Modelling during an emergency: the 2009 H1N1 influenza pandemic

Framework to guide the use of mathematical modelling in evidence-based policy decision-making

INTRODUCTION

The COVID-19 pandemic highlighted the significance of mathematical modelling in decision-making and the limited capacity in many low-income and middle-income countries (LMICs). Thus, we studied how modelling supported policy decision-making processes in LMICs during the pandemic (details in a separate paper).We found that strong researcher-policymaker relationships and co-creation facilitated knowledge translation, while scepticism, political pressures and demand for quick outputs were barriers. We also noted that routine use of modelled evidence for decision-making requires sustained funding, capacity building for policy-facing modelling, robust data infrastructure and dedicated knowledge translation mechanisms.These lessons helped us co-create a framework and policy roadmap for improving the routine use of modelling evidence in public health decision-making. This communication paper describes the framework components and provides an implementation approach and evidence for the recommendations. The components include (1) funding, (2) capacity building, (3) data infrastructure, (4) knowledge translation platforms and (5) a culture of evidence use.

KEY ARGUMENTS

Our framework integrates the supply (modellers) and demand (policymakers) sides and contextual factors that enable change. It is designed to be generic and disease-agnostic for any policy decision-making that modelling could support. It is not a decision-making tool but a guiding framework to help build capacity for evidence-based policy decision-making. The target audience is modellers and policymakers, but it could include other partners and implementers in public health decision-making.

CONCLUSION

The framework was created through engagements with policymakers and researchers and reflects their real-life experiences during the COVID-19 pandemic. Its purpose is to guide stakeholders, especially in lower-resourced settings, in building modelling capacity, prioritising efforts and creating an enabling environment for using models as part of the evidence base to inform public health decision-making. To validate its robustness and impact, further work is needed to implement and evaluate this framework in diverse settings.

Infection Studies with Airway Organoids from Carollia perspicillata Indicate That the Respiratory Epithelium Is Not a Barrier for Interspecies Transmission of Influenza Viruses

Bats are a natural reservoir for many viruses and are considered to play an important role in the interspecies transmission of viruses. To analyze the susceptibility of bat airway cells to infection by viruses of other mammalian species, we developed an airway organoid culture model derived from airways of Carollia perspicillata. Application of specific antibodies for fluorescent staining indicated that the cell composition of organoids resembled those of bat trachea and lungs as determined by immunohistochemistry. Infection studies indicated that Carollia perspicillata bat airway organoids (AOs) from the trachea or the lung are highly susceptible to infection by two different porcine influenza A viruses. The bat AOs were also used to develop an air-liquid interface (ALI) culture system of filter-grown epithelial cells. Infection of these cells showed the same characteristics, including lower virulence and enhanced replication and release of the H1N1/2006 virus compared to infection with H3N2/2007. These observations agreed with the results obtained by infection of porcine ALI cultures with these two virus strains. Interestingly, lectin staining indicated that bat airway cells only contain a small amount of alpha 2,6-linked sialic acid, the preferred receptor determinant for mammalian influenza A viruses. In contrast, large amounts of alpha 2,3-linked sialic acid, the preferred receptor determinant for avian influenza viruses, are present in bat airway epithelial cells. Therefore, bat airway cells may be susceptible not only to mammalian but also to avian influenza viruses. Our culture models, which can be extended to other parts of the airways and to other species, provide a promising tool to analyze virus infectivity and the transmission of viruses both from bats to other species and from other species to bats. IMPORTANCE We developed an organoid culture system derived from the airways of the bat species Carollia perspicillata. Using this cell system, we showed that the airway epithelium of these bats is highly susceptible to infection by influenza viruses of other mammalian species and thus is not a barrier for interspecies transmission. These organoids provide an almost unlimited supply of airway epithelial cells that can be used to generate well-differentiated epithelial cells and perform infection studies. The establishment of the organoid model required only three animals, and can be extended to other epithelia (nose, intestine) as well as to other species (bat and other animal species). Therefore, organoids promise to be a valuable tool for future zoonosis research on the interspecies transmission of viruses (e.g., bat → intermediate host → human).

