[arXiv, 31 July 2023]

Stock and flow diagrams are widely used in epidemiology to model the dynamics of populations. Although tools already exist for building these diagrams and simulating the systems they describe, we have created a new package called StockFlow, part of the AlgebraicJulia ecosystem, which uses ideas from category theory to overcome notable limitations of existing software. Compositionality is provided by the theory of decorated cospans: stock and flow diagrams can be composed to form larger ones in an intuitive way formalized by the operad of undirected wiring diagrams. Our approach also cleanly separates the syntax of stock and flow diagrams from the semantics they can be assigned. We consider semantics in ordinary differential equations, although others are possible. As an example, we explain code in StockFlow that implements a simplified version of a COVID-19 model used in Canada.

[PLOS Computational Biology, 27 February 2023]

Transmission of many communicable diseases depends on proximity contacts among humans. Modeling the dynamics of proximity contacts can help determine whether an outbreak is likely to trigger an epidemic. While the advent of commodity mobile devices has eased the collection of proximity contact data, battery capacity and associated costs impose tradeoffs between the observation frequency and scanning duration used for contact detection. The choice of observation frequency should depend on the characteristics of a particular pathogen and accompanying disease. We downsampled data from five contact network studies, each measuring participant-participant contact every 5 minutes for durations of four or more weeks. These studies included a total of 284 participants and exhibited different community structures. We found that for epidemiological models employing high-resolution proximity data, both the observation method and observation frequency configured to collect proximity data impact the simulation results. This impact is subject to the population’s characteristics as well as pathogen infectiousness. By comparing the performance of two observation methods, we found that in most cases, half-hourly Bluetooth discovery for one minute can collect proximity data that allows agent-based transmission models to produce a reasonable estimation of the attack rate, but more frequent Bluetooth discovery is preferred to model individual infection risks or for highly transmissible pathogens. Our findings inform the empirical basis for guidelines to inform data collection that is both efficient and effective.

[Systems, 17 July 2023]

We detail a case study using a participatory modeling approach in the development and use of discrete-event simulations to identify intervention strategies aimed at reducing emergency department (ED) wait times in a Canadian health policy setting. A four-stage participatory modeling approach specifically adapted to the local policy environment was developed to engage stakeholders throughout the modeling processes. The participatory approach enabled a provincial team to engage a broad range of stakeholders to examine and identify the causes and solutions to lengthy ED wait times in the studied hospitals from a whole-system perspective. Each stage of the approach was demonstrated through its application in the case study. A novel and key feature of the participatory modeling approach was the development and use of a multi-criteria framework to identify and prioritize interventions to reduce ED wait times. We conclude with a discussion on lessons learned, which provide insights into future development and applications of participatory modeling methods to facilitate policy development and build multi-stakeholder consensus.

[Frontiers in Digital Health, 28 June 2023]

Background: The use of social media data provides an opportunity to complement traditional influenza and COVID-19 surveillance methods for the detection and control of outbreaks and informing public health interventions.

Objective: The first aim of this study is to investigate the degree to which Twitter users disclose health experiences related to influenza and COVID-19 that could be indicative of recent plausible influenza cases or symptomatic COVID-19 infections. Second, we seek to use the Twitter datasets to train and evaluate the classification performance of Bidirectional Encoder Representations from Transformers (BERT) and variant language models in the context of influenza and COVID-19 infection detection.

Methods: We constructed two Twitter datasets using a keyword-based filtering approach on English-language tweets collected from December 2016 to December 2022 in Saskatchewan, Canada. The influenza-related dataset comprised tweets filtered with influenza-related keywords from December 13, 2016, to March 17, 2018, while the COVID-19 dataset comprised tweets filtered with COVID-19 symptom-related keywords from January 1, 2020, to June 22, 2021. The Twitter datasets were cleaned, and each tweet was annotated by at least two annotators as to whether it suggested recent plausible influenza cases or symptomatic COVID-19 cases. We then assessed the classification performance of pre-trained transformer-based language models, including BERT-base, BERT-large, RoBERTa-base, RoBERT-large, BERTweet-base, BERTweet-covid-base, BERTweet-large, and COVID-Twitter-BERT (CT-BERT) models, on each dataset. To address the notable class imbalance, we experimented with both oversampling and undersampling methods.

Results: The influenza dataset had 1129 out of 6444 (17.5%) tweets annotated as suggesting recent plausible influenza cases. The COVID-19 dataset had 924 out of 11939 (7.7%) tweets annotated as inferring recent plausible COVID-19 cases. When compared against other language models on the COVID-19 dataset, CT-BERT performed the best, supporting the highest scores for recall (94.8%), F1(94.4%), and accuracy (94.6%). For the influenza dataset, BERTweet models exhibited better performance. Our results also showed that applying data balancing techniques such as oversampling or undersampling method did not lead to improved model performance.

