Poster summary

The Stress and Recovery in Frontline Healthcare COVID-19 workers study was investigating wellbeing, mood, stress, sleep and activity in healthcare workers involved in frontline care during the COVID-19 pandemic. Myself and my colleagues at Cambridge Cognition focused on one aspect of this large study, aiming to establish whether discrete clusters of mental health exist in this population. We also aimed to investigate differences in temporal dynamics of mood, cognition and wearable sensor data.

Data was collected through an Oura smart ring, mood assessments on a smartphone and via Cambridge Cognition and Cognition Kit N-back, which measures working memory and attention. We also had baseline mental and physical health data from questionnaires such as PROMIS, PHQ-9 and GAD. Participants were grouped based on health through unsupervised machine learning. Patterns were also identified across timeseries variables and a general linear model was constructed to compare cluster Vector Autoregressive (VAR) coefficients. 

Our results show that healthcare workers on the COVID-19 frontline reached thresholds for anxiety and depression. Time-series analysis showed individual differences in stress, mood, cognition, sleep and activity.

The findings also show that Vector Autoregressive models can be applied to integrate sensor data, mood and cognitive measures and shows promise for use as an endpoint when studying people affected by mood disorders.

View the poster