An Agile Systems Modeling Framework for Bed Resource Planning During COVID-19 Pandemic in Singapore

Background

The COVID-19 pandemic has had a major impact on health systems globally. The sufficiency of hospitals' bed resource is a cornerstone for access to care which can significantly impact the public health outcomes.

Objective

We describe the development of a dynamic simulation framework to support agile resource planning during the COVID-19 pandemic in Singapore.

Materials and Methods

The study data were derived from the Singapore General Hospital and public domain sources over the period from 1 January 2020 till 31 May 2020 covering the period when the initial outbreak and surge of COVID-19 cases in Singapore happened. The simulation models and its variants take into consideration the dynamic evolution of the pandemic and the rapidly evolving policies and processes in Singapore.

Results

The models were calibrated against historical data for the Singapore COVID-19 situation. Several variants of the resource planning model were rapidly developed to adapt to the fast-changing COVID-19 situation in Singapore.

Conclusion

The agility in adaptable models and robust collaborative management structure enabled the quick deployment of human and capital resources to sustain the high level of health services delivery during the COVID-19 surge.

A Bayesian estimate of the early COVID-19 infection fatality ratio in Brazil based on a random seroprevalence survey

BACKGROUND

A number of estimates of the infection fatality ratio (IFR) of SARS-CoV-2 in different countries have been published. In Brazil, the fragile political situation, together with socioeconomic and ethnic diversity, could result in substantially different IFR estimates.

METHODS

We infer the IFR in Brazil in 2020 by combining three datasets. We compute the prevalence via the population-based seroprevalence survey, EPICOVID19-BR. For the fatalities we obtain the absolute number using the public Painel Coronavírus dataset and the age-relative number using the public SIVEP-Gripe dataset. The time delay between the development of antibodies and subsequent fatality is estimated via the SIVEP-Gripe dataset. We obtain the IFR for each survey stage and 27 federal states. We include the effect of fading IgG antibody levels by marginalizing over the test detectability time window.

RESULTS

We infer a country-wide average IFR (maximum posterior and 95% CI) of 1.03% (0.88-1.22%) and age-specific IFRs of 0.032% (0.023-0.041%) [< 30 years], 0.22% (0.18-0.27%) [30-49 years], 1.2% (1.0-1.5%) [50-69 years], and 3.0% (2.4-3.9%) [≥ 70 years]. We find that the fatality ratio in the country increased significantly at the end of June 2020, likely due to the increased strain on the health system.

CONCLUSIONS

Our IFR estimate is based on data and does not rely on extrapolating models. This estimate sets a baseline value with which future medications and treatment protocols may be confronted.

The impact of long-term non-pharmaceutical interventions on COVID-19 epidemic dynamics and control: the value and limitations of early models

Mathematical models of epidemics are important tools for predicting epidemic dynamics and evaluating interventions. Yet, because early models are built on limited information, it is unclear how long they will accurately capture epidemic dynamics. Using a stochastic SEIR model of COVID-19 fitted to reported deaths, we estimated transmission parameters at different time points during the first wave of the epidemic (March-June, 2020) in Santa Clara County, California. Although our estimated basic reproduction number ([Formula: see text]) remained stable from early April to late June (with an overall median of 3.76), our estimated effective reproduction number ([Formula: see text]) varied from 0.18 to 1.02 in April before stabilizing at 0.64 on 27 May. Between 22 April and 27 May, our model accurately predicted dynamics through June; however, the model did not predict rising summer cases after shelter-in-place orders were relaxed in June, which, in early July, was reflected in cases but not yet in deaths. While models are critical for informing intervention policy early in an epidemic, their performance will be limited as epidemic dynamics evolve. This paper is one of the first to evaluate the accuracy of an early epidemiological compartment model over time to understand the value and limitations of models during unfolding epidemics.

