Modelling: Understanding pandemics and how to control them.

New disease challenges, societal demands and better or novel types of data, drive innovations in the structure, formulation and analysis of epidemic models. Innovations in modelling can lead to new insights into epidemic processes and better use of available data, yielding improved disease control and stimulating collection of better data and new data types. Here we identify key challenges for the structure, formulation, analysis and use of mathematical models of pathogen transmission relevant to current and future pandemics.

Modelling varicella vaccination - What does a lack of surge in herpes zoster incidence tell us about exogenous boosting?

BACKGROUND: For decades, assessments of the impact of universal varicella vaccination on the epidemiology of varicella and herpes zoster (HZ) have been made using mathematical modelling. Decreased virus circulation and the resulting diminished exogenous boosting have been predicted to lead to a surge in HZ incidence. Lately, the exogenous boosting hypothesis has been challenged due to a lack of an extensive surge in HZ incidence in countries with, by now long-standing universal varicella vaccination. METHODS: In a deterministic compartmental transmission model of varicella zoster virus disease, we model various levels and duration of protection from boosting to explore the impact of successful childhood varicella vaccination on HZ incidence. RESULTS: Considering total HZ incidence, lifelong and strong protection from boosting give a stable incidence of HZ for about 60 years followed by a decline, whereas lifelong intermediate protection leads to a decline. So does weak protection of intermediate duration. Full and short protection, lead to a small surge, while full and intermediate protection lead to the largest HZ surge. HZ incidence by age group show that total incidence is the result of opposing increasing and decreasing trends in the various age groups over time. CONCLUSIONS: The absence of an extensive surge in HZ incidence after varicella vaccination can, especially during the first 20-30 years, occur in either strong, intermediate or weak boosting scenarios. The impact seems to depend on an interplay of the protective level and duration of the protection in determining the basic reactivation rate and the proportion of the population that is susceptible at the start of vaccination. However, the picture depends on whether the entire population or specific age groups are observed.

Key questions for modelling COVID-19 exit strategies.

Combinations of intense non-pharmaceutical interventions (lockdowns) were introduced worldwide to reduce SARS-CoV-2 transmission. Many governments have begun to implement exit strategies that relax restrictions while attempting to control the risk of a surge in cases. Mathematical modelling has played a central role in guiding interventions, but the challenge of designing optimal exit strategies in the face of ongoing transmission is unprecedented. Here, we report discussions from the Isaac Newton Institute 'Models for an exit strategy' workshop (11-15 May 2020). A diverse community of modellers who are providing evidence to governments worldwide were asked to identify the main questions that, if answered, would allow for more accurate predictions of the effects of different exit strategies. Based on these questions, we propose a roadmap to facilitate the development of reliable models to guide exit strategies. This roadmap requires a global collaborative effort from the scientific community and policymakers, and has three parts: (i) improve estimation of key epidemiological parameters; (ii) understand sources of heterogeneity in populations; and (iii) focus on requirements for data collection, particularly in low-to-middle-income countries. This will provide important information for planning exit strategies that balance socio-economic benefits with public health.

"How do online and offline sampling compare in a multinational study of drug use and nightlife behaviour?"

BACKGROUND: Online sampling is widely used to recruit hard to reach samples such as drug users at nightlife events. We conducted the first study comparing differences in demographics, drug use and nightlife behaviour between an online sample of young adults engaging with the European nightlife scene, and an offline sample recruited at nightclubs and festivals in Europe. METHODS: Online participants who attended at least six nightlife events in the past 12 months were recruited using social media advertising (May-November 2017). Offline participants were recruited at nightclubs and festivals using a random intercept method (May-November 2017). Samples were compared with respect to age, gender, past year use of alcohol, cannabis, cocaine, ecstasy/MDMA and amphetamines, and past year attendance at nightclubs, licensed festivals, illegal festivals, pubs and house-parties. RESULTS: 6153 online and 3529 offline participants were recruited. When adjusting for differences in age and gender, online participants were less likely to have used each drug and to have attended illegal festivals, pubs and house-parties in the past 12 months. The online sample also used each drug and attended each venue, with the exception of nightclubs, less frequently on average than offline participants. Adjusted odds ratios (range 0.37 to 1.39) and regression coefficients (range -0.84 to 0.07) indicate that the majority of observed differences between the samples were of a small effect size. CONCLUSIONS: Estimates of drug use and nightlife engagement are more conservative when using online sampling compared to venue based sampling. Observed differences were generally small in effect, indicating good overall representativeness when using online sampling in the European nightlife scene.

