The Montreal Cognitive Assessment–Basic (MoCA-B) is not a reliable screening tool for cognitive decline in HIV patients receiving combination antiretroviral therapy in rural South Africa

Background HIV-associated neurocognitive disorders (HAND) are frequently occurring comorbidities in HIV-positive patients, diagnosed by means of a neuropsychological assessment (NPA). Due to the magnitude of the HIV-positive population in Sub-Saharan Africa, easy-to-use cognitive screening tools are essential. Methods This was a cross-sectional clinical trial involving 44 HIV-positive patients (on stable cART) and 73 HIV-negative controls completing an NPA, the International HIV Dementia Scale (IHDS), and a culturally appropriate cognitive screening tool, the Montreal Cognitive Assessment–Basic (MoCA-B). HAND were diagnosed by calculating Z-scores using internationally published normative data on NPA, as well as by using data from the HIV-negative group to validate the MoCA-B. Results One hundred and seventeen patients were included (25% male, median age 35 years, median 11 years of education). A moderate correlation was found between the MoCA-B and NPA total Z-score (Pearson's r = 0.36, p = 0.02). Area under the curve (AUC) values for MoCA-B and IHDS were 0.59 and 0.70, respectively. The prevalence of HAND in HIV-positive patients was 66% when calculating Z-scores using published normative data versus 48% when using the data from the present HIV-negative cohort. Conclusion The MoCA-B appeared not to be a valid screening tool for HAND in this setting. The prevalence of HAND in this setting is high, but appeared overestimated when using published norms.