI < 26/30 (SN 90%, SP 75%; AUC 90%; 95% CI 82-95%, NPV = 92%) 

• MoCA best at differentiating PD-MCI
  • MoCA vs SCOPA-Cog (AUC difference of 12%, p < 0.05) 
  • MoCA vs MMSE-sevens (AUC difference 12%; p < 0.05) 
• MoCA screening cutoff for PD-D = 21/30 (SN 81%, SP 95%, AUC 97%; 95% CI 92-99%, NPV = 95%) 
  • MoCA vs MMSE-world (AUC difference 10%, p < 0.05) 
  • MoCA, MMSE, and SCOPA-COG, All 3 accurately discriminated for PD-D without statistical distinction on ROC values between measures

(Hoops et al, 2009; N = 132 with idiopathic PD; 75.8% male; 94.7% white; mean age 65.1 (9.7); PD duration 6.5 (5.3); Education level = 16.4 (3.1) years. 30% determined to have cognitive disorder using Dementia criteria: > 1.5 SD below normative mean in at least 2 of 4 cognitive domains, self-report of cognitive dysfunction, and cognition interfering with IADL. PD-Norm (n = 92): H&Y stage 1: 50%, 2: 41.3% 3: 7.6% 4: 1.1%; GDS-15 3.0 (3.4), (n = 40) PD-MCI and PD-D (n = 40): H&Y stage 1 = 17.5%, 2 = 62.5%, 3 = 15.0%, 4 = 2.5%, 5 = 2.5%; GDS-15 score 4.3 (4.0). 

• MoCA cutoff for any cognitive disorder (MCI or D) = 26/30 (sensitivity = 0.90, specificity = 0.53). PPV = 46; NPV = 92, and 64% accuracy. 
• In comparison, cutoff for MMSE = 29/30 (SN 0.9, SP 0.38), PPV = 0.39 and 54% accuracy 

(Robben et al, 2010; N = 41; Young group (< 66, n = 22. PDD, n = 5, HY stages [%]: 1 = 7, 2 = 12, 3 = 3) Older group (> 65, n = 19, PDD n = 10; HY stages [%]: 2 = 6, 3 = 6, 4 = 6, 5 = 1). Prospective study; Blinded examiner. Questionnaire, MoCA, FAB, ACE-R) 

• MoCA cutoff for PD-Dementia = 22/30 for the older group. (SN 100%, SP 100%, AUC (95% CI) = 1.0 (1.0-1.0) and 23/30 for PDD young group (SN 80%, SP 88.2%, AUC (95% CI) = 0.81 (0.58-1.0). Scores on MoCA, FAB, and ACE-R significantly lower in PDD young group (MW U: p < 0.05) Scores on MoCA, FAB, and ACE-R significantly lower in PDD older group (MW-U: p < 0.01) 
• MoCA did not show a false negative result but did take longer to administer (~16 min) than FAB and ACE-R. 
• Recommended sequence: 1) questionnaire, if positive, follow-up with 2) Screening with MoCA or other screening measure, if positive, followup with 3) full Neuropsychologic Exam assessment battery

Parkinson Disease:

(Hoops et al, 2009)

• PD norm = MoCA score 26.2 (2.9); MMSE score 28.7 (1.5), 
• PD-MCI and PD-D = MoCA score 22.2 (4.1); MMSE score 26.8 (2.3)

(Gill et al, 2008; N = 38 (17.5% female); mean age = 71.3 (10.5) yrs, Education 14.8 (3.1) yrs, , H&Y stage 2.9 (0.94), Schwab and England 79% (12), Symptom duration 6.6 (5.4) yrs, Geriatric Depression Scale 1.9 (1.3)) 

• MoCA displays lower scores across progressive disease stages and wider range of scores than MMSE. Range of scores: 6-28 for MoCA, while MMSE range was 16-30
• Mean MoCA score of 23.3 (2.1) was significantly lower than mean MMSE score of 27.4 (1.9) for this group (p < 0.01) 
  • HY Stages 1-2: Mean MoCA = 23.3 (4.1); MMSE = 27.6 (2.5) 
  • HY stage 3: mean MOCA = 21.2 (4.8) ; MMSE = 26.9 (3.5) 
  • HY stages 4-5: Mean MoCA = 19.9 (4.3); MMSE = 25.4 (3.0)

Parkinson's Disease:

(Gill et al, 2008) 

• Excellent test-retest reliability (n = 8): ICC = 0.79 (95% CI: 0.36¨C1.2); Tested on average 133 days apart

Parkinson's Disease: 

(Gill et al, 2008) 