Evaluating costs and health consequences of sick leave strategies against pandemic and seasonal influenza in Norway using a dynamic model.

OBJECTIVES: To quantify population-level health and economic consequences of sick leave among workers with influenza symptoms. INTERVENTIONS: Compared with current sick leave practice (baseline), we evaluated the health and cost consequences of: (1) increasing the proportion of workers on sick leave from 65% (baseline) to 80% or 90%; (2) shortening the maximum duration from symptom onset to sick leave from 4 days (baseline) to 2 days, 1.5 days, 1 day and 0.5 days; and (3) combinations of 1 and 2. METHODS: A dynamic compartmental influenza model was developed using Norwegian population data and survey data on employee sick leave practices. The sick leave interventions were simulated under 12 different seasonal epidemic and 36 different pandemic influenza scenarios. These scenarios varied in terms of transmissibility, the proportion of symptomatic cases and illness severity (risk of primary care consultations, hospitalisations and deaths). Using probabilistic sensitivity analyses, a net health benefit approach was adopted to assess the cost-effectiveness of the interventions from a societal perspective. RESULTS: Compared with current sick leave practice, sick leave interventions were cost-effective for 31 (65%) of the pandemic scenarios, and 11 (92%) of the seasonal scenarios. Economic benefits from sick leave interventions were greatest for scenarios with low transmissibility, high symptomatic proportions and high illness severity. Overall, the health and economic benefits were greatest for the intervention involving 90% of sick workers taking sick leave within one-half day of symptoms. Depending on the influenza scenario, this intervention resulted in a 44.4%-99.7% reduction in the attack rate. Interventions involving sick leave onset beginning 2 days or later, after the onset of symptoms, resulted in economic losses. CONCLUSIONS: Prompt sick leave onset and a high proportion of sick leave among workers with influenza symptoms may be cost-effective, particularly during influenza epidemics and pandemics with low transmissibility or high morbidity.

Influenza in workplaces: transmission, workers' adherence to sick leave advice and European sick leave recommendations.

BACKGROUND: Knowledge about influenza transmission in the workplace and whether staying home from work when experiencing influenza-like illness can reduce the spread of influenza is crucial for the design of efficient public health initiatives. AIM: This review synthesizes current literature on sickness presenteeism and influenza transmission in the workplace and provides an overview of sick leave recommendations in Europe for influenza. METHODS: A search was performed on Medline, Embase, PsychINFO, Cinahl, Web of Science, Scopus and SweMed to identify studies related to workplace contacts, -transmission, -interventions and compliance with recommendations to take sick leave. A web-based survey on national recommendations and policies for sick leave during influenza was issued to 31 European countries. RESULTS: Twenty-two articles (9 surveys; 13 modelling articles) were eligible for this review. Results from social mixing studies suggest that 20-25% of weekly contacts are made in the workplace, while modelling studies suggest that on average 16% (range 9-33%) of influenza transmission occurs in the workplace. The effectiveness of interventions to reduce workplace presenteeism is largely unknown. Finally, estimates from studies reporting expected compliance with sick leave recommendations ranged from 71 to 95%. Overall, 18 countries participated in the survey of which nine (50%) had issued recommendations encouraging sick employees to stay at home during the 2009 A(H1N1) pandemic, while only one country had official recommendations for seasonal influenza. CONCLUSIONS: During the 2009 A(H1N1) pandemic, many European countries recommended ill employees to take sick leave. Further research is warranted to quantify the effect of reduced presenteeism during influenza illness.

Four key challenges in infectious disease modelling using data from multiple sources.

Public health-related decision-making on policies aimed at controlling epidemics is increasingly evidence-based, exploiting multiple sources of data. Policy makers rely on complex models that are required to be robust, realistically approximating epidemics and consistent with all relevant data. Meeting these requirements in a statistically rigorous and defendable manner poses a number of challenging problems. How to weight evidence from different datasets and handle dependence between them, efficiently estimate and critically assess complex models are key challenges that we expound in this paper, using examples from influenza modelling.

Estimates of cocaine use in Milan.

