Lightnews — Scholar-powered news

James Munday @jdmunday.bsky.social · 2d

We promote further investigation of integrating expert opinion in short-term forecasting, especially the development of scalable systems to elicit forecasts. 5/5

James Munday @jdmunday.bsky.social · 2d

In general, the experts tended to over-predict small numbers of cases in more geographical regions than the models but were less likely to predict larger flare-ups. 4/5

1

James Munday @jdmunday.bsky.social · 2d

We found that while no individual expert could consistently compete with the mathematical models, the performance of the ensemble forecast of the experts was comparable. 3/5

1

James Munday @jdmunday.bsky.social · 2d

We evaluated the relative performance of a panel of experts with two mathematical models when forecasting the number and spatial distribution of future cases of Ebola Virus Disease in real-time during the 2018-2020 outbreak in North-eastern DRC. 2/5

1

James Munday @jdmunday.bsky.social · 2d

Our paper "Forecasting the spatial spread of an Ebola epidemic in real-time: comparing predictions of mathematical models and experts" is now published as a 'version of record' in @elife.bsky.social (with @sbfnk.bsky.social Alicia Rosello and John Edmunds) doi.org/10.7554/eLif... 1/5

Forecasting the spatial spread of an Ebola epidemic in real time: Comparing predictions of mathematical models and experts

Models were more consistent than individual experts but performed comparably to ensembles of many experts and relative performance varied by month, justifying the use of combined model and expert fore...

doi.org

1 2 8

James Munday @jdmunday.bsky.social · 2d

We found that while no individual expert could consistently compete with the mathematical models, the performance of the ensemble forecast of the experts was comparable. 3/5

James Munday @jdmunday.bsky.social · 2d

We evaluated the relative performance of a panel of experts with two mathematical models when forecasting the number and spatial distribution of future cases of Ebola Virus Disease in real-time during the 2018-2020 outbreak in North-eastern DRC. 2/5

1

James Munday @jdmunday.bsky.social · 15d

The study provides a nice insight into the pathway between respiratory outbreaks in schools and voluntary absence. Further research is warranted to understand the role of asymptomatic students in transmission and the potential of promoting voluntary absence as part of infection control.

1

James Munday @jdmunday.bsky.social · 15d

In students who took absence, those testing positive for Influenza B almost all reported a fever, runny nose and cough, whereas those testing positive for Influenza A were more likely to report fatigue, with a lower proportion (~60%) reporting fever.

1 1

James Munday @jdmunday.bsky.social · 15d

Absences were most clearly correlated with Influenza B (mean of 80% of infections [37 - 96% CI)] associated with absence) followed by Parainfluenza virus and Influenza A with means of 50% and 48% of episodes associated with absence.

1

James Munday @jdmunday.bsky.social · 15d

Using data from a longitudinal study in four classes in a secondary school, molecular (sputum) samples were tested for 24 resp. pathogens. Children who took time off school due to symptoms were surveyed for symptoms.

1

James Munday @jdmunday.bsky.social · 15d

New paper led by Nicolas Banholzer "Absences, Symptoms and Respiratory Viruses in a Swiss School: Longitudinal Study With Serial Saliva Sampling" Now published in *Influenza and Other Respiratory Viruses* doi.org/10.1111/irv....

Absences, Symptoms and Respiratory Viruses in a Swiss School: Longitudinal Study With Serial Saliva Sampling

Background Viral respiratory infections contribute to sick days in school children. We monitored respiratory infections, absences, and symptoms in a Swiss school. Methods Serial saliva sampling (t...

doi.org

1 3

James Munday @jdmunday.bsky.social · Sep 4

We evaluate outbreak risk across the school network in each academic year. We find that while the first children born since the previous outbreak remain in primary school, outbreaks remain small. As they enter secondary school the expected outbreak sizes return to the order of that seen in 2013/14.

James Munday @jdmunday.bsky.social · Sep 4

We constructed a school-household network for the Netherlands using national school records data. Using school level MMR coverage estimates from 2013 we propagate susceptible children through schools affected by the outbreak as new cohorts progress through the school system.

1

James Munday @jdmunday.bsky.social · Sep 4

This follows our previous work on understanding risk of measles outbreaks in Dutch schools: doi.org/10.1371/jour....

