Abstract for: Value of Anticipating Chaotic Post Pandemic Surges and Collapses of Depression in London Neighbourhoods

In March 2020 UK psychological depression prevalence surged from 10% to almost 30% in a matter of weeks, then it collapsed to 16%, surged again to 23% and has continued surging and collapsing in waves and remains on average at much higher levels than pre-pandemic. Long Covid is inducing further surges as each new wave of Covid infections passes. Depression is contagious and episodic causing the waves move around and intersect to produce complex, sustained chaotic eruptions and collapses that can vary massively between neighbourhoods right next to each other. The estimated potential UK Economic impact in the absence of mitigating actions could exceed £1 Trillion -- equivalent to a major Economic Depression for several years. This presentation shows how Geospatial Statistical, System Dynamics and Agent Based models working together to represent the epidemiology of depression, geospatial contagion and local environmental conditions in 30 neighbourhoods of London can be used to anticipate locations where eruptions are likely to happen and help heath services prepare proactive mitigating actions. The presentation will also discuss ongoing work to apply Machine Learning to automate and scale up for widespread use and generalise for multiple applications.