A model and predictions for COVID-19 considering population behavior and vaccination

The effect of vaccination coupled with the behavioral response of the population is not well understood. Our model incorporates two important dynamically varying population behaviors: level of caution and sense of safety. Level of caution increases with infectious cases, while an increasing sense of safety with increased vaccination lowers precautions. Our model accurately reproduces the complete time history of COVID-19 infections for various regions of the United States. We propose a parameter [Formula: see text] as a direct measure of a population's caution against an infectious disease that can be obtained from the infectious cases. The model provides quantitative measures of highest disease transmission rate, effective transmission rate, and cautionary behavior. We predict future COVID-19 trends in the United States accounting for vaccine rollout and behavior. Although a high rate of vaccination is critical to quickly ending the pandemic, a return towards pre-pandemic social behavior due to increased sense of safety during vaccine deployment can cause an alarming surge in infections. Our results predict that at the current rate of vaccination, the new infection cases for COVID-19 in the United States will approach zero by August 2021. This model can be used for other regions and for future epidemics and pandemics.

Diagnostic Accuracy of Web-Based COVID-19 Symptom Checkers: Comparison Study

Background

A large number of web-based COVID-19 symptom checkers and chatbots have been developed; however, anecdotal evidence suggests that their conclusions are highly variable. To our knowledge, no study has evaluated the accuracy of COVID-19 symptom checkers in a statistically rigorous manner.

Objective

The aim of this study is to evaluate and compare the diagnostic accuracies of web-based COVID-19 symptom checkers.

Methods

We identified 10 web-based COVID-19 symptom checkers, all of which were included in the study. We evaluated the COVID-19 symptom checkers by assessing 50 COVID-19 case reports alongside 410 non–COVID-19 control cases. A bootstrapping method was used to counter the unbalanced sample sizes and obtain confidence intervals (CIs). Results are reported as sensitivity, specificity, F1 score, and Matthews correlation coefficient (MCC).

Results

The classification task between COVID-19–positive and COVID-19–negative for “high risk” cases among the 460 test cases yielded (sorted by F1 score): Symptoma (F1=0.92, MCC=0.85), Infermedica (F1=0.80, MCC=0.61), US Centers for Disease Control and Prevention (CDC) (F1=0.71, MCC=0.30), Babylon (F1=0.70, MCC=0.29), Cleveland Clinic (F1=0.40, MCC=0.07), Providence (F1=0.40, MCC=0.05), Apple (F1=0.29, MCC=-0.10), Docyet (F1=0.27, MCC=0.29), Ada (F1=0.24, MCC=0.27) and Your.MD (F1=0.24, MCC=0.27). For “high risk” and “medium risk” combined the performance was: Symptoma (F1=0.91, MCC=0.83) Infermedica (F1=0.80, MCC=0.61), Cleveland Clinic (F1=0.76, MCC=0.47), Providence (F1=0.75, MCC=0.45), Your.MD (F1=0.72, MCC=0.33), CDC (F1=0.71, MCC=0.30), Babylon (F1=0.70, MCC=0.29), Apple (F1=0.70, MCC=0.25), Ada (F1=0.42, MCC=0.03), and Docyet (F1=0.27, MCC=0.29).

Conclusions

We found that the number of correctly assessed COVID-19 and control cases varies considerably between symptom checkers, with different symptom checkers showing different strengths with respect to sensitivity and specificity. A good balance between sensitivity and specificity was only achieved by two symptom checkers.

Influencing public health policy with data-informed mathematical models of infectious diseases: Recent developments and new challenges

Modern data and computational resources, coupled with algorithmic and theoretical advances to exploit these, allow disease dynamic models to be parameterised with increasing detail and accuracy. While this enhances models' usefulness in prediction and policy, major challenges remain. In particular, lack of identifiability of a model's parameters may limit the usefulness of the model. While lack of parameter identifiability may be resolved through incorporation into an inference procedure of prior knowledge, formulating such knowledge is often difficult. Furthermore, there are practical challenges associated with acquiring data of sufficient quantity and quality. Here, we discuss recent progress on these issues.

Using Simulation Modeling to Guide the Design of the Girl Scouts Fierce & Fit Program

OBJECTIVE

The study aim was to help the Girl Scouts of Central Maryland evaluate, quantify, and potentially modify the Girl Scouts Fierce & Fit program.

METHODS

From 2018 to 2019, our Public Health Informatics, Computational, and Operations Research team developed a computational simulation model representing the 250 adolescent girls participating in the Fierce & Fit program and how their diets and physical activity affected their BMI and subsequent outcomes, including costs.

