Comparison of machine learning methods for estimating case fatality ratios: An Ebola outbreak simulation study

Background Machine learning (ML) algorithms are now increasingly used in infectious disease epidemiology. Epidemiologists should understand how ML algorithms behave within the context of outbreak data where missingness of data is almost ubiquitous. Methods Using simulated data, we use a ML algorithmic framework to evaluate data imputation performance and the resulting case fatality ratio (CFR) estimates, focusing on the scale and type of data missingness (i.e., missing completely at random—MCAR, missing at random—MAR, or missing not at random—MNAR). Results Across ML methods, dataset sizes and proportions of training data used, the area under the receiver operating characteristic curve decreased by 7% (median, range: 1%–16%) when missingness was increased from 10% to 40%. Overall reduction in CFR bias for MAR across methods, proportion of missingness, outbreak size and proportion of training data was 0.5% (median, range: 0%–11%). Conclusion ML methods could reduce bias and increase the precision in CFR estimates at low levels of missingness. However, no method is robust to high percentages of missingness. Thus, a datacentric approach is recommended in outbreak settings—patient survival outcome data should be prioritised for collection and random-sample follow-ups should be implemented to ascertain missing outcomes.

A Delay Differential Equation approach to model the COVID-19 pandemic

SEIR (Susceptible - Exposed - Infected - Recovered) approach is a classic modeling method that has frequently been applied to the study of infectious disease epidemiology. However, in the vast majority of SEIR models and models derived from them transitions from one population group to another are described using the mass-action law which assumes population homogeneity. That causes some methodological limitations or even drawbacks, particularly inability to reproduce observable dynamics of key characteristics of infection such as, for example, the incubation period and progression of the disease's symptoms which require considering different time scales as well as probabilities of different disease trajectories. In this paper, we propose an alternative approach to simulate the epidemic dynamics that is based on a system of differential equations with time delays to precisely reproduce a duration of infectious processes (e.g. incubation period of the virus) and competing processes like transition from infected state to the hospitalization or recovery. The suggested modeling approach is fundamental and can be applied to the study of many infectious disease epidemiology. However, due to the urgency of the COVID-19 pandemic we have developed and calibrated the delay-based model of the epidemic in Germany and France using the BioUML platform. Additionally, the stringency index was used as a generalized characteristic of the non-pharmaceutical government interventions implemented in corresponding countries to contain the virus spread. The numerical analysis of the calibrated model demonstrates that adequate simulation of each new wave of the SARS-CoV-2 virus spread requires dynamic changes in the parameter values during the epidemic like reduction of the population adherence to non-pharmaceutical interventions or enhancement of the infectivity parameter caused by an emergence of novel virus strains with higher contagiousness than original one. Both models may be accessed and simulated at https://gitlab.sirius-web.org/covid-19/dde-epidemiology-model utilizing visual representation as well as Jupyter Notebook.

Demographic methods in epidemiology

Demography is both shaped by and shapes infectious disease dynamics. Infectious pathogens can increase host mortality. Host birth rates introduce new susceptible individuals into the population, which allows infections to persist in the face of the depletion of susceptible individuals that can result from mortality or immunity that can follow infection. Many important processes in infectious disease epidemiology, from transmission to vaccination, vary as a function of age or life stage. Epidemiology thus requires demographic methods. This chapter introduces broad expectations for patterns emerging from the intersection between demography and epidemiology and presents a set of structured population modelling tools that can be used to dissect important processes, including next generation methods, and estimation of R0 in the context of stage structure and with important differences in time-scale between host demography and pathogen life cycle.

Dynamical Modeling as a Tool for Inferring Causation

Dynamical models, commonly used in infectious disease epidemiology, are formal mathematical representations of time-changing systems or processes. For many chronic disease epidemiologists, the link between dynamical models and predominant causal inference paradigms is unclear. This commentary explains the use of dynamical models for representing causal systems and the relevance of dynamical models for causal inference. In certain simple settings, dynamical modeling and conventional statistical methods (e.g., regression-based methods) are equivalent, but dynamical modeling has advantages over conventional statistical methods for many causal inference problems. Dynamical models can be used to transparently encode complex biological knowledge, interference and spillover, effect modification, and variables that influence each other in continuous time. As our knowledge of biological and social systems and access to computational resources increases, there will be a growing utility for a variety of mathematical modeling tools in epidemiology.

