Diagnostic test accuracy of the Montreal Cognitive Assessment in the detection of post-stroke cognitive impairment under different stages and cutoffs: a systematic review and meta-analysis.

The purposes of this review were to give the optimal cutoffs of the Montreal Cognitive Assessment (MoCA) by comparing sensitivity and specificity under different cutoffs and compare the MoCA with other screening tools in post-stroke cognitive impairment (PSCI) determined by a neuropsychological evaluation. Articles were derived from a systematic search in PubMed, Web of science, Embase, and CINAHL and were assessed for internal validity by the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2). The figure of risk of bias was made by Review Manager 5.3, and data of selected studies were synthesized by MetaDisc 1.4. Twelve diagnostic studies, involving 2130 patients, were included. The area under the curve (AUC) under cutoffs of 20v19, 21v20, and 26v25 are 0.90, 0.90, and 0.95, showing high predictive validity for PSCI screening within 1 month. When the sensitivity and specificity are equal important, the optimal cutoff is 20v19 (Youden Index = 0.58). Compared to the Mini-Mental State Examination (MMSE), the MoCA has higher sensitivity but lower specificity. The optimal cutoff differs in different stages of stroke. Both the MMSE and MoCA are appropriate screening tools for PSCI, and the use of these two tools should be in accordance with the aim of screening. The Addenbrooke's Cognitive Examination-Revised (ACE-R) can act as a supplement for the MoCA.