RESULTS

Changing the Fierce & Fit program from a 6-week program meeting twice a week, with 5 minutes of physical activity each session, to a 12-week program meeting twice a week with 30 minutes of physical activity saved an additional $84,828 ($80,130-$89,526) in lifetime direct medical costs, $81,365 ($76,528-$86,184) in lifetime productivity losses, and 7.85 (7.38-8.31) quality-adjusted life-years. The cost-benefit of implementing this program was $95,943. Based on these results, the Girl Scouts of Central Maryland then implemented these changes in the program.

CONCLUSIONS

This is an example of using computational modeling to help evaluate and revise the design of a program aimed at increasing physical activity among girls.

Exploring the role of mass immunisation in influenza pandemic preparedness: A modelling study for the UK context

The nature and timing of the next influenza pandemic is unknown. This makes it difficult for policy makers to assess whether spending money now to prepare for mass immunisation in the event of a pandemic is worthwhile. We used simple epidemiological modelling and health economic analysis to identify the range of pandemic and policy scenarios under which plans to immunise the general UK population would have net benefit if a stockpiled vaccine or, alternatively, a responsively purchased vaccine were used. Each scenario we studied comprised a combination of pandemic, vaccine and immunisation programme characteristics in presence or absence of access to effective antivirals, with the chance of there being a pandemic each year fixed. Monetarised health benefits and cost savings from any influenza cases averted were set against the option, purchase, storage, distribution, administration, and disposal costs relevant for each scenario to give a discounted net present value over 10 years for planning to immunise, accounting for the possibility that there may be no pandemic over the period considered. To support understanding and exploration of model output, an interactive visualisation tool was devised and made available online. We evaluated over 29 million combinations of pandemic and policy characteristics. Preparedness plans incorporating mass immunisation show positive net present value for a wide range of scenarios, predominantly in the absence of effective antivirals. Plans based on the responsive purchase of vaccine have wider benefit than plans reliant on the purchase and maintenance of a stockpile if immunisation can start without extensive delays. This finding is not dependent on responsively purchased vaccine being more effective than stockpiled vaccine, but rather is driven by avoiding the costs of storing and replenishing a stockpile.

The Potential Health Care Costs And Resource Use Associated With COVID-19 In The United States

With the coronavirus disease 2019 (COVID-19) pandemic, one of the major concerns is the direct medical cost and resource use burden imposed on the US health care system. We developed a Monte Carlo simulation model that represented the US population and what could happen to each person who got infected. We estimated resource use and direct medical costs per symptomatic infection and at the national level, with various "attack rates" (infection rates), to understand the potential economic benefits of reducing the burden of the disease. A single symptomatic COVID-19 case could incur a median direct medical cost of $3,045 during the course of the infection alone. If 80 percent of the US population were to get infected, the result could be a median of 44.6 million hospitalizations, 10.7 million intensive care unit (ICU) admissions, 6.5 million patients requiring a ventilator, 249.5 million hospital bed days, and $654.0 billion in direct medical costs over the course of the pandemic. If 20 percent of the US population were to get infected, there could be a median of 11.2 million hospitalizations, 2.7 million ICU admissions, 1.6 million patients requiring a ventilator, 62.3 million hospital bed days, and $163.4 billion in direct medical costs over the course of the pandemic.

Using "outbreak science" to strengthen the use of models during epidemics

Infectious disease modeling has played a prominent role in recent outbreaks, yet integrating these analyses into public health decision-making has been challenging. We recommend establishing ‘outbreak science’ as an inter-disciplinary field to improve applied epidemic modeling.

Are we prepared for the next influenza pandemic? Lessons from modelling different preparedness policies against four pandemic scenarios

