Great discussion with @billhanage.bsky.social and the audience at #ESCMIDGlobal2026 during our session “Pathogens and people: modelling epidemics for preparedness”. Strong focus on how to improve preparedness in collaboration with public health institutes - lots to take forward.
A panel of immunohistochemistry images showing infection of a panel of IAV viruses (as 6:2 reassortants on a PR8 backbone) in epithelia from the teat and gland cisterns of Aberdeen Angus, Limousin and Holstein Friesian cattle
🚨New pre-print!🚨
Using a panel of different H5N1 clade 2.3.4.4b viruses and human seasonal influenza, and mammary explants from common beef and dairy cattle breeds, we add to the growing data showing that H5N1 spillovers into cattle should be seen as an ongoing risk
www.biorxiv.org/content/10.6...
Delighted to share the (preprint) output of @mhairijan.bsky.social's Oxford visit:
SARS-CoV-2 Neutralising Antibody Profiles Reveal Variant Specific Antibody Dynamics and Regional Differences in Infection Histories in Malawi. dx.doi.org/10.2139/ssrn...
📰 UK HPAI H5N1 case report: 17 Apr 2026 #IDsky #OneHealth🧪
🐔 Large commercial poultry unit near Gainsborough, Lincolnshire.
📊 100 confirmed HPAI H5N1 cases in poultry & other captive birds from 01 Oct 2025
🏴 England: 79
🏴 Scotland: 9
🏴 Wales: 7
🇬🇧 Northern Ireland: 5
🔗: www.gov.uk/government/n...
🎓 One last PhD studentship on offer on health analytics and modelling, jointly between @hpruham.bsky.social and @hprurespinf.bsky.social. Real-time use of data from household studies to inform pandemic and epidemic preparedness, with Anne Cori, @seabbs.bsky.social and Chris Overton. Link ↓
❓ Interested in contributing to the Global Society for Infectious Disease Dynamics (GSIDD; @gs-idd.bsky.social) mission, values & strategic plan?
📧 Sign up to the GSIDD mailing list to receive updates: www.gsidd.org/join-us-old
👇 View GSIDD mission, values & strategic plan details.
New UK study shows the MVA-BN #mpox vaccine generates antibodies against emerging clade Ib, though at lower levels than clade IIb.
Findings highlight the importance of vaccination, surveillance, and ongoing research as the virus evolves.
Read more: www.thepandemicinstitute.org/news/study-s...
Postdoctoral Research Associate in Public Engagement (Part time) – Grade 7 University of Liverpool - Faculty of Health and Life Sciences / Institute of Population Health / Department of Public Health / Policy and Systems We are seeking a highly motivated researcher to join a multi-disciplinary team to engage with pregnant women, families and stakeholders to deliver and evaluate public engagement activities across the Liverpool City Region to build on the Young Voices, Big Ideas project and foster ownership of the Children Growing Up in Liverpool birth cohort study in the wider community, and understand how people perceive research into health inequalities and data. You will work collaboratively as part of a wider team, with specific responsibility to co-create a series of arts-based engagement activities and events to engage with expectant and new parents, children and young people and their families, particularly those from minoritized and marginalised groups. You should have demonstrable skills in public engagement, ideally with a focus on creative activities, and engaging research methods (e.g. participatory video, photovoice, participatory workshops, Lego Serious Play). Experience of writing research outputs, such as journal articles and reports, creating or commissioning creative outputs, and good project management skills are required. Ideally you will have experience of and/or be familiar with patient and community engagement in applied health care research and the literature on health inequalities. You should have a PhD degree in relevant areas of Public Health or relevant disciplines e.g. sociology, psychology, anthropology, performing arts, arts-based health research and experience of organising and conducting research or public engagement in a community setting. The post is available until 31 August 2027 and is part time (0.6 FTE). Closing Date: 27 April 2026 23:30
🚨 Hiring: Postdoctoral Research Associate in Public Engagement - engage with pregnant women, families & stakeholders across Liverpool City Region! #AcademicJobs #PublicHealth
💰 Part time role (0.6FTE) until 31 Aug 2027
📅 Apply by 23:30 BST, 27 Apr 2026
🔗 to apply: my.corehr.com/pls/ulivrecr...