Conclusions: Utilizing domain-specific language models for monitoring users’ health experiences related to influenza and COVID-19 on social media shows improved classification performance and has the potential to supplement real-time disease surveillance.

[Systems Research and Behavioral Science, 19 August 2022]

This study systematically reviews applications of three simulation approaches, that is, system dynamics model (SDM), agent-based model (ABM) and discrete event simulation (DES), and their hybrids in COVID-19 research and identifies theoretical and application innovations in public health. Among the 372 eligible papers, 72 focused on COVID-19 transmission dynamics, 204 evaluated both pharmaceutical and non-pharmaceutical interventions, 29 focused on the prediction of the pandemic and 67 investigated the impacts of COVID-19. ABM was used in 275 papers, followed by 54 SDM papers, 32 DES papers and 11 hybrid model papers. Evaluation and design of intervention scenarios are the most widely addressed area accounting for 55% of the four main categories, that is, the transmission of COVID-19, prediction of the pandemic, evaluation and design of intervention scenarios and societal impact assessment. The complexities in impact evaluation and intervention design demand hybrid simulation models that can simultaneously capture micro and macro aspects of the socio-economic systems involved.

[Journal of Simulation, 20 March 2023]

In this study, we present a hybrid agent-based model (ABM) and discrete event simulation (DES) framework where ABM captures the spread dynamics of COVID-19 via asymptomatic passengers and DES captures the impacts of environmental variables, such as service process capacity, on the results of different containment measures in a typical high-speed train station in China. The containment and control measures simulated include as-is (nothing changed) passenger flow control, enforcing social distancing, adherence level in face mask-wearing, and adding capacity to current service stations. These measures are evaluated individually and then jointly under a different initial number of asymptomatic passengers. The results show how some measures can consolidate the outcomes for each other, while combinations of certain measures could compromise the outcomes for one or the other due to unbalanced service process configurations. The hybrid ABM and DES models offer a useful multi-function simulation tool to help inform decision/policy makers of intervention designs and implementations for addressing issues like public health emergencies and emergency evacuations. Challenges still exist for the hybrid model due to the limited availability of simulation platforms, extensive consumption of computing resources, and difficulties in validation and optimisation.

[Journal of Cancer Education, 3 November 2023]

Health systems are interested in increasing colorectal cancer (CRC) screening rates as CRC is a leading cause of preventable cancer death. Learning health systems are ones that use data to continually improve care. Data can and should include qualitative local perspectives to improve patient and provider education and care. This study sought to understand local perspectives on CRC screening to inform future strategies to increase screening rates across our integrated health system. Health insurance plan members who were eligible for CRC screening were invited to participate in semi-structured phone interviews. Qualitative content analysis was conducted using an inductive approach. Forty member interviews were completed and analyzed. Identified barriers included ambivalence about screening options (e.g., “If it had the same performance, I’d rather do home fecal sample test. But I’m just too skeptical [so I do the colonoscopy].”), negative prior CRC screening experiences, and competing priorities. Identified facilitators included a positive general attitude towards health (e.g., “I’m a rule follower. There are certain things I’ll bend rules. But certain medical things, you just got to do.”), social support, a perceived risk of developing CRC, and positive prior CRC screening experiences. Study findings were used by the health system leaders to inform the selection of CRC screening outreach and education strategies to be tested in a future simulation model. For example, the identified barrier related to ambivalence about screening options led to a proposed revision of outreach materials that describe screening types more clearly.

[Mathematics of Public Health, 7 August 2023]

Agent-based modeling is a computational dynamic modeling technique that may be less familiar to some readers. Agent-based modeling seeks to understand the behavior of complex systems by situating agents in an environment and studying the emergent outcomes of agent-agent and agent-environment interactions. In comparison with compartmental models, agent-based models offer simpler, more scalable, and flexible representation of heterogeneity, the ability to capture dynamic and static network and spatial context, and the ability to consider history of individuals within the model. In contrast, compartmental models offer faster development time with less programming required, lower computational requirements that do not scale with population, and the option for concise mathematical formulation with ordinary, delay, or stochastic differential equations supporting derivation of properties of the system behavior.

