Improving estimates of epidemiological quantities by combining reported cases with wastewater data: a statistical framework with applications to COVID-19 in Aotearoa New Zealand (preprint)

Background: Timely and informed public health responses to infectious diseases such as COVID-19 necessitate reliable information about infection dynamics. The case ascertainment rate (CAR), the proportion of infections that are reported as cases, is typically much less than one and varies with testing practices and behaviours, making reported cases unreliable as the sole source of data. The concentration of viral RNA in wastewater samples provides an alternate measure of infection prevalence that is not affected by human behaviours. Here, we investigated how these two data sources can be combined to inform estimates of the instantaneous reproduction number, R, and track changes in the CAR over time. Methods: We constructed a state-space model that we solved using sequential Monte Carlo methods. The observed data are the levels of SARS-CoV-2 in wastewater and reported case incidence. The hidden states that we estimate are R and CAR. Model parameters are estimated using the particle marginal Metropolis Hastings algorithm. Findings: We analysed data from 1 January 2022 to 31 March 2023 from Aotearoa New Zealand. Our model estimates that R peaked at 2.76 (95% CrI 2.20, 3.83) around 18 February 2022 and the CAR peaked around 12 March 2022. Accounting for reduced CAR, we estimate that New Zealand's second Omicron wave in July 2022 was similar in size to the first, despite fewer reported cases. We estimate that the CAR in the BA.5 Omicron wave in July 2022 was approximately 50% lower than in the BA.1/BA.2 Omicron wave in March 2022. The CAR in subsequent waves around November 2022 and April 2023 was estimated to be comparable to that in the second Omicron wave. Interpretation: This work on wastewater-based epidemiology (WBE) can be used to give insight into key epidemiological quantities. Estimating R, CAR, and cumulative number of infections provides useful information for planning public health responses and understanding the state of immunity in the population. This model is a useful disease surveillance tool, improving situational awareness of infectious disease dynamics in real-time, which may be increasingly useful as intensive pandemic surveillance programmes are wound down.

Maori and Pacific People in New Zealand have higher risk of hospitalisation for COVID-19 (preprint)

Aims. We aim to quantify differences in clinical outcomes from COVID-19 infection by ethnicity with a focus on risk of hospitalisation. Methods. We used data on age, ethnicity, deprivation index, pre-existing health conditions, and clinical outcomes on 1,829 COVID-19 cases reported in New Zealand. We used a logistic regression model to calculate odds ratios for the risk of hospitalisation by ethnicity. We also consider length of hospital stay and risk of fatality. Results. M[a]ori have 2.5 times greater odds of hospitalisation than non-M[a]ori, non-Pacific people, after controlling for age and pre-existing conditions. Similarly, Pacific people have 3 times greater odds. Conclusions. Structural inequities and systemic racism in the healthcare system mean that M[a]ori and Pacific communities face a much greater health burden from COVID-19. Older people and those with pre-existing health conditions are also at greater risk. This should inform future policy decisions including prioritising groups for vaccination.

Early intervention is the key to success in COVID-19 control (preprint)

New Zealand responded to the COVID-19 pandemic with a combination of border restrictions and an Alert Level system that included strict stay-at-home orders. These interventions were successful in containing the outbreak and ultimately eliminating community transmission of COVID-19. The timing of interventions is crucial to their success. Delaying interventions for too long may both reduce their effectiveness and mean that they need to be maintained for a longer period of time. Here, we use a stochastic branching process model of COVID-19 transmission and control to simulate the epidemic trajectory in New Zealand and the effect of its interventions during its COVID-19 outbreak in March-April 2020. We use the model to calculate key outcomes, including the peak load on the contact tracing system, the total number of reported COVID-19 cases and deaths, and the probability of elimination within a specified time frame. We investigate the sensitivity of these outcomes to variations in the timing of the interventions. We find that a delay to the introduction of Alert Level 4 controls results in considerably worse outcomes. Changes in the timing of border measures have a smaller effect. We conclude that the rapid response in introducing stay-at-home orders was crucial in reducing the number of cases and deaths and increasing the probability of elimination.