• Excellent interrater reliability (n = 11): ICC = 0.81 (95%, CI: 0.41¨C1.2); tested on average 129 days apart

Parkinson's Disease: 

(Gill et al, 2008) 

• Excellent correlation with neuropsychologic battery ICC = 0.72 (p < 0.0001). Neuropsychology Battery included Hopkins Verbal Learning Test-Revised, the Letter Number Sequencing subtest of the Wechsler Adult Intelligence Scale, the Comalli¨CKaplan adaptation of the Stroop, and the Phonemic and Category Verbal Fluency tests 
• Excellent correlation between MMSE and MoCA ICC = 0.66 (p < 0.0001)

Parkinson's Disease: 

(Dalyrimple et al, 2010) 

• Poor correlation with premorbid IQ (r = 0.19; p < 0.05) 

(Nazem et al, 2009): N = 100 with idiopathic PD (70% male, 96% white); Age 65.3 (11.5); Education level 15.7 (3.6) years; disease duration 7.7 (6.4) years; median Hoehn & Yahr stage = 2; mean GDS-15 score 3.4 (3.8) (26% showing clinically sig depression); evaluated ¡°on¡± medication state; 19% with DBS; intact global cognition (MMSE > 25; in top 75^th percentile when adjusted for age & education)

• Despite normal MMSE scores (> 25), 52% scored positive for cognitive impairment on MoCA (< 26) indicating greater potential sensitivity of MoCA. 
• Association between MoCA and UPDRS (Odds Ratio 1.07 (95% CI = 1.02¨C1.11) p = 0.006 was determined to be due to the motor items of the MoCA. 
• Regression analysis demonstrated: Poor correlation MoCA (visuospatial and executive subscores) and UPDRS motor score (r = -0.14, p = 0.17)

Ceiling Effects:

Parkinson¡¯s Disease (Kletzel et al, 2017)

• Poor: 37% of participants performed all items correctly indicating a large ceiling effect.

Parkinson's Disease: 

(Hoops et al, 2009) 

• There was no ceiling effect for MoCA but there was ceiling effect for MMSE. MoCA results involved larger range of scores with 19-point spread (12-30) while the MMSE range was narrower at 9-point spread (22-30) 

(Zadikoff et al, 2008; N = 88 (M: 62, F: 26); mean age = 65 +/-10 yrs; mean disease duration 9.5 (5) years; mean UPDRS III 20.7 (11.6); Education level identified for adjustment; Tested ¡°on¡± med status; Tested using combined version of MoCA and MMSE using cutoff of 26 for each measure) 

• MoCA demonstrated less of a ceiling effect when compared to the MMSE (when controlled for educational level). 
• More subjects scored < 26 on MoCA than on MMSE (x² = 22.5, p < 000002) 
• If subject scored > 25 on MoCA, they did not score < 26 on the MMSE. 
• In contrast, 36% of those who scored >25 on MMSE had score of <26 on MoCA (p < 0.0001)

Vascular cognitive impairment: (Koski, 2013)  

• When using the originally published cut-off scores, the prevalence of cognitive impairment among individuals with cerebrovascular disease is greater when using the MoCA than the MMSE
• For any impairment versus no impairment, the optimal cut-off score for the MoCA yielded a sensitivity of 0.72 and a specificity of 0.67
  • When discriminating between the MCI (CDR 0.5) and no impairment (CDR 0), sensitivity declined to 0.63, while specificity stayed at 0.67
• In a total of 90 patients with mild-to-moderate stroke severity, a cut-off score of <24 was determined on the MoCA, resulting in a sensitivity of 0.88 and a specificity of 0.71 for detecting cognitive impairment (including dementia)
  • Lowering the cut-off score to <23 caused a specificity of 0.77 and lowered the sensitivity to 0.78
  • MMSE cut-off scores obtained comparable sensitivities and specificities to the MoCA
• In 91 non-demented stroke or TIA patients, a cut-off score of <26 resulted in a sensitivity of 0.87 and a specificity of 0.63, while decreasing the cut-off score to <25 increased the specificity to 0.82 and reduced sensitivity to 0.77
  • With a cut-off score of <29 for the MMSE, the sensitivity was 0.77 and the specificity was 0.81
• Optimal cut-off scores for detecting cognitive impairments tends to be lower when the MoCA is utilized in populations that include individuals with higher cognitive impairment

Chronic kidney disease: (Tiffin-Richards et al., 2014; chronic kidney disease patients undergoing hemodialysis (HD); n = 43; age = 58.3 years (patients) and 57.9 years (control); gender = 52.1% male (patients) and 47.6% male (control); education = 12.0 years (patients) and 13.0 years (control); time on dialysis = 4.0 hours)

• An optimal cut-off score of ¡Ü24 was identified for the MoCA, with a sensitivity of 76.7%, specificity of 78.6%, and AUC of 0.755 (95% CI, 0.602 ¨C 0.872)
  • The optimal cut-off score of ¡Ü24 points is lower than the cut-off score of ¡Ü26 determined in populations of patients with Alzheimer¡¯s disease and mild cognitive impairment (MCI).