A comparative analysis of three different estimation methods of cocaine use in Milan, Italy, is carried out, including an analysis of the size and trends of the underlying reference population. The three cocaine use estimates are derived from wastewater analysis, a "street" survey and one-source capture-recapture analysis of administrative sanctions for drug possession. All three data sources span several years during the decade 2000-2010. For each method, assumptions and limitations are discussed. It is concluded, although the amount of data regarding cocaine use in Milan during the years 2000-2010 is considerable, that none of the estimates is completely reliable, mainly because of the many assumptions needed for inference to the whole community, that all three estimates however agree on a substantial reduction in cocaine use and users between the years 2007 and 2010. The recently developed wastewater analysis technique yields estimates comparable to those derived from the street based survey, while capture-recapture analysis of administrative data on identification by law enforcement agencies as cocaine user probably targets only a subpopulation of all users.

Methods for estimating incidence of drug use: a review.

Several methods for estimating drug use incidence, that have been used in the literature or that could be used, having been used in a different framework, are described and commented. The applicability of the different methods depends on available data and knowledge about relevant parameters. The many similarities between drug use incidence estimation and estimation of disease incidence are highlighted, but also the distinguishing aspects that make drug use incidence estimation a challenge to standard statistical methods.

Alternative splicing tends to avoid partial removals of protein-protein interaction sites.

BACKGROUND: Anecdotal evidence of the involvement of alternative splicing (AS) in the regulation of protein-protein interactions has been reported by several studies. AS events have been shown to significantly occur in regions where a protein interaction domain or a short linear motif is present. Several AS variants show partial or complete loss of interface residues, suggesting that AS can play a major role in the interaction regulation by selectively targeting the protein binding sites. In the present study we performed a statistical analysis of the alternative splicing of a non-redundant dataset of human protein-protein interfaces known at molecular level to determine the importance of this way of modulation of protein-protein interactions through AS. RESULTS: Using a Cochran-Mantel-Haenszel chi-square test we demonstrated that the alternative splicing-mediated partial removal of both heterodimeric and homodimeric binding sites occurs at lower frequencies than expected, and this holds true even if we consider only those isoforms whose sequence is less different from that of the canonical protein and which therefore allow to selectively regulate functional regions of the protein. On the other hand, large removals of the binding site are not significantly prevented, possibly because they are associated to drastic structural changes of the protein. The observed protection of the binding sites from AS is not preferentially directed towards putative hot spot interface residues, and is widespread to all protein functional classes. CONCLUSIONS: Our findings indicate that protein-protein binding sites are generally protected from alternative splicing-mediated partial removals. However, some cases in which the binding site is selectively removed exist, and here we discuss one of them.

Effect of vaccines and antivirals during the major 2009 A(H1N1) pandemic wave in Norway--and the influence of vaccination timing.

To evaluate the impact of mass vaccination with adjuvanted vaccines (eventually 40% population coverage) and antivirals during the 2009 influenza pandemic in Norway, we fitted an age-structured SEIR model using data on vaccinations and sales of antivirals in 2009/10 in Norway to Norwegian ILI surveillance data from 5 October 2009 to 4 January 2010. We estimate a clinical attack rate of approximately 30% (28.7-29.8%), with highest disease rates among children 0-14 years (43-44%). Vaccination started in week 43 and came too late to have a strong influence on the pandemic in Norway. Our results indicate that the countermeasures prevented approximately 11-12% of potential cases relative to an unmitigated pandemic. Vaccination was found responsible for roughly 3 in 4 of the avoided infections. An estimated 50% reduction in the clinical attack rate would have resulted from vaccination alone, had the campaign started 6 weeks earlier. Had vaccination been prioritized for children first, the intervention should have commenced approximately 5 weeks earlier in order to achieve the same 50% reduction. In comparison, we estimate that a non-adjuvanted vaccination program should have started 8 weeks earlier to lower the clinical attack rate by 50%. In conclusion, vaccination timing was a critical factor in relation to the spread of the 2009 A(H1N1) influenza. Our results also corroborate the central role of children for the transmission of A(H1N1) pandemic influenza.

Social contacts and mixing patterns relevant to the spread of infectious diseases.