In the event of a novel influenza strain that is markedly different to the current strains circulating in humans, the population have little/no immunity and infection spreads quickly causing a global pandemic. Over the past century, there have been four major influenza pandemics: the 1918 pandemic ("Spanish Flu"), the 1957-58 pandemic (the "Asian Flu"), the 1967-68 pandemic (the "Hong Kong Flu") and the 2009 pandemic (the "Swine flu"). To inform planning against future pandemics, this paper investigates how different is the net-present value of employing pre-purchase and responsive- purchased vaccine programmes in presence and absence of anti-viral drugs to scenarios that resemble these historic influenza pandemics. Using the existing literature and in discussions with policy decision makers in the UK, we first characterised the four past influenza pandemics by their transmissibility and infection-severity. For these combinations of parameters, we then projected the net-present value of employing pre-purchase vaccine (PPV) and responsive-purchase vaccine (RPV) programmes in presence and absence of anti-viral drugs. To differentiate between PPV and RPV policies, we changed the vaccine effectiveness value and the time to when the vaccine is first available. Our results are "heat-map" graphs displaying the benefits of different strategies in pandemic scenarios that resemble historic influenza pandemics. Our results suggest that immunisation with either PPV or RPV in presence of a stockpile of effective antiviral drugs, does not have positive net-present value for all of the pandemic scenarios considered. In contrast, in the absence of effective antivirals, both PPV and RPV policies have positive net-present value across all the pandemic scenarios. Moreover, in all considered circumstances, vaccination was most beneficial if started sufficiently early and covered sufficiently large number of people. When comparing the two vaccine programmes, the RPV policy allowed a longer timeframe and lower coverage to attain the same benefit as the PPV policy. Our findings suggest that responsive-purchase vaccination policy has a bigger window of positive net-present value when employed against each of the historic influenza pandemic strains but needs to be rapidly available to maximise benefit. This is important for future planning as it suggests that future preparedness policies may wish to consider utilising timely (i.e. responsive-purchased) vaccines against emerging influenza pandemics.

Assessing the Use of Influenza Forecasts and Epidemiological Modeling in Public Health Decision Making in the United States

Although forecasts and other mathematical models have the potential to play an important role in mitigating the impact of infectious disease outbreaks, the extent to which these tools are used in public health decision making in the United States is unclear. Throughout 2015, we invited public health practitioners belonging to three national public health organizations to complete a cross-sectional survey containing questions on model awareness, model use, and communication with modelers. Of 39 respondents, 46.15% used models in their work, and 20.51% reported direct communication with those who create models. Over half (64.10%) were aware that influenza forecasts exist. The need for improved communication between practitioners and modelers was overwhelmingly endorsed, with over 50% of participants indicating the need for models more relevant to public health questions, increased frequency of telecommunication, and more plain language in discussing models. Model use for public health decision making must be improved if models are to reach their full potential as public health tools. Increased quality and frequency of communication between practitioners and modelers could be particularly useful in achieving this goal. It is important that improvements be made now, rather than waiting for the next public health crisis to occur.

A systems approach to obesity

Obesity has become a truly global epidemic, affecting all age groups, all populations, and countries of all income levels. To date, existing policies and interventions have not reversed these trends, suggesting that innovative approaches are needed to transform obesity prevention and control. There are a number of indications that the obesity epidemic is a systems problem, as opposed to a simple problem with a linear cause-and-effect relationship. What may be needed to successfully address obesity is an approach that considers the entire system when making any important decision, observation, or change. A systems approach to obesity prevention and control has many benefits, including the potential to further understand indirect effects or to test policies virtually before implementing them in the real world. Discussed here are 5 key efforts to implement a systems approach for obesity prevention: 1) utilize more global approaches; 2) bring new experts from disciplines that do not traditionally work with obesity to share experiences and ideas with obesity experts; 3) utilize systems methods, such as systems mapping and modeling; 4) modify and combine traditional approaches to achieve a stronger systems orientation; and 5) bridge existing gaps between research, education, policy, and action. This article also provides an example of how a systems approach has been used to convene a multidisciplinary team and conduct systems mapping and modeling as part of an obesity prevention program in Baltimore, Maryland.