Heart Disease: 

(Rossetti et al, 2011; n = 2653; mean age 50.03, range 18-85; sample from Dallas Heart Study incorporating multiple ethnicities: 25% Caucasian, 52% African-American, 11% Hispanic)

• Suggests the cut-off score recommended by developers may be too low, since 62% of the sample would be classified with cognitive impairment even with points added based on years of education.

• Normative values provided in study may be a better guide for performance especially for different ethnic backgrounds.

Normative sample:

(Rossetti et al, 2011) 

Vascular cognitive impairment: (Koski, 2013)  

• Excellent correlation between ischemic stroke patients with small vessel disease and controls tested two weeks apart (r=0.96)
• Poor correlation in older persons from community or memory clinic-based populations due to longer test-retest intervals of 4-8 weeks (r=0.75-0.92)

Vascular cognitive impairment: (Koski, 2013)  

• Moderate-to-high internal consistency 
  • Adequate internal consistency for small vessel disease (Cronbach¡¯s alpha = 0.72)
  • Adequate internal consistency for mild subacute stroke patients undergoing rehab (Cronbach¡¯s alpha = 0.78)
  • Excellent internal consistency for patients studied 3 months after a stroke (Cronbach¡¯s alpha = 0.86)
  • Excellent internal consistency for patients with vascular dementia (Cronbach¡¯s alpha = 0.83)
• Positive correlation between individual MoCA subtests scores and the total MoCA score in individuals with vascular dementia
• Coefficients ranged from poor 0.49 (memory) to excellent 0.78 (attention, concentration, and working memory)

Concurrent validity:

Chronic kidney disease

(Tiffin-Richards et al., 2014)

• Adequate to excellent Correlation between the composite scores of the MoCA and the test battery in the memory (r_s = .53, p<.001) and executive function (r_s = .60, p<.001) domains.
• Adequate positive correlation between the MoCA and MMSE total scores (r_s = .54, p<.001).
• No significant association between the attention, language, and visuospatial composites were exhibited.

Discriminant validity:

Chronic kidney disease

(Tiffin-Richards et al., 2014)

• Adequate association in the patient group between MoCA total score and age of HD patients (r_s = -.38, p<.05).
• Poor association between MoCA and education (r_s = .28, p<.01).
• Adequate correlation between the CCI score for HD patients and MoCA total scores (r_s = -.53, p<.001).

Interrater Reliability:

Stroke (Cumming et al, 2018)

• Interrater reliability between original and blind scoring for the visuospatial/executive MoCA items was the following:
  • Almost Perfect for trail-making ( Kappa=0.94, 95% confidence interval [CI] 0.92-0.96)
  • Substantial for cube copy (Kappa=0.80, 95% CI 0.76-0.83)
  • Substantial for clock numbers (Kappa=0.67, 95% CI 0.62-0.73)
  • Moderate for clock hands (Kappa=0.46, 95% CI 0.42-0.51)
  • Moderate for clock contour (Kappa=0.49, 95% CI 0.40-0.58)

Subacute mild stroke: 

(Toglia et al, 2011; n = 72; mean age 70 (17) years; 8.5 days post-stroke with mild neurological (NIHSS 4) and cognitive (MMSE median 25) deficits)

• Excellent internal consistency (Chronbach¡¯s alpha = 0.78)

• Higher internal consistency than MMSE (a = 0.60)

Subacute mild stroke:

(Toglia et al, 2011)

• Excellent correlation of MoCA with MMSE (r = 0.79) and Cognitive FIM scores (r = 0.67)

• Adequate correlation with discharge status (r = 0.40), which is higher than MMSE (r = 0.30)

• Stronger relationship of MoCA scores to rate of functional improvement (formula using admission and discharge FIM scores, LOS) than the MMSE

• MoCA visuoexecutive subscore was strongest predictor of functional status and improvement in FIM scores

Subacute mild stroke:

(Toglia et al, 2011)

• Identified more patients as having cognitive impairment than usual cutoff points for MMSE (89% vs. 63%)

• Attributed to greater ceiling effects with MMSE

• Mean scores for delayed recall, visuoexecutive and verbal fluency were all < 50% of maximum score