Excited to advertise a fully funded PhD with @socialinfluenza.bsky.social investigating how influenza virus evolution differs between different avian hosts (and how this impacts pandemic potential). Joint between @pirbrightinst.bsky.social + @cvrinfo.bsky.social
www.findaphd.com/phds/project...
It's not gone: #USDA reports 5 more dairy herds in Idaho have tested positive for #H5N1 #birdflu, the first in months.
In the 2+ years since the virus was first detected in cows in the US, 1,093 herds in 19 states have tested positive for the virus. www.aphis.usda.gov/livestock-po...
📰 UK HPAI H5N1 reported cases: 14 Apr 2026 #IDsky #OneHealth 🧪
🐔 Two poultry premises (near Gainsborough, Lincolnshire; near Great Shelford, Cambridgeshire)
📊 99 confirmed HPAI H5N1 cases in poultry & other captive birds from 01 Oct 2025
🏴 England: 78
🏴 Scotland: 9
🏴 Wales: 7
🇬🇧 Northern Ireland: 5
Treatment-specific mean utility and regret weights from the mixed logit and mixed RRM models, pooled Italy and France samples.
📣 New pre-print out!
➡️ Dengue risk perception and public preferences for vector control in Italy and France: utility and regret-based choice experiments: www.medrxiv.org/content/10.6...
A great collaboration with @gaveltri.bsky.social and @filippoandrei.bsky.social
📰 UK HPAI H5N1 reported cases: 11 Apr 2026 #IDsky #OneHealth🧪
🐔 Commercial poultry near Market Rasen, Lincolnshire.
📊 97 confirmed HPAI H5N1 cases in poultry & other captive birds from 01 Oct 2025
🏴 England: 76
🏴 Scotland: 9
🏴 Wales: 7
🇬🇧 Northern Ireland: 5
🔗: www.gov.uk/government/n...
Great speakers at this symposium - 22 April 2026 at LSHTM. Join online or in person (email if in person) to hear great talks at the cutting edge of Inequalities in infectious disease dynamics. www.lshtm.ac.uk/newsevents/e...
13/ Study info
📄 Estimating the strength of symptom propagation from primary-secondary case pair data
✍️ Authors: Phoebe Asplin | Rebecca Mancy | @mattkeeling.bsky.social | @edmhill.bsky.social.
🔗 Preprint: doi.org/10.64898/202...
💻 Study code: github.com/pasplin/symp...
/end
12/ 📌 Study limitations 📌
▶️ Potential biases in household data: Many individuals are likely to be genetically related. These genetic similarities may cause correlations in symptom severity.
[NB: Preprint, not yet peer-reviewed]
11/ 📌 Study limitations 📌
▶️ Limitations when using household data compared to contract tracing data: Household data has higher uncertainty regarding who infected whom, or even whether the secondary case resulted from within-household transmission.
[NB: Preprint, not yet peer-reviewed]
10/ 📌 Study limitations 📌
▶️ Two of the data sets in our real-world data analysis consisted of purely household data, as opposed to contact tracing data.
[NB: Preprint, not yet peer-reviewed]
9/ 💡 Why This Matters 💡
✅ Helping inform data reporting: Despite the low data requirements of our methodology, freely available data of this type is currently rare. Encourage reporting of these statistics to enable scientific analysis on symptom propagation.
[NB: Preprint, not yet peer-reviewed]
8/ 💡 Why This Matters 💡
✅ Methodological advancement: Robustly estimate symptom propagation strength using data on primary-secondary case pairs.
✅ Public health applications: Practical tool to study symptom propagation across a range of pathogens & settings.
