Related Experiment Video
Updated: Nov 9, 2025
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
Evaluation of a Telephone Version for the Montreal Cognitive Assessment: Establishing a Cutoff for Normative Data
Sivan Klil-Drori1, Natalie Phillips2, Alita Fernandez1
1Rotman Research Institute, 7942Baycrest Health Sciences, University of Toronto, Ontario, Canada.
Objective:
Compare a telephone version and full version of the Montreal Cognitive Assessment (MoCA).
Methods:
Cross-sectional analysis of a prospective study. A 20-point telephone version of MoCA (Tele-MoCA) was compared to the Full-MoCA and Mini Mental State Examination.
Results:
Total of 140 participants enrolled. Mean scores for language were significantly lower with Tele-MoCA than with Full-MoCA (P = .003). Mean Tele-MoCA scores were significantly higher for participants with over 12 years of education (P < .001). Cutoff score of 17 for the Tele-MoCA yielded good specificity (82.2%) and negative predictive value (84.4%), while sensitivity was low (18.2%).
Conclusions:
Remote screening of cognition with a 20-point Tele-MoCA is as specific for defining normal cognition as the Full-MoCA. This study shows that telephone evaluation is adequate for virtual cognitive screening. Our sample did not allow accurate assessment of sensitivity for Tele-MoCA in detecting MCI or dementia. Further studies with representative populations are needed to establish sensitivity.
