The Colorado Cognitive Assessment (CoCA): Development of an Advanced Neuropsychological Screening Tool.

Abstract

The Colorado Cognitive Assessment (CoCA) was designed to improve upon existing screening tests in a number of ways, including enhanced psychometric properties and minimization of bias across diverse groups. This paper describes the initial validation study of the CoCA, which seeks to describe the test; demonstrate its construct validity; measurement invariance to age, education, sex, and mood symptoms; and compare it to the Montreal Cognitive Assessment (MoCA). Participants included 151 older adults (MAge=71.21, SD=8.05) who were administered the CoCA, MoCA, Judgment test from the Neuropsychological Assessment Battery (NAB), 15-item version of the Geriatric Depression Scale (GDS-15), and 10-item version of the Geriatric Anxiety Scale (GAS-10). A single-factor confirmatory factor analysis model of the CoCA fit the data well, CFI=0.955; RMSEA=0.033. The CoCA factor score reliability was .84, compared to .74 for the MoCA. The CoCA had stronger disattenuated correlations with the MoCA (r=.79) and NAB Judgment (r=.47) and weaker correlations with the GDS-15 (r=-.36) and GAS-10 (r=-.15), supporting its construct validity. Finally, when analyzed using multiple-indicators, multiple-causes (MIMIC) modeling, the CoCA showed no evidence of measurement noninvariance, unlike the MoCA. These results provide initial evidence to suggest that the CoCA is a valid cognitive screening tool that offers numerous advantages over the MoCA, including superior psychometric properties and measurement noninvariance. Additional validation and normative studies are warranted.