Montreal Cognitive Assessment Hearing Impairment (MoCA-H) in Brazilian Portuguese: criterion and construct validity.

PURPOSE: To find evidence of criterion and construct validity for the Montreal Cognitive Assessment Hearing Impairment (MoCA-H) protocol in Brazilian Portuguese. METHODS: The sample consisted of 70 elderly people divided into two groups: Group 1-50 subjects with hearing loss and no cognitive decline; Group 2-20 subjects with hearing loss and cognitive decline. Criterion validity was obtained by comparing Group 1 and 2 considering the overall score and the eight domains assessed in the MoCA-H. The data were analyzed using the Mann-Whitney U-test and Student's T-test, respecting the characteristics of the data collected. To verify construct validity, the correlation between the total scores of the Mini-Mental State Examination (MMSE) and the MoCA-H obtained by Group 2 was analyzed. Spearman's Correlation Test was used for this purpose. RESULTS: The analysis of criterion validity showed a difference between the groups with and without decline in naming, attention, language, abstraction, memory and delayed recall skills, as well as the MoCA-H total score, indicating significantly higher performance of Group 1. The construct validity correlation analysis was weak and non-significant (Rho=0.384; p=0.095) between the MoCA-H and MMSE scores. CONCLUSION: The MoCA-H protocol showed good criterion validity for this specific population, making it a reliable tool for screening mild cognitive decline. However, it did not show satisfactory construct validity, indicating the need for further studies with this instrument using another protocol as a reference.