[Preventive Medicine Reports, December 2022]

The long-term dynamics of COVID-19 disease incidence and public health measures may impact individuals’ precautionary behaviours as well as support for measures. The objectives of this study were to assess longitudinal changes in precautionary behaviours and support for public health measures. Survey data were collected online from 1030 Canadians in each of 5 cycles in 2020: June 15-July 13; July 22-Aug 8; Sept 7–15; Oct 14–21; and Nov 12–17. Precautionary behaviour increased over the study period in the context of increasing disease incidence. When controlling for the stringency of public health measures and disease incidence, mixed effects logistic regression models showed these behaviours did not significantly change over time. Odds ratios for avoiding contact with family and friends ranged from 0.84 (95% CI 0.59–1.20) in September to 1.25 (95% CI 0.66–2.37) in November compared with July 2020. Odds ratios for attending an indoor gathering ranged from 0.86 (95% CI 0.62–1.20) in August to 1.71 (95% CI 0.95–3.09) in October compared with July 2020. Support for non-essential business closures increased over time with 2.33 (95% CI 1.14–4.75) times higher odds of support in November compared to July 2020. Support for school closures declined over time with lower odds of support in September (OR 0.66 [95% CI 0.45–0.96]), October (OR 0.48 [95% CI 0.26–0.87]), and November (OR 0.39 [95% CI 0.19–0.81]) compared with July 2020. In summary, respondents’ behaviour mirrored government guidance between July and November 2020 and supported individual precautionary behaviour and limitations on non-essential businesses over school closures.

[Journal of Public Health Research, 12 May 2023]

Background: Public health surveillance data do not always capture all cases, due in part to test availability and health care seeking behaviour. Our study aimed to estimate under-ascertainment multipliers for each step in the reporting chain for COVID-19 in Toronto, Canada.

Design and methods: We applied stochastic modeling to estimate these proportions for the period from March 2020 (the beginning of the pandemic) through to May 23, 2020, and for three distinct windows with different laboratory testing criteria within this period.

Results: For each laboratory-confirmed symptomatic case reported to Toronto Public Health during the entire period, the estimated number of COVID-19 infections in the community was 18 (5th and 95th percentile: 12, 29). The factor most associated with under-reporting was the proportion of those who sought care that received a test.

Conclusions: Public health officials should use improved estimates to better understand the burden of COVID-19 and other similar infections.

[PLoS ONE, 23 February 2023]

Background: Antimicrobial Resistance (AMR) is a global problem with large health and economic consequences. Current gaps in quantitative data are a major limitation for creating models intended to simulate the drivers of AMR. As an intermediate step, expert knowledge and opinion could be utilized to fill gaps in knowledge for areas of the system where quantitative data does not yet exist or are hard to quantify. Therefore, the objective of this study was to identify quantifiable data about the current state of the factors that drive AMR and the strengths and directions of relationships between the factors from statements made by a group of experts from the One Health system that drives AMR development and transmission in a European context.

Methods: This study builds upon previous work that developed a causal loop diagram of AMR using input from two workshops conducted in 2019 in Sweden with experts within the European food system context. A secondary analysis of the workshop transcripts was conducted to identify semi-quantitative data to parameterize drivers in a model of AMR.

Main findings: Participants spoke about AMR by combining their personal experiences with professional expertise within their fields. The analysis of participants’ statements provided semi-quantitative data that can help inform a future of AMR emergence and transmission based on a causal loop diagram of AMR in a Swedish One Health system context.

Conclusion: Using transcripts of a workshop including participants with diverse expertise across the system that drives AMR, we gained invaluable insight into the past, current, and potential future states of the major drivers of AMR, particularly where quantitative data are lacking.

[Policy Insights from the Behavioral and Brain Sciences, 15 March 2023]

Psychosocial factors are related to immune, viral, and vaccination outcomes. Yet, this knowledge has been poorly represented in public health initiatives during the COVID-19 pandemic. This review provides an overview of biopsychosocial links relevant to COVID-19 outcomes by describing seminal evidence about these associations known prepandemic as well as contemporary research conducted during the pandemic. This focuses on the negative impact of the pandemic on psychosocial health and how this in turn has likely consequences for critically relevant viral and vaccination outcomes. We end by looking forward, highlighting the potential of psychosocial interventions that could be leveraged to support all people in navigating a postpandemic world and how a biopsychosocial approach to health could be incorporated into public health responses to future pandemics.

[Zoonoses and Public Health, 8 February 2024]

Introduction: Public health preparedness is based on timely and accurate information. Time series forecasting using disease surveillance data is an important aspect of preparedness. This study compared two approaches of time series forecasting: seasonal auto-regressive integrated moving average (SARIMA) modelling and the artificial neural network (ANN) algorithm. The goal was to model weekly seasonal influenza activity in Canada using SARIMA and compares its predictive accuracy, based on root mean square prediction error (RMSE) and mean absolute prediction error (MAE), to that of an ANN.

Methods: An initial SARIMA model was fit using automated model selection by minimizing the Akaike information criterion (AIC). Further inspection of the autocorrelation function and partial autocorrelation function led to ‘manual’ model improvements. ANNs were trained iteratively, using an automated process to minimize the RMSE and MAE.

Results: A total of 378, 462 cases of influenza was reported in Canada from the 2010–2011 influenza season to the end of the 2019–2020 influenza season, with an average yearly incidence risk of 20.02 per 100,000 population. Automated SARIMA modelling was the better method in terms of forecasting accuracy (per RMSE and MAE). However, the ANN correctly predicted the peak week of disease incidence while the other models did not.

Conclusion: Both the ANN and SARIMA models have shown to be capable tools in forecasting seasonal influenza activity in Canada. It was shown that applying both in tandem is beneficial, SARIMA better forecasted overall incidence while ANN correctly predicted the peak week.