Abstract Body

Background: The NEU Screen is a method recently proposed to help in the diagnosis of HIV-associated neurocognitive disorders (HAND). We tested this tool in a sample of virologically suppressed HIV-infected patients to investigate both sensitivity and specificity in comparison with those found in the original study. Methodology: In a sample of 156 virologically suppressed HIV-infected outpatients the accuracy of the NEU Screen was analyzed, establishing as a gold standard the existence of neurocognitive impairment (NCI). All subjects had available neurocognitive information through the application of a comprehensive battery of neuropsychological tests, and clinical and demographical information was accessible as well. Sensitivity and specificity tests were applied to study the accuracy of the NEU Screen detecting NCI, and logistic regression was used to analyze clinical and demographic variables linked to the correct classification. Results: Subjects were mostly men (81%), with a median (IQR) age of 43 (38;50) years, current CD4 cell count of 522 (380;718) cells/μL, nadir CD4 cell count of 188 (80;285) cells/mL, and HCV seronegative (73%). The rate of NCI was 52% and appeared significantly associated with time since HIV diagnosis (p=0.01). A trend towards worse neurocognitive functioning was also observed in relationship with lower nadir cell counts (p=0.17) and existence of potential confounding comorbidities for NCI (p=0.15). When the combination of scores included in the NEU Screen was analyzed for the detection of NCI, the sensitivity (95% CI) observed was 73.1% (62%;82%), specificity 74.3% (62.6%;83.4%), positive predictive value 75.9% (64.7%;84.5%), and negative predictive value 71.4% (59.8%;80.8%). According to logistic regression models, the correct classification of NCI by the NEU Screen was unrelated to any relevant demographic or clinical variable. Conclusions: The NEU screen confirms in this work fairly high sensitivity and specificity to detect NCI in HIV population. This method includes 3 paperbased neurocognitive tests, has an expected time for administration of ≤10 minutes, and appears to be also useful detecting NCI in virologically suppressed HIV-infected patients.