Nine-month Trend of Time-Varying Reproduction Numbers of COVID-19 in West of Iran

Background: The basic reproduction number (R0) is an important concept in infectious disease epidemiology and the most important parameter to determine the transmissibility of a pathogen. This study aimed to estimate the nine-month trend of time-varying R of COVID-19 epidemic using the serial interval (SI) and Markov Chain Monte Carlo in Lorestan, west of Iran. Study design: Descriptive study. Methods: This study was conducted based on a cross-sectional method. The SI distribution was extracted from data and log-normal, Weibull, and Gamma models were fitted. The estimation of time-varying R0, a likelihood-based model was applied, which uses pairs of cases to estimate relative likelihood. Results: In this study, Rt was estimated for SI 7-day and 14-day time-lapses from 27 February-14 November 2020. To check the robustness of the R0 estimations, sensitivity analysis was performed using different SI distributions to estimate the reproduction number in 7-day and 14-day time-lapses. The R0 ranged from 0.56 to 4.97 and 0.76 to 2.47 for 7-day and 14-day time-lapses. The doubling time was estimated to be 75.51 days (95% CI: 70.41, 81.41). Conclusions: Low R0 of COVID-19 in some periods in Lorestan, west of Iran, could be an indication of preventive interventions, namely quarantine and isolation. To control the spread of the disease, the reproduction number should be reduced by decreasing the transmission and contact rates and shortening the infectious period.

The Saudi Critical Care Society Practice Guidelines on the Management of COVID-19 in the ICU: Therapy Section

BackgroundThe rapid increase in coronavirus disease 2019 (COVID-19) cases during the subsequent waves in Saudi Arabia and other countries prompted the Saudi Critical Care Society (SCCS) to put together a panel of experts to issue evidence-based recommendations for the management of COVID-19 in the intensive care unit (ICU).MethodsThe SCCS COVID-19 panel included 51 experts with expertise in critical care, respirology, infectious disease, epidemiology, emergency medicine, clinical pharmacy, nursing, respiratory therapy, methodology, and health policy. All members completed an electronic conflict of interest disclosure form. The panel addressed 9 questions that are related to the therapy of COVID-19 in the ICU. We identified relevant systematic reviews and clinical trials, then used the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach as well as the evidence-to-decision framework (EtD) to assess the quality of evidence and generate recommendations.ResultsThe SCCS COVID-19 panel issued 12 recommendations on pharmacotherapeutic interventions (immunomodulators, antiviral agents, and anticoagulants) for severe and critical COVID-19, of which 3 were strong recommendations and 9 were weak recommendations. ConclusionThe SCCS COVID-19 panel used the GRADE approach to formulate recommendations on therapy for COVID-19 in the ICU. The EtD framework allows adaptation of these recommendations in different contexts. The SCCS guideline committee will update recommendations as new evidence becomes available.

Implementing Mandatory Testing and a Public Health Commitment to Control COVID-19 on a College Campus

