Objectives To estimate the prevalence of HIV-associated neurocognitive disorders (HAND) among HIV patients in a multiethnic South Asian population, describe the pattern of neurocognitive impairment in HAND and the factors associated with HAND. Design A cross-sectional survey of HIV-positive outpatients and inpatients. Setting The sole referral centre for HIV/AIDS treatment in Singapore. Participants Inclusion criteria were HIV positive, age between 21 and 80 years old and at least 3 years of education. Exclusion criteria included concomitant delirium, serious systemic disease or major psychiatric illness. 265 patients did not meet criteria or declined to participate. The final sample size was 132. Outcome measures The primary outcome measure was cognitive impairment based on performance on the Montreal Cognitive Assessment, International HIV Dementia Scale and Instrumental Activities of Daily Living. The secondary outcome measure was the classification of impairment based on the 2007 updated research nosology for HAND. Results The prevalence of HAND was 22.7% of which 70% (15.9% of total) were asymptomatic neurocognitive impairment, 23.3% (5.3% of total) were mild neurocognitive disorder and 6.7% (1.5% of total) were HIV-associated dementia. Increasing age (OR 1.104, 95% CI 1.054 to 1.155, p<0.001), less education (OR 0.78, 95% CI 0.69 to 0.89, p<0.001) and low baseline CD4 count (OR 0.15, 95% CI 0.03 to 0.74, p=0.019) were associated with HAND. Delayed recall, language and abstract thinking were the domains most commonly affected, but impairment in visuospatial ability (RC 3.013, 95% CI 1.954 to 4.073, p<0.001) and attention (RC 2.205, 95% CI 1.043 to 3.367, p<0.001) were most strongly associated with HAND. Conclusion HAND is common among HIV patients in a South Asian sample, most of whom are asymptomatic. Older patients with less education and severe illness at diagnosis are at highest risk of HAND. Delayed recall is most commonly affected, but visuospatial dysfunction is most strongly associated with prevalent HAND.

Post-stroke depression and anxiety, collectively known as post-stroke adverse mental outcome (PSAMO) are common sequelae of stroke. About 30% of stroke survivors develop depression and about 20% develop anxiety. Stroke survivors with PSAMO have poorer health outcomes with higher mortality and greater functional disability. In this study, we aimed to develop a machine learning (ML) model to predict the risk of PSAMO. We retrospectively studied 1780 patients with stroke who were divided into PSAMO vs. no PSAMO groups based on results of validated depression and anxiety questionnaires. The features collected included demographic and sociological data, quality of life scores, stroke-related information, medical and medication history, and comorbidities. Recursive feature elimination was used to select features to input in parallel to eight ML algorithms to train and test the model. Bayesian optimization was used for hyperparameter tuning. Shapley additive explanations (SHAP), an explainable AI (XAI) method, was applied to interpret the model. The best performing ML algorithm was gradient-boosted tree, which attained 74.7% binary classification accuracy. Feature importance calculated by SHAP produced a list of ranked important features that contributed to the prediction, which were consistent with findings of prior clinical studies. Some of these factors were modifiable, and potentially amenable to intervention at early stages of stroke to reduce the incidence of PSAMO.