Potential reduction in transmission of COVID-19 by digital contact tracing systems (preprint)

Digital tools are being developed to support contact tracing as part of the global effort to control the spread of COVID-19. These include smartphone apps, Bluetooth-based proximity detection, location tracking, and automatic exposure notification features. Evidence on the effectiveness of alternative approaches to digital contact tracing is so far limited. We use an age-structured branching process model of the transmission of COVID-19 in different settings to estimate the potential of manual contact tracing and digital tracing systems to help control the epidemic. We investigate the effect of the uptake rate and proportion of contacts recorded by the digital system on key model outputs: the effective reproduction number, the mean outbreak size after 30 days, and the probability of elimination. We show that effective manual contact tracing can reduce the effective reproduction number from 2.4 to around 1.5. The addition of a digital tracing system with a high uptake rate over 75% could further reduce the effective reproduction number to around 1.1. Fully automated digital tracing without manual contact tracing is predicted to be much less effective. We conclude that, for digital tracing systems to make a significant contribution to the control of COVID-19, they need be designed in close conjunction with public health agencies to support and complement manual contact tracing by trained professionals.

Effective reproduction number for COVID-19 in Aotearoa New Zealand (preprint)

The effective reproduction number, Reff, is the average number of secondary cases infected by a primary case, a key measure of the transmission potential for a disease. Compared to many countries, New Zealand has had relatively few COVID-19 cases, many of which were caused by infections acquired overseas. This makes it difficult to use standard methods to estimate Reff. In this work, we use a stochastic model to simulate COVID-19 spread in New Zealand and report the values of Reff from simulations that gave best fit to case data. We estimate that New Zealand had an effective reproduction number Reff = 1.8 for COVID-19 transmission prior to moving into Alert Level 4 on March 25 2020 and that after moving into Alert level 4 this was reduced to Reff = 0.35. Our estimate Reff = 1.8 for reproduction number before Alert Level 4, is relatively low compared to other countries. This could be due, in part, to measures put in place in early- to mid-March, including: the cancellation of mass gatherings, the isolation of international arrivals, and employees being encouraged to work from home.

Probability of elimination for COVID-19 in Aotearoa New Zealand (preprint)

On 25th March 2020, New Zealand implemented stringent lockdown measures (Alert Level 4, in a four-level alert system) with the goal of eliminating community transmission of COVID-19. Once new cases are no longer detected over consecutive days, the probability of elimination is an important measure for informing decisions on when certain COVID-19 restrictions should be relaxed. Our model of COVID-19 spread in New Zealand estimates that after 2-3 weeks of no new reported cases, there is a 95% probability that COVID-19 has been eliminated. We assessed the sensitivity of this estimate to varying model parameters, in particular to different likelihoods of detection of clinical cases and different levels of control effectiveness. Under an optimistic scenario with high detection of clinical cases, a 95% probability of elimination is achieved after 10 consecutive days with no new reported cases, while under a more pessimistic scenario with low case detection it is achieved after 22 days.

The effect of border controls on the risk of COVID-19 reincursion from international arrivals (preprint)

In an attempt to maintain elimination of COVID-19, the New Zealand government has closed the border to everyone except citizens and residents. All arrivals are required to spend 14 days in government-managed isolation/quarantine and to be tested for COVID-19 on day 3 and on day 12 of their stay. We model the testing, isolation and potential transmission of COVID-19 within managed isolation facilities to estimate the risk of undetected cases and the risk of infectious cases being released into the community. We use a stochastic individual-based that includes a time-dependent probability of a false negative test result, complete isolation of confirmed and probable cases, and secondary transmission of COVID-19 between close contacts. We show that the combination of 14-day quarantine with day 3 and day 12 testing reduces risk of releasing an infectious case to around 0.1% per infected arrival. Shorter quarantine periods, or reliance on testing only with no quarantine, substantially increase this risk. It is important to avoid contacts between individuals staying in quarantine to minimise the risk of secondary transmission. We calculate the ratio of cases detected on day 3 to cases detected on day 12 in the model and show that this may be a useful indicator of the likelihood of secondary transmission occurring within quarantine. We do not explicitly model transmission of COVID-19 from individuals in quarantine to staff, but this is likely to present a significant risk. This needs to be minimised by strict infection control, use of personal protective equipment by staff at all times, and avoiding close contact between staff and hotel guests.

