Abstract Body

Background: HIV-associated neurocognitive disorders (HAND) are prevalent and often under-diagnosed complications of HIV. The International HIV Dementia Scale (IHDS) has been validated as a useful screening tool for HIV-associated dementia (HAD) in a number of settings, including Kampala, Uganda. However, little is known about the utility of the IHDS in rural sub-Saharan Africa. The objective of this study was to assess the validity of the IHDS as an appropriate screening tool for detecting different levels of HAND in the rural district of Rakai, Uganda.

Methods: 299 HIV+ antiretroviral treatment (ART) naïve participants in the Rakai Community Cohort Study underwent comprehensive standardized neurological, neuropsychological and functional assessments, including the IHDS. HAND stages were determined based on the Frascati criteria. The diagnostic validity of the IHDS was assessed by determining the sensitivity, specificity, predictive values and area under the ROC curve (AUC) for IHDS cut-off scores of 9, 9.5 and 10 (lower score= worse performance).

Results: Participants’ demographics were: Mean(SD) age = 36(9) years; Male = 51%; Education ≤ 4th grade = 47%; Mean(SD) CD4 count = 309(159). For detecting any level of HAND [asymptomatic neurocognitive impairment (ANI), mild neurocognitive disorder (MND), and HAD], an IHDS cut off of ≤9 had a sensitivity and specificity of 60% and 71% (AUC 0.66), while a cut-off of ≤9.5 had a sensitivity and specificity of 67% and 64% (AUC 0.66), and a cut-off of ≤10 had a sensitivity and specificity of 80% and 55% (AUC 0.68). For detecting symptomatic HAND (MND and HAD), a ≤9 cut-off had a sensitivity and specificity of 62% and 62% (AUC 0.62), ≤9.5 had a sensitivity and specificity of 70% and 54% (AUC 0.62), and ≤10 had a sensitivity and specificity of 83% and 44% (AUC 0.64). For detecting HAD, an IHDS ≤9 cut-off had a sensitivity and specificity of 75% and 54% (AUC 0.65), ≤9.5 had a sensitivity and specificity of 80% and 46% (AUC 0.63), and ≤10 had a sensitivity and specificity of 90% and 32% (AUC 0.61).

Conclusions: The IHDS is a sensitive and potentially useful screening tool for neurocognitive impairment in rural Uganda. When used for screening in a setting with comprehensive neuropsychological testing to confirm a HAND diagnosis, the ≤10 cut-off may be most useful as it provides the highest sensitivity. However, if further neuropsychological testing is unavailable, the higher specificity of a ≤9 cut-off may be more appropriate, particularly for HIV dementia screening.