Background: Uncertainties related to COVID-19 have strained the mental health of healthcare workers (HCWs) worldwide. Gaining the ability to adapt and thrive under pressure will be key to addressing this. We explore what characterises risk, vulnerability and resilient responses of HCWs during the early stages of the outbreak in Singapore. Methods: We undertook qualitative theory-guided thematic analysis of e-diary entries from HCWs who navigated the outbreak from June–August 2020. Data were extracted from a subset of an online survey of n = 3616 participants collected across 9 institutions, including restructured hospitals, hospices and affiliated primary care partners. Results: N = 663 or 18% submitted qualitative journal entries included for analyses. All professional cadres, local as well as foreign HCWs participated. Themes are reported according to the Loads–Levers–Lifts model of resilience and highlighted in italics. The model assumes that resilience is a dynamic process. Key factors threatening mental health (loading) risk included a notable rise in anxiety, the effects of being separated from loved ones, and experiencing heightened emotions and emotional overload. Bad situations were made worse, prompting vulnerable outcomes when HCWs experienced stigma in the community and effects of “public paranoia”; or under conditions where HCWs ended up feeling like a prisoner with little control or choice when either confined to staff accommodation or placed on quarantine/Stay Home Notices. Those with strife in their place of residence also described already difficult situations at work being aggravated by home life. Protection (lifts) came from being able to muster a sense of optimism about the future or feeling grateful for the pace of life slowing down and having the space to reprioritise. In contrast, when risk factors were present, balancing these in the direction of resilient outcomes was achieved by choosing to re-direct stress into positive narratives, drawing on inner agency, uptake of therapeutic activities, social support as well as faith and prayer and drawing comfort from religious community among other factors. Conclusion: The Loads–Levers–Lifts model is used to guide analysis to inform intervention designs. Levers promoting resilience through targeting therapies, workplace policies and awareness campaigns accounting for identified loads are proposed.

Practical screening strategies are necessary to detect neurocognitive impairment of all severities in HIV populations, which remains prevalent despite highly active antiretroviral therapy and requires full neuropsychological testing for diagnosis. We aimed to develop a brief and clinically feasible battery to screen for HIV-associated neurocognitive disorders (HAND) in resource-limited settings even where English is not the native language. A total of 53 outpatients were recruited from a multi-ethnic Southeast Asian HIV-positive cohort. Performance on a neuropsychological protocol was used to define cognitive impairment, of which 28 patients (52.8%) were identified with HAND. Receiver operating characteristic analysis was used to determine the best combinations of cognitive tests for the screening battery. 3 different combinations of cognitive tests that required minimal literacy, time to administer, and administrator training were found to classify HAND with fair accuracy. Montreal Cognitive Assessment (MoCA), in combination with tests of psychomotor coordination, verbal learning and speed processing, yielded area under curve scores of above 0.75, the primary outcome of receiver operating characteristic analysis. The 3-test combinations presented in this study appear to be promising screening options for HAND in HIV-infected patients. The addition of 2 tests to MoCA improves the overall accuracy while retaining its convenience, giving more potential for the inclusion of cognitive screening in routine clinical care. Further validation of the batteries in specific settings is warranted to determine specific screening cut-offs to a global cognitive score.

Objectives: Research suggests a high prevalence of depression and anxiety in people living with HIV, resulting in negative health outcomes and poorer help-seeking behaviours when undetected. Subsequent disease progression and non-adherence to treatment constitute a significant barrier to HIV treatment. This paper aims to identify the risk factors for the development of psychological distress and non-adherence to antiretroviral medication in people living with HIV.Methods: An HIV outpatient clinical service screened for anxiety and depressive symptoms. As part of a retrospective analysis of the cohort, independent sample t-test and χ²test were conducted to examine differences between symptomatic and asymptomatic patients in demographic variables such as mode of transmission and disclosure to family; clinical indicators such as psychiatric history and history of alcohol and substance use; and outcome variables such as current psychological distress and non-adherence. Binary logistic regression was conducted to determine predictors of psychological distress and non-adherence.Results: After adjusting for age, no history of alcohol use and psychiatric history were found to be significant risk factors for psychological distress during the programme. Older patients were less likely to be symptomatic during the programme. After adjusting for age, having received intervention and psychiatric history, significant risk factors for non-adherence to antiretroviral medication were mode of transmission, history of smoking and being symptomatic during the programme.Conclusion: Significant psychological distress occurring early in HIV care predicts future non-adherence to antiretroviral treatment, highlighting the importance of early detection and intervention for psychological distress in people living with HIV. Mental health interventions should be intercalated with treatment adherence interventions to improve HIV treatment outcomes.