This report presents advice delivered to the NZ Ministry of Health in December 2022. It was prepared in response to requests for modelling advice around how COVID-19 case numbers and hospitalisations might be affected by policy changes enacted at different dates in the future, as well as a new variant of concern becoming dominant in November.
The initial report was delivered to the NZ Ministry of Health on 1st December 2022. It was revised following feedback from an independent peer review panel within COVID-19 Modelling Aotearoa. This version was finalised 26th June 2023.
This report explains how we use COVID-19 Modelling Aotearoa’s (CMAs) Network Contagion Model (NCM) and Ordinary Differential Equation (ODE) model to estimate the impacts of different scenarios of policy induced behaviour change, on future infections, case numbers, and hospitalisations. The behaviour changes we consider include symptomatic testing, case isolation behaviour, household contact testing and quarantine, and general population transmission reduction behaviours.
In order to model how the effect of future changes in policy induced public behaviour change, we needed first to infer how transmission reduction behaviour had changed since the September removal of the Covid Protection Framework (CPF).
Initially we use the ODE model to infer the change in transmission rate in Aotearoa since the removal of the Covid Protection Framework at midnight on 12th September 2022 up to the time of writing (late November 2022). These results indicate that the transmission increase resulting from the September policy change was around 19%; much higher than the previous estimate of 8.5% from CMA in August 2022.
With reference to previous NCM results, we then discuss possible changes in community, contact, and case behaviour that could have produced this inferred change in transmission rate.
This leads us to an estimate the impact of the complete removal of case isolation as being in the range of a further 5% to 15% increase in transmission.
We then use the ODE model to illustrate the potential impact of five scenarios for future policy changes (varying between a 0% and 10% increase in Rt) occurring in either December 2022, January 2023 or April 2023, all with the addition of a new variant of concern (VOC) becoming the main variant on 22 November 2022.
As expected, larger relaxations in behaviour lead to larger increases in transmission in the ODE results. Another key conclusion from the results is that if background transmission is increasing, and policies are changed that result in an increase in transmission rate, then we see higher peaks of infections and hospitalisations.
Throughout this report we note that it is very difficult to determine what behaviour changes drove the observed change in transmission rate following September 2022 policy changes, and therefore how much transmission may change with the removal of case isolation. We can propose likely bundles of behaviour change that could lead to the effect size that we have observed up to the time of writing. But without knowing ‘where we are’ it is even harder to model ‘where we will go’ when a new policy change is enacted.
Some key unknowns are:
- The infection rate currently, and how the case ascertainment rate (CAR) has changed over time.
- Exactly how the transmission rate has changed (although we estimate this with the ODE).
- How different transmission reduction behaviours have changed since September 2022.
Some ways that these information gaps could be addressed in future are:
- A repeated seroprevalence survey
- A repeated survey on transmission reduction behaviours
Related reports:
- Estimating the effect of Covid Protection Framework policy scenarios on the effective reproduction number of COVID-19 in Aotearoa: August 2022
- Estimating the effect of changes in case isolation on the effective reproduction number of COVID-19 in Aotearoa: September 2022
- Addendum: Assumption of perfect case isolation within the home when estimating the effect of a change in case isolation policy