Older Adults (Feeney, et al., 2016; n=130; participants aged 55 or older)

• SEM of the MoCA was 1.5

Older Adults (Feeney, et al., 2016; n=130; participants aged 55 or older)

• An individual score would have to change by greater than or equal to 4 points on the MoCA for the rater to be confident that the change was not due to measurement error.
• 3.54 and 4.21 at the 90% and 95% level

Cut-Off Scores:

Older Adults (Carson, Leach, & Murphy, 2017; N= 304 articles; Mean age 75)

• The cutoff of 23 optimally balanced sensitivity and specificity and provided the highest classification accuracy
  • overall sensitivity of .83 and specificity of .88
• A MoCA cutoff score of 23, rather than the initially recommended score of 26, lowers the false positive rate and shows overall better diagnostic accuracy

Normative Data:

Older Adults (Malek-Ahmadi et al, 2015)

Normative Data based on age and education

Older Adults (Feeney, et al., 2016; n=130; participants aged 55 or older)

• Excellent reliability of 0.81

Cardiovascular sample: (McLennan et al., 2011; n = 110; mean age = 67.9 (11.7) years; gender = 40% (44) male; education = 10.5 (3.2) years)

• The recommended cut-off score was <26.

HIV: (Fazeli et al, 2017; n = 100; mean age = 58.2(6.5) years; sex = 88% male; education = 14.3 (2.6) years; race = 82% white)

• The receiver operating characteristic (ROC) analysis for the MoCA determined the cut-off score of ¡Ü26. This is the most optimal balance of sensitivity (84.21%) and specificity (55.81%) (AUC = 70.42 (95% CI = -0.49 to -0.14, p<.01)).
• The ROC yielded the same cut-off score of ¡Ü26 in a sample 84 participants without severe contributing neuromedical comorbidities, determining a sensitivity of 86.36% and specificity of 57.50% (AUC = 71.19 (95% CI = -0.52 to -0.14, p<.01)).

Cardiovascular sample: (McLennan et al., 2011; n = 110)

• Poor Cronbach¡¯s alpha was 0.55

Predictive Validity

HIV: (Fazeli et al, 2017)

• Poor correlation with IADL declines (r = -0.29), Karnofsky score (r = 0.28), and reported cognitive symptoms (r = 0.28).
• Poor correlation with current depressive symptoms (r = -0.04, p = .72)

Convergent Validity

Cardiovascular sample  (McLennan et al., 2011)

• Moderate positive correlation between the MoCA and the Neuropsychological Assessment Battery Screening Module (NAB-SM) global score (p=0.49, P<0.001)
• Moderate positive correlation between the MoCA and each of the Neuropsychological Assessment Battery Screening Module (NAB-SM) domain scores: attention (r_s = .34, P < .001); memory (r_s .25, P=.007); spatial (r_s = .48, P < .001); and executive (r_s = .41, P < .001)

HIV: (Fazeli et al, 2017)

• Adequate correlations with the following domains of neurocognitive performances: verbal (r = -0.34), working memory (r = -0.38), executive functions (r = -0.49), learning (r = -0.39), speed of information processing (r = -0.31), and global (r = -0.44).
• Poor correlation with the domain of neurocognitive performance: motor (r = -0.11)
• Adequate correlation with education levels (r = 0.35)

Huntington¡¯s disease (Bezdicek et al., 2013; n = 20 (HD) and 23 (normal control (NC)); gender = 12 males, 8 females (HD) and 12 males, 11 females (NC))

• The optimal cut-off score was 25/26, with a sensitivity of 0.94 and a specificity of 0.84 (PPV = 0.81, NPV ¨C 0.95) for 3 measures: optimal diagnostic cut-off, optimal screening cut-off, and point of maximal combined sensitivity and specificity

Huntington¡¯s disease: (Bezdicek et al., 2013; n = 20 (HD) and 23 (normal control (NC)); gender = 12 males, 8 females (HD) and 12 males, 11 females (NC))

• Adequate internal consistency for seven subtests of MoCA for 20 HD patients (Cronbach¡¯s alpha = .82) and 23 controls (Cronbach¡¯s alpha = .56)

Concurrent validity:

Huntington¡¯s disease

• Modest-to-strong correlations between the MoCA and other measures of cognitive functioning (Bezdicek et al., 2013)
  • Excellent correlation between visuospatial/executive and Rey Complex Figure Test (RCFT) Immediate Recall (r=.64)
  • Poor correlation between naming and Rey Complex Figure Test (RCFT) Immediate Recall (r=.30)
  • Excellent correlation between attention and free recall from Free and Cued Selective Reminding Test (FCSRT) (r=.63)
  • Excellent correlation between language and Controlled Oral Word Association Test (COWAT) (r=.81)
  • Excellent correlation between abstraction and Controlled Oral Word Association Test (COWAT) (r=.62)
  • Excellent correlation between delayed recall and Symbol-Digit Modalities Test (SDMT) (r=.72)
  • Adequate correlation between orientation and Symbol-Digit Modalities Test (SDMT) (r=.47)

