
Wastewater is increasingly being recognized as an important data stream that can contribute to infectious disease surveillance and forecasting. With this recognition, a growing number of statistical inference approaches are being developed to use wastewater data to provide quantitative insights into epidemiological dynamics. However, few existing approaches have allowed for systematic integration of data streams for inference, for example by combining case incidence data and/or serological data with wastewater data. Furthermore, only a subset of existing approaches have been able to handle missing data without imputation and to handle datasets with different sampling times or intervals. Here, we develop a statistically rigorous, yet lightweight, approach to infer and forecast time-varying effective reproduction numbers (Re values) using longitudinal wastewater virus concentrations either alone or jointly with additional data streams including case incidence data and serological data. Our approach relies on a state-space modeling approach for inference and forecasting, within the context of a simple bootstrap particle filter. We first describe the structure of our underlying disease transmission process model as well as our observation models. Using a mock dataset, we then show that Re can be accurately estimated by interfacing this model with case incidence data, wastewater data, or a combination of these two data streams using the bootstrap particle filter. Of note, we show that these data streams alone do not allow for reconstruction of underlying infection dynamics due to parameter unidentifiability. We then apply our particle filter to a previously analyzed SARS-CoV-2 dataset from Zurich that includes case data and wastewater data. Our analyses of these empirical datasets indicate that incorporation of process noise into the state-space model improves our ability to reconstruct the latent variables of the model. We further show that underlying infection dynamics can be made identifiable through the incorporation of serological data. Finally, we show how the bootstrap particle filter can be used to make forecasts of case incidence and wastewater virus concentrations, although we leave a comprehensive analysis of the additional predictive value of wastewater data to future work. We hope that the inference approach presented here will lead to greater reliance on wastewater data for disease surveillance and forecasting that will aid public health practitioners in responding to infectious disease threats.
Background Sexual partnerships are central to sexually transmitted infection (STI) transmission. Modeling studies and prevention programs often assume stable sexual activity over time, potentially misrepresenting transmission dynamics. We quantified longitudinal changes in sexual activity levels and associated factors among Canadian gay, bisexual, and other men who have sex with men (GBM). Methods The Engage Cohort Study recruited GBM using respondent-driven sampling in Montréal, Toronto, and Vancouver (Canada). Sexual activity was categorized into low (1st quartile), medium (2nd-3rd quartile), and high (4th quartile) levels based on the quartiles of the distribution of self-reported number of male sexual partners over the past six months, defined separately before/since the COVID-19 pandemic. Transition probabilities were estimated using a Bayesian continuous-time Markov chain model, jointly accounting for visit intensity and attrition. Results A total of 2001 participants with ≥2 visits (2017–2024) were included. Changes in sexual activity were frequent. Over one year, 33% (95% credible intervals [CrI]: 31–35%), 47% (CrI: 45–49%), and 41% (CrI: 39–44%) of participants with low, medium, and high activity changed levels. Overall, unregulated substance use showed the strongest association with increases in activity level (risk ratio [RR] 1.37–2.52 across directions). Similar patterns were observed for younger age and being in good health. Having a stable partner reduced the probability of transitioning outside of low-activity level (RR=0.80). Conclusion Changes in sexual activity among GBM are common and occur over short time periods. Considering these transitions is essential to better represent STI transmission, evaluate intervention impact, and design and deliver public health interventions.
Infectious disease modelling plays a critical role in understanding and addressing global health challenges. However, the field faces persistent barriers related to Equity, Diversity and Inclusion (EDI), including language and cultural biases, underrepresentation of certain groups and systemic inequities in access to resources and opportunities. This paper identifies key challenges and proposes actionable guidelines to foster inclusivity, collaboration and innovation in the field. Unlike generic EDI guidelines, we provide context-specific recommendations for three key settings in infectious disease research: team dynamics, conferences and virtual collaborations. We emphasise practical steps that can be taken such as diversifying conference organising committees, accommodating participants’ needs and integrating mentorship programmes. While recognising that EDI initiatives must be tailored to specific cultural and institutional settings, we highlight the importance of measurable progress through continuous reflection, assessment and adaptation. By embracing EDI principles, the field of infectious disease modelling can better harness the strengths of a diverse global community, leading to more equitable science, more representative models and improved health outcomes worldwide.
Tuberculosis remains a serious disease that persists around the world today because of ongoing transmission of Mycobacterium tuberculosis among cases and their contacts in the community. To curb the tuberculosis epidemic, new strategies are needed to reduce transmission in the community. Since quantifying exposures and transmission in a dynamic community is difficult, we propose that the mobility of a case during the infectious period may be a proxy measure of exposure and subsequent transmission. In a study performed in Kampala, Uganda, where the prevalence of tuberculosis disease was 750 cases/100,000 population, we collected cellular telephone call-detail records from 191 culture-confirmed tuberculosis cases and used a novel artificial intelligence algorithm, BERT4Traj, to estimate geographic trajectories of each case and characterized their mobility using metrics derived from the trajectories. We found that mobility of cases was both spatially and temporally heterogeneous. Some cases traced spatial patterns that bridged various locations and expanded their contact networks, whereas others traced regular movements that led to repeated contacts. This heterogeneity was partially explained by socio-demographic characteristics of the cases. The heterogeneity in the mobility of tuberculosis cases suggests that a single strategy to reduce transmission of tuberculosis may not suffice. Instead, future interventions should address the variable patterns of case mobility in the community.