Modelling the impact of COVID-19 in Australia to inform transmission reducing measures and health system preparedness

Robert Moss; James Wood; Damien Brown; Freya Shearer; Andrew J Black; Allen Cheng; James M McCaw; Jodie McVernon

Quantitative findings from our model are limited by ongoing uncertainties about the true disease 'pyramid' for COVID-19, and a lack of nuanced information about determinants of severe disease, here represented by age as a best proxy. The clinical pathways model assumes that half of available bed capacity is available for patients with the disease but does not anticipate the seasonal surge in influenza admissions that may be overlaid with the epidemic peak, although even in our most recent severe season (2017) only 6% of hospital beds were occupied by influenza cases (16) . Available beds will likely be increased by other factors such as secondary reductions in all respiratory infections and road trauma resulting from social restrictions, and purposive decisions to cancel non-essential surgery. Importantly, we do not consider health care worker absenteeism due to illness, carer responsibilities or burnout -all of which are anticipated challenges over a very prolonged epidemic accompanied by marked social disruption. We also cannot account for shortages in critical medical supplies as the true extent of these and their likely future impacts on service provision are presently unknown. 7 The model indicates that a combination of case-targeted and social measures will need to be applied over an extended period to reduce the rate of epidemic growth. In reality, it is likely that the stringency of imposed controls, their public acceptability and compliance will all vary over time. In Australia, compliance with isolation and self-quarantining was largely on the basis of trust in early response (February through March) but active monitoring and enforcement of these public health measures is now occurring in many jurisdictions. Hong Kong and Singapore initiated electronic monitoring technologies from the outset to track the location of individuals and enforce compliance (17). Proxy indicators of compliance such as transport and mobile phone data have informed understanding of the impact of social and movement restrictions on mobility and behaviour in other settings (11) , and will be further investigated in the Australian context.