Citerank: (15) 679712CANMOD – PeopleCANMOD is a national network, with members located across the country and associated with a broader Emerging Infectious Disease Modelling (EIDM) initiative. We are a community of modellers, statisticians, epidemiologists, public health decision-makers, and those implementing and delivering interventions.10019D3ABAB, 679752Amy HurfordAmy Hurford is an Associate Professor jointly appointed in the Department of Biology and the Department of Mathematics and Statistics at Memorial University of Newfoundland and Labrador. 10019D3ABAB, 679757Beate SanderCanada Research Chair in Economics of Infectious Diseases and Director, Health Modeling & Health Economics and Population Health Economics Research at THETA (Toronto Health Economics and Technology Assessment Collaborative).10019D3ABAB, 679761Caroline ColijnDr. Caroline Colijn works at the interface of mathematics, evolution, infection and public health, and leads the MAGPIE research group. She joined SFU's Mathematics Department in 2018 as a Canada 150 Research Chair in Mathematics for Infection, Evolution and Public Health. She has broad interests in applications of mathematics to questions in evolution and public health, and was a founding member of Imperial College London's Centre for the Mathematics of Precision Healthcare.10019D3ABAB, 679775David BuckeridgeDavid is a Professor in the School of Population and Global Health at McGill University, where he directs the Surveillance Lab, an interdisciplinary group that develops, implements, and evaluates novel computational methods for population health surveillance. He is also the Chief Digital Health Officer at the McGill University Health Center where he directs strategy on digital transformation and analytics and he is an Associate Member with the Montreal Institute for Learning Algorithms (Mila).10019D3ABAB, 679776David EarnProfessor of Mathematics and Faculty of Science Research Chair in Mathematical Epidemiology at McMaster University.10019D3ABAB, 679844Mathieu Maheu-GirouxCanada Research Chair (Tier 2) in Population Health Modeling and Associate Professor, McGill University.10019D3ABAB, 679856Naveed Zafar JanjuaDr. Naveed Zafar Janjua is an epidemiologist and senior scientist at the BC Centre for Disease Control and Clinical Associate Professor at School of Population and Public Health, University of British Columbia. Dr. Janjua is a Medical Doctor (MBBS) with a Masters of Science (MSc) degree in Epidemiology & Biostatistics and Doctorate in Public Health (DrPH). 10019D3ABAB, 679880Sharmistha MishraSharmistha Mishra is an infectious disease physician and mathematical modeler and holds a Tier 2 Canadian Research Chair in Mathematical Modeling and Program Science.10019D3ABAB, 679893Kumar MurtyProfessor Kumar Murty is in the Department of Mathematics at the University of Toronto. His research fields are Analytic Number Theory, Algebraic Number Theory, Arithmetic Algebraic Geometry and Information Security. He is the founder of the GANITA lab, co-founder of Prata Technologies and PerfectCloud. His interest in mathematics ranges from the pure study of the subject to its applications in data and information security.10019D3ABAB, 685387Michael Y LiProfessor of Mathematics in the Department of Mathematical and Statistical Sciences at the University of Alberta, and Director of the Information Research Lab (IRL).10019D3ABAB, 701020CANMOD – PublicationsPublications by CANMOD Members144B5ACA0, 701037MfPH – Publications144B5ACA0, 704045Covid-19859FDEF6, 728553Pandemic modeling questions?140D5CB99

[Canadian Journal of Public Health, 25 July 2024]

Setting: Mathematical modelling played an important role in the public health response to COVID-19 in Canada. Variability in epidemic trajectories, modelling approaches, and data infrastructure across provinces provides a unique opportunity to understand the factors that shaped modelling strategies.

Intervention: Provinces implemented stringent pandemic interventions to mitigate SARS-CoV-2 transmission, considering evidence from epidemic models. This study aimed to summarize provincial COVID-19 modelling efforts. We identified modelling teams working with provincial decision-makers, through referrals and membership in Canadian modelling networks. Information on models, data sources, and knowledge translation were abstracted using standardized instruments.

Outcomes: We obtained information from six provinces. For provinces with sustained community transmission, initial modelling efforts focused on projecting epidemic trajectories and healthcare demands, and evaluating impacts of proposed interventions. In provinces with low community transmission, models emphasized quantifying importation risks. Most of the models were compartmental and deterministic, with projection horizons of a few weeks. Models were updated regularly or replaced by new ones, adapting to changing local epidemic dynamics, pathogen characteristics, vaccines, and requests from public health. Surveillance datasets for cases, hospitalizations and deaths, and serological studies were the main data sources for model calibration. Access to data for modelling and the structure for knowledge translation differed markedly between provinces.

Implication: Provincial modelling efforts during the COVID-19 pandemic were tailored to local contexts and modulated by available resources. Strengthening Canadian modelling capacity, developing and sustaining collaborations between modellers and governments, and ensuring earlier access to linked and timely surveillance data could help improve pandemic preparedness.