COVID-19: New Zealand Outcome Modelling and Visualisation

Mortality

Hospitalisations

Introduction

What could happen if the COVID-19 epidemic is not contained?

The New Zealand government has adopted a strategy of eliminating the virus from the population. Should elimination not be achieved, the strategy is to mitigate the spread of the disease throughout the population.

This is not straightforward as recent overseas experience has demonstrated. Reports of the COVID-19 epidemic from overseas indicate that it is having different impacts on different populations, ages, and regions. If the epidemic is not controlled in New Zealand, we can expect similar differences in impacts for different populations based on current overseas experience and the high levels of unequal impact of previous New Zealand epidemics. The 1918 flu epidemic had a death rate for Māori that was 7.2 times that of non-Māori, while the 1956 epidemic it was 6.2 times that of non-Māori. As recently as 2009, this ratio was 2.6 times, but this was only discovered and documented well after the epidemic was over (Wilson et al, 2012a).

Inequitable outcomes by age, gender, region or ethnicity are not inevitable. Extra support focused on and working with Māori, Pacific and low-income communities can ensure that all those aged 60+ or with chronic conditions can be maximally protected if elimination fails. Preventing the high levels of inequity of previous epidemics will require flexible responses informed by the best available information on the potential spread and impact of this disease. This reduction in inequity will also significantly reduce the impact of this disease on the entire population.

This tool enables the modelling of a range of possible scenarios if elimination is not successful. It provides an easy to use way to model various outcomes at national and DHB level using the currently available knowledge about the disease. This information is intended to inform intervention at the early stages of the epidemic using low population infection rates. Using higher population infection rates can demonstrate the potential impact if both elimination and mitigation strategies were not successful and the epidemic ran its course in the NZ population.

The first outcomes to be modelled are hospitalisations, requiring intensive care use and mortality. These models use information on population age structure, previous NZ epidemic experience and published COVID-19 outcome information from overseas to model the variation and equity of potential outcomes. The tool uses default values based on the most recent documented experience.

This tool will be updated and extended as more information becomes available. This will include testing and confirmed cases – but that information is not currently available in a form that can be visualised.

This work sits alongside other forthcoming work on COVID-19 outcome equity being led by Te Pūnaha Matatini.

This work was undertaken by iNZight Analytics with funding from McDonaldSporle Ltd.

Contact: Andrew Sporle ( a.sporle -at- auckland.ac.nz )

Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...
Loading...

Authors

This interactive app was developed by Daniel Barnett and Andrew Sporle (funding from McDonaldSporle Ltd.)

Citing

To cite results from this software, use the following citation:

Barnett, D. and Sporle, A. (2020). New Zealand COVID-19 Outcome Modelling and Visualisation (version 2.5). Available from

Contact

Andrew Sporle ( a.sporle -at- auckland.ac.nz )

Acknowledgements

The authors would like to extend their many thanks to:

  • Professors Thomas Lumley and Nick Wilson, Associate-Professor Ricci Harris and Dr Melissa McLeod for their very helpful comments on earlier versions of this app. Any errors or omissions with this application remain the responsibility of the authors.

  • Professor Shaun Hendy, Kate Hannah, and Te Pūnaha Matatini for their help in making this application happen.

  • Martin Feller, Yvette Wharton, Marcus Gustafsson, Professor Mark Gahegan and the Centre for eResearch at the University of Auckland for their help in hosting and making this application publicly accessible.

This work was undertaken by iNZight Analytics with funding from McDonaldSporle Ltd.

References

Software