COMMUNAL SUPPORT PREDICTS BETTER MENTAL HEALTH: KNOWLEDGE TRANSLATION AMONG OLDER ADULTS DURING COVID-19

The well-being of older adults has been linked to the quality of their neighbourhood environment. Given that COVID-19 affected poorer neighbourhoods disproportionately, we partnered with community organizations to identify meso-level psychosocial factors that may improve loneliness, depressive mood, and cognitive function. Five variables were identified through focus groups with older adults and community organizations. These variables were drawn from validated scales, including communal provisions, neighbourhood friendship, self-expression, social experiences, and time outdoors. This paper presents preliminary findings from surveys administered to 151 community-dwelling older adults across British Columbia and interviews in four neighbourhoods.Purposeful and snowball sampling were used to recruit older adults (age 55+) from community centres and neighbourhood houses. Online surveys measured the five meso-level psychosocial exposure variables. Outcome variables included an index of loneliness, depressive mood, self-rated memory, semantic fluency and delayed recall. Data was geocoded and aggregated by Forward Sortation Area. Regression and cross-level mediation analysis were conducted. Four neighbourhoods were selected from a 2x2 matrix of high and low neighbourhood deprivation (CANUE, 2016). Mental health was associated with better social experiences (B=.26, p=.003). Time outdoors (B=.35, p=.047) was associated with better delayed recall. Mental health was better in poorer neighbourhoods (B=.20, p=.015). This was partially mediated by communal provisions (B=.19, p=.032). Social experiences (B=.23, p=.009) fully mediated these effects on mental health. Participants described being of local community services and took on opportunities to volunteer. Social experiences and neighbourhood resources may help support mental health and well-being among older adults during the pandemic and beyond.

MULTISTAGE SAMPLING FOR TRANSLATIONAL COMMUNITY RESEARCH: ZOOMING OUT TO HONE IN ON PLACE-BASED HEALTH DISPARITIES

​​Whereas researchers strive for generalizability, community-engaged research (CEnR) typically involves only a few specific communities. Drawing on Weberian ideal type, I outline the use of an innovative blended-methods approach to sample the communities in which CEnR practitioners would collect in-depth data. To complement typical practices of entering a community without preconceived ideas, understanding how communities in the sampling frame relate to one another is important for equigenic (place-based health equity) implementations. The selection of neighborhood communities from quadrants in 2x2 matrices allows pertinent concepts to emerge and relevant solutions to be drawn from thriving communities to aid program co-creation and implementation in other communities. For example, this has led to the identification of communities in British Columbia with differing socioeconomic status, social capital, and coping during COVID-19. This methodological innovation is congruent with asset-based community development (ABCD) to minimize arbitrariness in sampling decisions and advance health equity in our cities.

Beyond COVID-19 Diagnosis: Prognosis with Hierarchical Graph Representation Learning

Coronavirus disease 2019 (COVID-19), the pandemic that is spreading fast globally, has caused over 181 million confirmed cases. Apart from the reverse transcription polymerase chain reaction (RT-PCR), the chest computed tomography (CT) is viewed as a standard and effective tool for disease diagnosis and progression monitoring. We propose a diagnosis and prognosis model based on graph convolutional networks (GCNs). The chest CT scan of a patient, typically involving hundreds of sectional images in a sequential order, is formulated as a densely connected weighted graph. A novel distance aware pooling is proposed to abstract the node information hierarchically, which is robust and efficient for such densely connected graphs. Our method, combining GCNs and distance aware pooling, can integrate the information from all slices in the chest CT scans for optimal decision making, which leads to the state-of-the-art accuracy in the COVID-19 diagnosis and prognosis. With less than 1% of the total number of parameters in the baseline 3D ResNet model, our method achieves 94.8% accuracy for diagnosis, which represents a 2.4% improvement over the baseline on the same dataset. In addition, we can localize the most informative slices with disease lesions for COVID-19 within a large sequence of chest CT images. The proposed model can produce visual explanations for the diagnosis and prognosis, making the decision more transparent and explainable, while RT-PCR only leads to the test result with no prognosis information. The prognosis analysis can help hospitals or clinical centers designate medical resources more efficiently and better support clinicians to determine the proper clinical treatment. © 2021, Springer Nature Switzerland AG.