Successful contact tracing systems for COVID-19 rely on effective quarantine and isolation (preprint)

Background. Test, trace and isolate are the three crucial components of the response to COVID-19 identified by the World Health Organisation. Mathematical models of contact tracing often over-simplify the ability of traced contacts to quarantine or isolate. Method. We use an age-structured branching process model of individual disease transmission combined with a detailed model of symptom onset, testing, contact quarantine and case isolation to model each aspect of the test, trace, isolate strategy. We estimated the effective reproduction number under a range of scenarios to understand the importance of each aspect of the system. Findings. People's ability to quarantine and isolate effectively is a crucial component of a successful contact tracing system. 80% of cases need to be quarantined or isolated within 4 days of quarantine or isolation of index case to be confident the contact tracing system is effective. Interpretation. Provision of universal support systems to enable people to quarantine and isolate effectively, coupled with investment in trained public health professionals to undertake contact tracing, are crucial to success. We predict that a high-quality, rapid contact tracing system with strong support structures in place, combined with moderate social distancing measures, is required to contain the spread of COVID-19. Funding. This work was funded by the Ministry of Business, Innovation and Employment, New Zealand and Te Punaha Matatini, the NZ Centre of Research Excellence for Complex Systems.

A structured model for COVID-19 spread: modelling age and healthcare inequities (preprint)

We use a stochastic branching process model, structured by age and level of healthcare access, to look at the heterogeneous spread of COVID-19 within a population. We examine the effect of control scenarios targeted at particular groups, such as school closures or social distancing by older people. Although we currently lack detailed empirical data about contact and infection rates between age groups and groups with different levels of healthcare access within New Zealand, these scenarios illustrate how such evidence could be used to inform specific interventions. We find that an increase in the transmission rates amongst children from reopening schools is unlikely to significantly increase the number of cases, unless this is accompanied by a change in adult behaviour. We also find that there is a risk of undetected outbreaks occurring in communities that have low access to healthcare and that are socially isolated from more privileged communities. The greater the degree of inequity and extent of social segregation, the longer it will take before any outbreaks are detected. Well-established evidence for health inequities, particularly in accessing primary healthcare and testing, indicates that M[a]ori and Pacific peoples are at higher risk of undetected outbreaks in Aotearoa New Zealand. This highlights the importance of ensuring that community needs for access to healthcare, including early proactive testing, rapid contact tracing, and the ability to isolate, are being met equitably. Finally, these scenarios illustrate how information concerning contact and infection rates across different demographic groups may be useful in informing specific policy interventions.

Effect of Alert Level 4 on effective reproduction number: review of international COVID-19 cases (preprint)

The effective reproduction number, Reff, is an important measure of transmission potential in the modelling of epidemics. It measures the average number of people that will be infected by a single contagious individual. A value of Reff > 1 suggests that an outbreak will occur, while Reff< 1 suggests the virus will die out. In response to the COVID-19 pandemic, countries worldwide are implementing a range of intervention measures, such as population-wide social distancing and case isolation, with the goal of reducing Reff to values below one, to slow or eliminate transmission. We analyse case data from 25 international locations to estimate their Reff values over time and to assess the effectiveness of interventions, equivalent to New Zealand's Alert Levels 1-4, for reducing transmission. Our results show that strong interventions, equivalent to NZ's Alert Level 3 or 4, have been successful at reducing Reff below the threshold for outbreak. In general, countries that implemented strong interventions earlier in their outbreak have managed to maintain case numbers at lower levels. These estimates provide indicative ranges of Reff for each Alert Level, to inform parameters in models of COVID-19 spread under different intervention scenarios in New Zealand and worldwide. Predictions from such models are important for informing policy and decisions on intervention timing and stringency during the pandemic.

