Diagnostic accuracy of IHDS and MoCA in screening for HIV-associated neurocognitive disorders among adults living with HIV in Ethiopia.

Abstract

Background: HIV-associated neurocognitive disorders (HAND) remain prevalent among adults living with HIV, even in the era of sustained antiretroviral therapy (ART). Accurate screening for HAND remains a challenge in resource-limited settings like Ethiopia, as comprehensive neuropsychological assessments thought effective are often limited by their high demands for time, specialized personnel, and logistical resources. The International HIV Dementia Scale (IHDS) and the Montreal Cognitive Assessment (MoCA) are brief cognitive screening tools commonly used in clinical settings; however, their diagnostic accuracy for detecting HAND in the Ethiopian population has not been well established. This study aims to evaluate and compare the sensitivity and specificity of the IHDS and MoCA in identifying HIV-associated neurocognitive disorder among adults living with HIV in Ethiopia.

Methods: A multicenter, cross-sectional study was conducted among 290 adults living with HIV attending three primary hospitals in Ethiopia. Participants aged ≥ 18 years were recruited using systematic random sampling. Clinical data, including baseline CD4⁺ T-cell counts and recent viral load measurements, were extracted from medical records. Cognitive function was assessed using the IHDS and MoCA, with the Frascati criteria used as the diagnostic reference standard, and factors independently associated with IHDS scores were identified through multivariable linear regression.

Results: The study included 290 adults living with HIV, with a mean age of 40.6 ± 8.04 years. Based on the Frascati criteria, 43.4% of participants exhibited HIV-associated neurocognitive disorders (HAND), with asymptomatic neurocognitive impairment being the most common subtype (25.4%), followed by mild neurocognitive disorder (13.3%) and HIV-associated dementia (4.7%). Using the International HIV Dementia Scale (IHDS), 47.9% of participants were identified as cognitively impaired, indicating possible HAND, whereas the Montreal Cognitive Assessment (MoCA) identified 36.2% as impaired. The IHDS demonstrated superior diagnostic performance, with an optimal cutoff score of ≤ 10 (sensitivity 85.3% and specificity 77%), compared to the MoCA at a cutoff score of 21 (sensitivity 62.1% and specificity 81%). Multivariable linear regression analysis revealed that higher educational attainment (β = 0.30, p = 0.005) and male sex (β = 0.41, p = 0.045) were independently associated with better neurocognitive performance, while being unmarried (β = -0.29, p = 0.01) and a longer duration of antiretroviral therapy (β = -0.26, p = 0.04) were significantly associated with poorer cognitive outcomes.

Conclusion: HIV-associated neurocognitive disorders remain highly prevalent among adults living with HIV in Ethiopia. The IHDS demonstrated better sensitivity than the MoCA for detecting HAND, making it a more suitable screening tool in resource-limited settings. Educational status, sex, marital status, and duration of ART were significant predictors of cognitive performance. Integrating routine cognitive screening using IHDS may improve early identification and management of HAND.