Discriminant Validity:

Huntington¡¯s Disease

• Excellent AUC (95%) was 0.90 (0.809-0.997), p<.001 for the MoCA in comparison with the brief cognitive battery(Bezdicek et al., 2013)

Cutoff Scores:

Schizophrenia (Yang et al, 2018; N=64 with schizophrenia; Aged 16-65; Mean 13.58 (2.72) years of education; Mean 9.10 (7.29) duration of iIlness; Mean 12.86 (6.85) duration of psychiatric treatment)

• For severe cognitive impairments, the optimal MoCA cut-off score was <23. AUC for the MoCA was 0.814 (95% CI=0.710-0.918, p<0.001)
• For mild cognitive impairments, the optimal MoCA cut-off score was <25. AUC for the MoCA was 0.816 (95% CI=0.698-0.934, p<0.001).

Substance Use Disorder (Bruijnen et al., 2018; n = 82; mean age = 44.1; gender = 68.3% male; primary substance abuse problem = alcohol (70.7%); 54.9% had neurocognitive disorder)

• AUC value of .676 was found and an optimal cut-off score of 24 resulted in a sensitivity of .56 and a specificity of .62 (PPV = 64.1% and NPV = 53.5%)
  • 39 out of 82 patients were classified as having a NCD when using this cut-off score
• AUC value of .745 was found and an optimal cut-off score of 25 resulted in a sensitivity of .67 and a specificity of .73 (PPV = 75.0% and NPV = 64.3%)
  • 40 out of 82 patients were classified as having NCD when using this cut-off score

Substance Use Disorder (Ewert et al., 2017; n=56; 31 with cognitive impairment and 25 without; mean age= 49.5; gender= 40 men, 16 women)

Using the 1©\point correction for education level increases the cutoff score by 1 point, it is suggested that the uncorrected score be used with the usual cutoff, that is, 26

Concurrent Validity:

Schizophrenia (Yang et al, 2018)

• Excellent correlations between the MoCA adjusted total score and the Brief Assessment of Cognition in Schizophrenia composite Z-score (r=0.61, p<0.001).
• Adequate correlations between the MoCA and the Z-scores of five Brief Assessment of Cognition in Schizophrenia subscales:
  • Adequate correlations with the Verbal Memory subscale (r=.44, p<0.001).
  • Adequate correlations with the Digit Sequencing subscale (r = .57, p < .001).
  • Adequate correlations with the  Semantic Fluency subscale (r = .47, p < .001).
  • Adequate correlations with the  Symbol Coding subscale (r = .35, p = .004).
  • Adequate correlations with the Tower of London subscale (r = .45, p < .001).
• Adequate correlations between the adjusted MoCA total score and the Brief UCSD Performance-based Skills Assessment (r(60)=0.51, p<0.001).

Concurrent validity:

Substance Use Disorder (Bruijnen et al., 2018)

• Significant correlations between performance on the MoCA and NPA at baseline
  • Poor correlation for executive functioning (r=.238, p=.032)
  • Poor correlation for abstract reasoning (r=.300, p=.006),
  • Adequate correlation for memory (r=.423, p<.001)
• Near perfect correlation between performance on the MoCA and NPA at follow-up
  • Significant correlation between all MoCA domain and corresponding NPA domains
    • Adequate correlation for executive functioning: r=.328, p=.003
    • Poor correlation for visuospatial abilities: r=.241, p=.029
    • Adequate correlation for attention: r=.396, p<.001
    • Adequate correlation for abstract reasoning: r=.542, p<.001
    • Adequate correlation for memory: r=.455, p<.001
    • Poor correlation for orientation: r=.229, p=.043

Predictive validity:

Substance Use Disorder (Bruijnen et al., 2018)

• During baseline and follow-up, patients with a NCD performed significantly worse on the MoCA than patients without a NCD
  • Similar results were found for the executive functioning domain on the MoCA (p=.003; p=.009)
  • Performance on the memory domain on the MoCA was significant at baseline between both groups (p=.014; p=.073)
  • Performance on the language and abstract reasoning domains on the MoCA were significantly different between both groups at follow-up (p=.984; p=.043 & p=.423; p=.010)
• Performances on other domains on the MoCA did not differ significantly between both patient groups