Table 5 Cross-tabulation between the output from PENSIEVE-AI and the final diagnosis in Test sample (n = 658)

From: PENSIEVE-AI a brief cognitive test to detect cognitive impairment across diverse literacy

Output from PENSIEVE-AI a

Final diagnosis

Normal cognition

MCI

Dementia

Less likely to have CI, n (%)b

461 (97.5)

12 (2.5)

0 (0.0)

Higher risk of CI, n (%)b

106 (76.8)

28 (20.3)

4 (2.9)

Likely to have CI, n (%)b

7 (14.9)

28 (59.6)

12 (25.5)

Sensitivity analysis 1 (Prevalence of MCI/dementia = 20%)c

Less likely to have CI, n (%)d

209 (95.9)

9 (4.1)

0 (0.0)

Higher risk of CI, n (%)d

42 (70.0)

17 (28.3)

1 (1.7)

Likely to have CI, n (%)d

5 (11.9)

22 (52.4)

15 (35.7)

Sensitivity analysis 2 (Prevalence of MCI/dementia = 35%)e

Less likely to have CI, n (%)f

82 (92.1)

7 (7.9)

0 (0.0)

Higher risk of CI, n (%)f

16 (55.2)

12 (41.4)

1 (3.4)

Likely to have CI, n (%)f

6 (14.3)

21 (50.0)

15 (35.7)

  1. CI cognitive impairment, MCI mild cognitive impairment.
  2. aTwo-cutoff approach was adopted for PENSIEVE-AI. The lower cutoff has high sensitivity and negative predictive value (>85% respectively), and is used to rule out MCI/dementia (for individuals with probability scores below the cutoff). The upper cutoff has high specificity and positive predictive value (>85% respectively), and identifies those who are likely to have MCI/dementia. These two cutoffs provide an intermediate range between them, identifying those who may be at higher risk and require further monitoring or assessment.
  3. bProbability cutoff for the main results (i.e. prevalence of MCI/dementia=13.6%): <13% (Less likely to have CI), 13–44% (Higher risk of CI), ≥45% (Likely to have CI).
  4. cPrevalence of MCI/dementia was readjusted to 20% in the Test sample, based on prior meta-analytic findings that community prevalence was ~15% for MCI and ~5% for dementia. In the Test sample, a subset of participants with MCI and dementia were randomly selected to readjust the prevalence in the dataset (see Methods section for further details). The resulting dataset comprised 256 participants with normal cognition (80%), 48 participants with MCI (15%), and 16 participants with dementia (5%).
  5. dProbability cutoff for the first sensitivity analysis (i.e. prevalence of MCI/dementia=20%): <13% (Less likely to have CI), 13–34% (Higher risk of CI), ≥35% (Likely to have CI).
  6. ePrevalence of MCI/dementia was readjusted to 35% in the Test sample, based on prior meta-analytic findings that community prevalence could be as high as ~25% for MCI and ~10% for dementia. In the Test sample, a subset of participants with MCI and dementia were randomly selected to readjust the prevalence in the dataset (see Methods section for further details). The resulting dataset comprised 104 participants with normal cognition (65%), 40 participants with MCI (25%), and 16 participants with dementia (10%).
  7. fProbability cutoff for the second sensitivity analysis (i.e. prevalence of MCI/dementia=35%): <11% (Less likely to have CI), 11–27% (Higher risk of CI), ≥28% (Likely to have CI).
Back to article page