Estimated inequities in COVID-19 infection fatality rates by ethnicity for Aotearoa New Zealand (preprint)

There is limited evidence as to how COVID-19 infection fatality rates (IFR) may vary by ethnicity. We combine demographic and health data for ethnic groupings in Aotearoa New Zealand with international data on IFR for different age groups to estimate inequities in IFR by ethnicity. We find that, if age is the dominant factor determining IFR, estimated IFR for M[a]ori is around 50% higher than non-M[a]ori. If underlying health conditions are more important than age per se, then estimated IFR for M[a]ori is more than 2.5 times that of New Zealand European, and estimated IFR for Pasifika is almost double that of New Zealand European. IFRs for M[a]ori and Pasifika are likely to be increased above these estimates by racism within the healthcare system and other inequities not reflected in official data. IFR does not account for differences among ethnicities in COVID-19 incidence, which could be higher in M[a]ori and Pasifika as a result of crowded housing and higher inter-generational contact rates. These factors should be included in future disease incidence modelling. The communities at the highest risk will be those with elderly populations, and M[a]ori and Pasifika communities, where the compounded effects of underlying health conditions, socioeconomic disadvantage, and structural racism result in imbricated risk of contracting COVID-19, becoming unwell, and death.

A stochastic model for COVID-19 spread and the effects of Alert Level 4 in Aotearoa New Zealand (preprint)

While case numbers remain low, population-wide control methods combined with efficient tracing, testing, and case isolation, offer the opportunity for New Zealand to contain and eliminate COVID-19. We use a stochastic model to investigate containment and elimination scenarios for COVID-19 in New Zealand, as the country considers the exit from its four week period of strong Level 4 population-wide control measures. In particular we consider how the effectiveness of its case isolation operations influence the outcome of lifting these strong population-wide controls. The model is parameterised for New Zealand and is initialised using current case data, although we do not make use of information concerning the geographic dispersion of cases and the model is not stratified for age or co-morbidities. We find that fast tracing and case isolation (i.e. operations that are sustained at rates comparable to that at the early stages of New Zealand's response) can lead to containment or elimination, as long as strong population-wide controls remain in place. Slow case isolation can lead to containment (but not elimination) as long as strong Level 4 population-wide controls remain in place. However, we find that relaxing strong population-wide controls after four weeks will most likely lead to a further outbreak, although the speed of growth of this outbreak can be reduced by fast case isolation, by tracing, testing, or otherwise. We find that elimination is only likely if case isolation is combined with strong population-wide controls that are maintained for longer than four weeks. Further versions of this model will include an age-structured population as well as considering the effects of geographic dispersion and contact network structure, the possibility of regional containment combined with inter-regional travel restrictions, and the potential for harm to at risk communities and essential workers.

Suppression and Mitigation Strategies for Control of COVID-19 in New Zealand (preprint)

A standard SEIR-type compartment model, parameterised for New Zealand, was used to simulate the spread of Covid19 in New Zealand and to test the effectiveness of various control strategies. Control aims can be broadly categorised as either suppression or mitigation. Suppression aims to keep cases to an absolute minimum for as long as possible. Mitigation aims to allow a controlled outbreak to occur, with the aim of preventing significant overloads on healthcare systems and gradually allowing the population to develop herd immunity. Both types of strategy are fraught with uncertainty. Suppression strategies can succeed in delaying an outbreak, but only for as long as such control measures can be sustained. Once controls are eased or restricted, an epidemic is likely to follow as no herd immunity has been acquired. The success or failure of mitigation strategies can depend sensitively on the timing and efficacy of control measures, and require the ability to bring rapidly growing outbreaks under immediate control when needed. This is as yet untested even for a combination of national interventions including case isolation, household quarantine, population-wide social distancing and closure of schools and universities. Although there are disadvantages to both types of approach, suppression has the advantage of buying time until a vaccine and/or treatment become available and allowing NZ to learn from rapidly unfolding events in other countries. A combination of successful suppression, strong border measures, and widespread contact tracing and testing resulting in containment could allow periods when control measures can be relaxed, but only if cases are reduced to a handful.