A systems approach to vaccine decision making

Vaccines reside in a complex multiscale system that includes biological, clinical, behavioral, social, operational, environmental, and economical relationships. Not accounting for these systems when making decisions about vaccines can result in changes that have little effect rather than solutions, lead to unsustainable solutions, miss indirect (e.g., secondary, tertiary, and beyond) effects, cause unintended consequences, and lead to wasted time, effort, and resources. Mathematical and computational modeling can help better understand and address complex systems by representing all or most of the components, relationships, and processes. Such models can serve as "virtual laboratories" to examine how a system operates and test the effects of different changes within the system. Here are ten lessons learned from using computational models to bring more of a systems approach to vaccine decision making: (i) traditional single measure approaches may overlook opportunities; (ii) there is complex interplay among many vaccine, population, and disease characteristics; (iii) accounting for perspective can identify synergies; (iv) the distribution system should not be overlooked; (v) target population choice can have secondary and tertiary effects; (vi) potentially overlooked characteristics can be important; (vii) characteristics of one vaccine can affect other vaccines; (viii) the broader impact of vaccines is complex; (ix) vaccine administration extends beyond the provider level; and (x) the value of vaccines is dynamic.

Detecting Presymptomatic Infection Is Necessary to Forecast Major Epidemics in the Earliest Stages of Infectious Disease Outbreaks

We assess how presymptomatic infection affects predictability of infectious disease epidemics. We focus on whether or not a major outbreak (i.e. an epidemic that will go on to infect a large number of individuals) can be predicted reliably soon after initial cases of disease have appeared within a population. For emerging epidemics, significant time and effort is spent recording symptomatic cases. Scientific attention has often focused on improving statistical methodologies to estimate disease transmission parameters from these data. Here we show that, even if symptomatic cases are recorded perfectly, and disease spread parameters are estimated exactly, it is impossible to estimate the probability of a major outbreak without ambiguity. Our results therefore provide an upper bound on the accuracy of forecasts of major outbreaks that are constructed using data on symptomatic cases alone. Accurate prediction of whether or not an epidemic will occur requires records of symptomatic individuals to be supplemented with data concerning the true infection status of apparently uninfected individuals. To forecast likely future behavior in the earliest stages of an emerging outbreak, it is therefore vital to develop and deploy accurate diagnostic tests that can determine whether asymptomatic individuals are actually uninfected, or instead are infected but just do not yet show detectable symptoms.

Emerging epidemics pose a significant challenge to human health worldwide. Accurate real-time forecasts of whether or not initial reports will be followed by a major outbreak are necessary for efficient deployment of control. For all infectious diseases, there is a delay between infection and the appearance of symptoms, i.e. an initial period following first infection during which infected individuals remain presymptomatic. We use mathematical modeling to evaluate the effect of presymptomatic infection on predictions of major epidemics. Our results show rigorously, for the first time, that precise estimates of the current number of infected individuals—and consequently the chance of a major outbreak in future—cannot be inferred from data on symptomatic cases alone. This is the case even if the values of epidemiological parameters, such as the average infection and death or recovery rates of individuals in the population, can be estimated accurately. Accurate prediction is in fact impossible without additional data from which the number of currently infected but as yet presymptomatic individuals can be deduced.

Quantitative evaluation of the strategy to eliminate human African trypanosomiasis in the Democratic Republic of Congo

Background

The virulent vector-borne disease, Gambian human African trypanosomiasis (HAT), is one of several diseases targeted for elimination by the World Health Organization. This article utilises human case data from a high-endemicity region of the Democratic Republic of Congo in conjunction with a suite of novel mechanistic mathematical models to address the effectiveness of on-going active screening and treatment programmes and compute the likely time to elimination as a public health problem (i.e. <1 case per 10,000 per year).

Methods

The model variants address uncertainties surrounding transmission of HAT infection including heterogeneous risk of exposure to tsetse bites, non-participation of certain groups during active screening campaigns and potential animal reservoirs of infection.

Results

Model fitting indicates that variation in human risk of tsetse bites and participation in active screening play a key role in transmission of this disease, whilst the existence of animal reservoirs remains unclear. Active screening campaigns in this region are calculated to have been effective, reducing the incidence of new human infections by 52–53 % over a 15-year period (1998–2012). However, projections of disease dynamics in this region indicate that the elimination goal may not be met until later this century (2059–2092) under the current intervention strategy.

Conclusions

Improvements to active detection, such as screening those who have not previously participated and raising overall screening levels, as well as beginning widespread vector control in the area have the potential to ensure successful and timely elimination.

Electronic supplementary material

The online version of this article (doi:10.1186/s13071-015-1131-8) contains supplementary material, which is available to authorized users.