BACKGROUND: Mathematical modelling of infectious diseases transmitted by the respiratory or close-contact route (e.g., pandemic influenza) is increasingly being used to determine the impact of possible interventions. Although mixing patterns are known to be crucial determinants for model outcome, researchers often rely on a priori contact assumptions with little or no empirical basis. We conducted a population-based prospective survey of mixing patterns in eight European countries using a common paper-diary methodology. METHODS AND FINDINGS: 7,290 participants recorded characteristics of 97,904 contacts with different individuals during one day, including age, sex, location, duration, frequency, and occurrence of physical contact. We found that mixing patterns and contact characteristics were remarkably similar across different European countries. Contact patterns were highly assortative with age: schoolchildren and young adults in particular tended to mix with people of the same age. Contacts lasting at least one hour or occurring on a daily basis mostly involved physical contact, while short duration and infrequent contacts tended to be nonphysical. Contacts at home, school, or leisure were more likely to be physical than contacts at the workplace or while travelling. Preliminary modelling indicates that 5- to 19-year-olds are expected to suffer the highest incidence during the initial epidemic phase of an emerging infection transmitted through social contacts measured here when the population is completely susceptible. CONCLUSIONS: To our knowledge, our study provides the first large-scale quantitative approach to contact patterns relevant for infections transmitted by the respiratory or close-contact route, and the results should lead to improved parameterisation of mathematical models used to design control strategies.

Epidemiology and costs of herpes zoster: background data to estimate the impact of vaccination.

In 2004, we conducted a study in Piemonte (Italy), in order to describe incidence, treatment, hospitalizations and costs of herpes zoster (HZ), in population over 14 years of age. Twenty-four regional general practitioners, with 26,394 patients >14 years in charge (0.71% of the regional population), reported prospectively all diagnosed HZ cases. In addition, all regional hospital discharge records were reviewed. Forty-six HZ cases treated at home were reported, accounting for a total incidence of 1.74 cases/1000 population >14 years per year. HZ rate standardized by age on regional population 14 years older is 1.59/1000. The cost per observed home case was 136.06 euros. The incidence of hospital admissions was 0.12/1000 inhabitants. The mean cost of hospitalized cases was 4082.59 euros. These results contribute to depict the impact of HZ in the general population, and to provide background data for setting-up either mathematical models aimed to estimate the impact of vaccination on HZ, and the cost-benefit analyses of various preventive and therapeutic scenarios.

Changes in the expression of hypothalamic lipid sensing genes in rat model of intrauterine growth retardation (IUGR).

Intrauterine growth retardation (IUGR) has been linked to the development of type 2 diabetes in later life. The mechanisms underlying this phenomenon are unknown. Recent data suggest that some of the molecular defects underlying type 2 diabetes reside in the CNS. The enzyme carnitine palmitoyltransferase-1 (CPT1) regulates long-chain fatty acid (LCFA) entry into mitochondria, where LCFA undergo beta-oxidation. Hypothalamic inhibition of CPT1 decreases food intake and suppresses endogenous glucose production. Our aim was to investigate the effects of uterine artery ligation, a procedure that mimics uteroplacental insufficiency, on the CNS expression of CPT1 and other key enzymes of LCFA metabolism. Bilateral uterine artery ligation was performed on d 19 of gestation in the pregnant rat; sham-operated pregnant rats served as controls. Hypothalamus, cerebellum, hippocampus, and cortex were dissected and analyzed at birth by real-time PCR. Nonesterified fatty acid (NEFA) serum levels were significantly higher in IUGR pups (p<0.0001). In IUGR rats, the hypothalamic expression of CPT1 isoform C (p=0.005) and acetyl-CoA carboxylase (ACC) isoforms alpha (p<0.05) and beta (p=0.005) were significantly decreased. The data presented here support the hypothesis that an abnormal intrauterine milieu can induce changes in hypothalamic lipid sensing.

[Assessment of the potential impact in Italy of extensive varicella vaccination programs based on a mathematical model]. / Valutazione del potenziale impatto in Italia di programmi estesi di vaccinazione antivaricella secondo un modello matematico.

A mathematical model has been developed in order to assess the effect of extended programs of varicella vaccination on the epidemiology of the disease in Italy. The effect of different vaccination options have been estimated by the change in incidence of the disease and age distribution of cases over a short and long period of time. The developed mathematical model reproduces chickenpox transmission and immunisation; five strategies different for target age of vaccination and/or proportion of vaccinated subjects have been considered. In all scenarios the model pointed out an initial decrease of case frequency observed in the first 3-5 years, followed by a series of epidemic peaks, variable in number and size by vaccination strategy. Moreover, as the number of cases among infants decreases, the number of cases among adults increases. Such event is minimised only by very high vaccination coverage (80% in the first year of life and 50% at 12 years of age). Extensive programmes of vaccination against chickenpox must reach a high coverage as soon as possible in order to avoid undesirable effects that may move forward the age of cases and therefore should be offered to target age groups easy to reach.