The COVID-19 pandemic has greatly impacted US colleges and universities. As The George Washington University (GWU), a large urban university, prepared to reopen for the Fall 2020 semester, GWU established protocols to protect the health and wellness of all members of campus community. Reopening efforts included a cadre of COVID-19 surveillance systems including development of a public health COVID-19 laboratory, weekly and symptomatic SARS-CoV-2 testing and daily risk screening and symptom monitoring. Other activities included completion of a mandatory COVID-19 training and influenza vaccination for the on-campus population, quarantining of students returning to campus, campus-focused case investigations and quarantining of suspected close contacts, clinical follow-up of infected persons, and regular communication and monitoring. A smaller on-campus population of 4,435 students, faculty and staff returned to campus with later expansion of testing to accommodate GWU students living in the surrounding area. Between August 17 and December 4, 2020, 38,288 tests were performed; 220 were positive. The surveillance program demonstrated a relatively low positivity rate, with temporal clustering of infected persons mirroring community spread, and little evidence for transmission among the GWU on-campus population. These efforts demonstrate the feasibility of safely partially reopening a large urban college campus by applying core principles of public health surveillance, infectious disease epidemiology, behavioral measures, and increased testing capacity, while continuing to promote educational and research opportunities. GWU will continue to monitor the program as the pandemic evolves and periodically reassess to determine if these strategies will be successful upon a full return to in-person learning.

Epidemic interventions: insights from classic results

Analytical expressions and approximations from simple models have performed a pivotal role in our understanding of infectious disease epidemiology. During the current COVID-19 pandemic, while there has been proliferation of increasingly complex models, still the most basic models have provided the core framework for our thinking and interpreting policy decisions. Here, classic results are presented that give insights into both the role of transmission-reducing interventions (such as social distancing) in controlling an emerging epidemic, and also what would happen if insufficient control is applied. Though these are simple results from the most basic of epidemic models, they give valuable benchmarks for comparison with the outputs of more complex modelling approaches. This article is part of the theme issue ‘Modelling that shaped the early COVID-19 pandemic response in the UK’.

The Epidemic of Mental Disorders in Business—How Depression, Anxiety, and Stress Spread across Organizations through Employee Mobility

Combining management research with infectious disease epidemiology, we propose a new perspective on mental disorders in a business context. We suggest that—similar to infectious diseases—clinical diagnoses of depression, anxiety, and stress-related disorders can spread epidemically across the boundaries of organizations via social contagion. We propose a framework for assessing the patterns of disease transmission, with employee mobility as the driver of contagion across organizations. We empirically test the proposed mental disorder transmission patterns by observing more than 250,000 employees and more than 17,000 Danish firms over a period of 12 years. Our findings reveal that when organizations hire employees from other, unhealthy organizations (those with a high prevalence of mental disorders), they “implant” depression, anxiety, and stress-related disorders into their workforces. Employees leaving unhealthy organizations act as “carriers” of these disorders regardless of whether they themselves have received a formal diagnosis of a mental disorder. The effect is especially pronounced if the newcomer holds a managerial position.

Heterogeneity in transmissibility and shedding SARS-CoV-2 via droplets and aerosols

Background: Which virological factors mediate overdispersion in the transmissibility of emerging viruses remains a longstanding question in infectious disease epidemiology.Methods: Here, we use systematic review to develop a comprehensive dataset of respiratory viral loads (rVLs) of SARS-CoV-2, SARS-CoV-1 and influenza A(H1N1)pdm09. We then comparatively meta-analyze the data and model individual infectiousness by shedding viable virus via respiratory droplets and aerosols.Results: The analyses indicate heterogeneity in rVL as an intrinsic virological factor facilitating greater overdispersion for SARS-CoV-2 in the COVID-19 pandemic than A(H1N1)pdm09 in the 2009 influenza pandemic. For COVID-19, case heterogeneity remains broad throughout the infectious period, including for pediatric and asymptomatic infections. Hence, many COVID-19 cases inherently present minimal transmission risk, whereas highly infectious individuals shed tens to thousands of SARS-CoV-2 virions/min via droplets and aerosols while breathing, talking and singing. Coughing increases the contagiousness, especially in close contact, of symptomatic cases relative to asymptomatic ones. Infectiousness tends to be elevated between 1-5 days post-symptom onset.Conclusions: Intrinsic case variation in rVL facilitates overdispersion in the transmissibility of emerging respiratory viruses. Our findings present considerations for disease control in the COVID-19 pandemic as well as future outbreaks of novel viruses.Funding: Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Grant program, NSERC Senior Industrial Research Chair program and the Toronto COVID-19 Action Fund.