Estimating the risk and spatial spread of measles in populations with high MMR uptake: Using school-household networks to understand the 2013 to 2014 outbreak in the Netherlands

Using Dutch household networks, James Munday and team investigate the risk of measles spreading in populations with high MMR uptake.

doi.org

1

James Munday @jdmunday.bsky.social · Sep 4

New Preprint 💫: *Projected sudden step-wise increase in risk of large measles outbreaks in the Netherlands as susceptible children enter secondary school in 2025/26* on MedRxiv doi.org/10.1101/2025... (NOT YET PEER REVIEWED)

Projected sudden step-wise increase in risk of large measles outbreaks in the Netherlands as susceptible children enter secondary school in 2025/26

Background Many unvaccinated children were naturally immunised through infection during the last national Measles epidemic in the Netherlands (2013/14). Over time susceptibility amongst school aged ch...

doi.org

1 3

Reposted by James Munday

Jana Huisman @jhuisman.bsky.social · Aug 22

Our paper on the transmission dynamics of Noro- and Enterovirus during the COVID-19 pandemic just came out: www.sciencedirect.com/science/arti...
We measured wastewater viral loads, estimated Re and assessed the impact of non-pharmaceutical interventions on the transmission of these pathogens.

Transmission dynamics of Norovirus GII and Enterovirus in Switzerland during the COVID-19 pandemic (2021–2022) as evidenced in wastewater

Noroviruses and enteroviruses are major causes of endemic gastrointestinal disease associated with substantial disease burden. However, viral gastroen…

www.sciencedirect.com

1 2 10

James Munday @jdmunday.bsky.social · Aug 22

With Lukas Fenner, @phips81.bsky.social, @ccattuto.bsky.social, Tina Hascher, Mathias Egger, @tanjastadler.bsky.social, Pascal Bittel, @lo-dallamico.bsky.social‬, Lavinia
Furrer and Charlyne Bürki

1

James Munday @jdmunday.bsky.social · Aug 22

Our results indicate that extended periods spent in poorly ventilated classrooms may be a stronger driver of within-school transmission than individual contact rates.

1 1

James Munday @jdmunday.bsky.social · Aug 22

We found that while close contact contributed to elevated risk (rate ratio 1.16 per doubling daily time, 95%-CI 1.01–1.33), time spent in shared classrooms and poor air quality had larger effects (RR 8.96, 95%-CI 4.85–16.88 and RR 5.59 95%-CI 3.25–9.83 respectively).

Association of respiratory virus transmission with close proximity, shared classroom environment and indoor air quality. (A–C) Estimated transmission risk by days of exposure over one school week, shown as mean estimates (lines) with 95% confidence intervals (shaded areas). (A) Transmission risk by daily time spent in close proximity, for both within- and between-class transmissions. (B) Transmission risk for students depending on whether they shared a classroom. (C) Transmission risk by daily duration that classroom CO₂ levels exceeded 1,000 ppm (indicative of suboptimal air quality), restricted to within-class transmissions. (D) Relative contribution of each risk factor to transmission, expressed as the mean (dots) and 95% confidence interval (lines) of the changes in the model log-likelihoods when adding each predictor separately to the pairwise accelerated failure time regression model without covariates (null model). Note that the contributions of shared classroom time and low air quality could not be assessed (n.a.) for all transmissions combined (within and between classes), and the effect of being in the same class could not be assessed for within-class transmissions alone.

1 2

James Munday @jdmunday.bsky.social · Aug 22

Combining these data we inferred the relative contributions of 🧑‍🤝‍🧑 close contact, 🧑‍🏫 shared class rooms and 🍃 air quality on transmission risk using a pairwise survival analysis framework.

Transmission networks of influenza A and respiratory syncytial virus based on epidemiological and genomic data. The networks show all plausible internal transmissions between students participating in the study based on epidemiological delays and school absences. It excludes implausible links through genomic analysis. The probability refers to the proportion of paired datasets in which the link was present, considering uncertainty in the timing of infections and infectious periods. The temporal axis shows the date of the first positive saliva sample, which can be several days after the date of infection, so that circular links are possible.

1 1

James Munday @jdmunday.bsky.social · Aug 22

We collected molecular (saliva 3x weekly), contact (prox. sens.) and air quality data in a school in Switzerland over 7 weeks. We tested saliva samples using a respiratory panel RT-PCR (24 respiratory viruses) and sequenced SARS-CoV-2, IAV, IBV and RSV positive (targeted Illumina panel).

Schematic overview of the collected and input data for the pairwise survival model. (1) We detected virus infections in saliva samples and constructed paired datasets with transmissions that were epidemiologically plausible, excluding pairs where either the infectious or exposed student was absent from school. We also excluded transmissions of influenza A and respiratory syncytial virus through genomic analysis. If an infection could not be linked internally to another participating student, it was considered a transmission from an external source such as a household or community member. For each student pair, we determined the time in close proximity during the exposure period, which was collected with wearable sensors worn by students throughout school lessons and break times. For student pairs within classes, we determined time spent in shared classrooms and in suboptimal air quality, which was measured with air quality monitors and aerosol devices (2). The exposure period is from the onset of infectiousness of student i until the date of infection of student j. If j is never infected, the exposure period is until the end of the infectious period of i. School-free days and absences are not included in the exposure period. The association of respiratory virus transmission with time in close proximity and other risk factors was estimated using accelerated failure time regression models, with an internal and external rate of transmission.

