Sensitivity of conventional cognitive tests in multiple sclerosis: Application of item response theory

Background: Cognitive impairment (CI) is common in Multiple Sclerosis (MS), and its prevalence rate ranges between 22% and 70%. Because CI significantly impacts vocational status, caregiver burden, and quality of life, an accurate neuropsychological assessment is required. Three widely used and validated batteries for MS-associated CI are the Brief Repeatable Neuropsychological Battery (BRN-B), the Minimal Assessment of Cognitive Function (MACFIMS), and the Brief International Cognitive Assessment (BICAMS). Although similar, these batteries differ in time-consuming and in specific tests employed. This study aims to assess the sensitivity of cognitive tests included in these batteries through an Item Response Theory approach. Methods: Ninety-seven patients with MS and 91 demographically matched controls (HC) were consecutively assessed using the three neuropsychological batteries (i.e., BRN-B, MACFIMS, and BICAMS). Continuous Response Model (CRM) was used to identify the cognitive test(s) that best discriminate patients with MS from HC. Receiver Operating Characteristic (ROC) curve analysis was used to determine the accuracy of the CRM results. Results: Cognitive tests loaded on two different latent variables: the 'higher-order executive functioning,' consisting of tests assessing concept formation, problem-solving, and inhibitory control, and the 'memory and information processing speed,' comprising tests assessing long-term, working memory, and information processing speed. The Delis Kaplan Executive Functioning System-Sorting Test and the Stroop Test were the most sensitive tests in differentiating cognitive functioning between MS and HC. Conclusions: This study confirms the importance of including a more extensive executive assessment in MS clinical practice since higher-order executive functions (e.g., abstraction and inhibitory control) significantly impact patients' quality of life and functional autonomy. Clinical implications of careful dissection of executive functioning in MS neuropsychological assessment are discussed.