Related Experiment Video
Updated: Aug 22, 2026
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
Characteristics of the GPCOG, a screening tool for cognitive impairment
Henry Brodaty1, Nicola M Kemp, Lee-Fay Low
1School of Psychiatry, University of New South Wales, Australia. h.brodaty@unsw.edu.au
Background:
Early dementia diagnosis is aided by the use of brief screening tests; scores can be biased by patient and informant characteristics such as age, gender and education.
Objective:
To assess whether the General Practitioner's Assessment of Cognition (GPCOG), a brief screening tool for detecting cognitive impairment comprising a patient cognitive test and questions to an informant, is biased by patient and informant characteristics.
Design:
Sixty-seven general practitioners recruited consecutive patients (with informants). Patients were subsequently assessed by a research psychologist, and DSM-IV diagnoses assigned following a case-conference.
Setting:
Primary Care.
Subjects:
Two hundred and eighty three home-dwelling individuals, 11.3% of whom were aged 50-74 years with suspected memory problems and the rest aged 75 or more.
Methods:
The GPCOG, Cambridge Mental Disorder of the Elderly Examination cognitive scale (CAMCOG), Geriatric Depression Scale (GDS), and the SF-12 Health Survey (SF-12) were administered and demographic data were collected and consensus DSM-IV diagnoses of dementia made. Relationships between patient and informant characteristics and the GPCOG measure were examined using Pearson correlations and linear regression analyses.
Results:
There were correlations in GPCOG-patient scores with age, education and depression scores but on regression analysis only age was associated with the GPCOG-patient section. The GPCOG-informant section was free of bias.
Conclusions:
The GPCOG has advantages for use in primary care and is free of many biases common in other scales.
