Rapid cognitive screening of patients with substance use disorders

Abstract

To date, there has not been a time-efficient and resource-conscious way to identify cognitive impairment in patients with substance use disorders (SUDs). In this study, we assessed the validity, accuracy, and clinical utility of a brief (10-min) screening instrument, the Montreal Cognitive Assessment (MoCA), in identifying cognitive impairment among patients with SUDs. The Neuropsychological Assessment Battery-Screening Module, a 45-min battery with known sensitivity to the mild to moderate deficits observed in patients with SUDs, was used as the reference criterion for determining agreement, rates of correct and incorrect decision classifications, and criterion-related validity for the MoCA. Classification accuracy of the MoCA, based on receiver operating characteristic (ROC) analysis, was strong, with an area under the ROC curve of 0.86, 95% confidence interval [0.75, 0.97]. The MoCA also showed acceptable sensitivity (83.3%) and specificity (72.9%) for the identification of cognitive impairment. Using a cutoff of 25 on the MoCA, the overall agreement was 75.0%; chance-corrected agreement (kappa) was 41.9%. These findings indicate that the MoCA provides a time-efficient and resource-conscious way to identify patients with SUDs and neuropsychological impairment, thus addressing a critical need in the addiction treatment research community.

Figure 1

Sensitivity and specificity of the Montreal Cognitive Assessment in classifying 60 patients as cognitively impaired versus unimpaired. The area under the receiver operating characteristic curve = 0.86, standard error = 0.06, asymptotic normal 95% confidence interval [0.75, 0.97].