1 1

James Munday @jdmunday.bsky.social · Aug 22

New Preprint! Co-led with Nicolas Banholzer *The relative contribution of close-proximity contacts, shared classroom exposure and indoor air quality to respiratory virus transmission in schools* on MedRxiv. doi.org/10.1101/2025... (YET TO BE PEER REVIEWED)

The relative contribution of close-proximity contacts, shared classroom exposure and indoor air quality to respiratory virus transmission in schools.

Close-proximity interactions are considered a key risk factor for respiratory virus transmission, but their importance relative to shared space and air quality remains unclear. We conducted a six-week...

doi.org

1 7 13

James Munday @jdmunday.bsky.social · Jul 12

Just when you think you are beginning to understand a sport...

3

Reposted by James Munday

Ed Hill @edmhill.bsky.social · Jun 17

📣 Funded 3.5yr PhD opportunity (UK tuition fee rate) in @hpruezi.bsky.social @liverpooluni.bsky.social. #PhDSky

🧪 On "Optimising testing & control strategies in the early stages of infectious disease outbreaks" #IDSky #EpiSky

⏳ Deadline to apply: 07 Jul 2025

🔗: www.liverpool.ac.uk/courses/opti...

Optimising testing and control strategies in the early stages of infectious disease outbreaks

Supervisors: Dr Emily Nixon, Professor Christl Donnelly, Dr Richard Vipond, Dr Emily Adams, Dr Chris Overton.

Subject area: Mathematics

Stipend Amount: Funded studentship
Study mode: Full-time
Duration: 3.5 years

Application deadline: 07 July 2025
Start date: 01 October 2025

Overview
Develop models to help make decisions early in a major disease outbreak when availability of new tests is limited, on where and how best to use them for maximum impact in controlling spread, e.g. in low versus high areas of transmission, in hospitals or in the community. These will depend on the pathogen type, its mode of transmission, reproduction number, etc.

About this opportunity
In the early stages of a major disease outbreak, the availability of suitable diagnostic tests is often limited, however, timely interventions are critical for controlling the spread of disease. In this PhD, you will develop mechanistic and statistical models to investigate the effectiveness of various diagnostic test deployment strategies for emerging and zoonotic diseases which are at risk of causing large outbreaks, such as respiratory viruses (Influenza, SARS-CoV-2, Respiratory syncytial virus), vector-borne diseases (Dengue, Tick-borne encephalitis) or viral haemorrhagic fevers (Nipah, Lassa fever and Crimean-Congo haemorrhagic fever). You will investigate the effectiveness of targeting use of tests in different settings, for example in areas of high versus low transmission, or in hospitals versus the community. The optimum strategy for diagnostic test deployment will vary depending on prevalence of disease, the sensitivity and specificity of the tests and the time it takes to receive and act on results. In addition, variation between pathogens, for example, in their mode of transmission, reproduction number and incubation period, will influence which strategy is the most effective in controlling the impact of the outbreak. These insights will inform stockpiling strategies of medical countermeasures for pandemic preparedness and in the event of a major outbreak, can be used and adapted to help inform decision making.

Who is this for?

Applicants must hold, in a relevant STEM subject, either a first-class honours degree, a distinction at master level, or equivalent achievements. A background in programming (such as R, Python, Julia) would be advantageous, along with a keen desire to develop those skills.

Prospective applicants are encouraged to contact the primary supervisor, Dr Emily Nixon, emily.nixon@liverpool.ac.uk prior to preparation of an application to discuss the fit of the project with your background and qualifications.

Funding your PhD

This PhD opportunity is funded by the National Institute for Health and Care Research (NIHR) Health Protection Research Unit (HPRU) in Emerging and Zoonotic Infections (EZI).

This studentship will be for a maximum of 3.5 years duration. The studentship includes tuition fees at the UK/home rate, stipend and research-related travel. International students may apply but must be able to themselves cover the additional tuition fee costs as outlined on the University website: https://www.liverpool.ac.uk/study/fees-and-funding/tuition-fees/postgraduate-research/.

The supervisory team have no additional funding to support this difference in tuition fee costs for international students. The stipend amount is aligned with UK Research and Innovation rates. For the academic year 2025-2026, the rate of the UKRI Stipend is £20,780 per annum (tax free).

1 9 10