[NB: Preprint, not yet peer-reviewed]
Figure 4: Error in the estimated values of alpha and the percentage of replicates with support for symptom propagation, using an age-dependent methodology. The top row shows the error in the estimates of alpha. A positive error corresponds to the estimated value being an overestimate of the true value of alpha. The solid grey regions correspond to infeasible errors, which are present due to the restriction of alpha being between 0 and 1. The points show the median errors and the lines show the interpolation between those points. The shaded area spans the 2.5th percentile to the 97.5th percentile (the 95% uncertainty interval). The bottom row gives the percentage of replicates (out of 1000) for which alpha=0 was not within the 95% confidence region (i.e. support for symptom propagation). Across panels, we vary the assessed sample sizes n (i.e. the number of primary-secondary case pairs we had data for in each replicate): (first column) 10, (second column) 100, (third column) 1,000.
7/ 🔍 Key Findings: Synthetic data analysis 🔍
3️⃣ Very rarely found false positives (i.e. support for symptom propagation when it was not present in the underlying data).
[NB: Preprint, not yet peer-reviewed]
Figure 2: 95% confidence regions for the predicted value of nu and alpha for five replicates with varying value of n (the number of primary-secondary case pairs we had data for in each replicate): (first column) 10, (second column) 100, (third column) 1,000. The cyan triangles indicate the MLE estimates for each replicate. The red dots correspond to the true values of alpha. The true value of alpha is varied across the rows: (top row) alpha = 0.2, (middle row) alpha = 0.5, (bottom row) alpha = 0.7.
Figure 3: Error in the estimated values of alpha and the percentage of replicates with support for symptom propagation. We consider four reporting bias scenarios: no reporting bias (black circles), primary case reporting bias (red triangles), secondary case reporting bias (blue squares) and primary and secondary case reporting bias (purple diamonds). The top row shows the error in the estimates of alpha. A positive error corresponds to the estimated value being an overestimate of the true value of alpha. The solid grey regions correspond to infeasible errors, which are present due to the restriction of alpha being between 0 and 1. The points show the median errors and the lines show the interpolation between those points. The shaded area spans the 2.5th percentile to the 97.5th percentile (the 95% uncertainty interval). The bottom row gives the percentage of replicates (out of 1000) for which alpha=0 was not within the 95% confidence region (i.e. support for symptom propagation). Across panels, we vary the assessed sample sizes n (i.e. the number of primary-secondary case pairs we had data for in each replicate): (first column) 10, (second column) 100, (third column) 1,000.
6/ 🔍 Key Findings: Synthetic data analysis 🔍
1️⃣ Accurate estimates for strength of symptom propagation strength attained with 100 to 1000 primary-secondary case pairs.
2️⃣ Methodology retained high accuracy to severity-dependent biases in case reporting.
[NB: Preprint, not yet peer-reviewed]
Figure 5: 95% confidence regions for the predicted value of nu and alpha for the three real-world data sets: England households, Israel households, Norway contact tracing. Here, n is the number of primary-secondary case pairs in the data set. Cyan triangles indicate the MLE estimates for each replicate, and the grey shaded regions show the 95% bivariate confidence regions.
Figure 6: Posterior distributions for alpha and the four age-dependent nu parameters for two real-world data sets: England households (top row) and Israel households (bottom row). The cyan triangles indicate the posterior means and the grey shaded regions indicate the 95% credible intervals. Note that the age groups corresponding to each age-dependent nu estimate vary between the two household data sets: 0-10, 11-18, 19-54 and 55+ for England, and 0-9, 10-19 and 20+ for Israel. Therefore, there is no nu_4 parameter for the Israel households data.
Table 4: Estimates for alpha and nu for two real-world data sets, comparing age-free vs age-dependent results. Here we give the values of the maximum likelihood estimates (MLEs) and their 95% confidence intervals for the age-free (MLE) estimates and the posterior means and 95% credible intervals for the age-free (MCMC) and age-dependent (MCMC) estimates. Note that the age groups corresponding to each age-dependent nu estimate vary between the two household data sets: 0-10 year olds, 11-18 year olds, 19-54 year olds and 55+ year olds for England, and 0-9 year olds, 10-19 year olds and 20+ year olds for Israel. Here, n is the number of primary-secondary case pairs in the data set.
5/ 🔍 Key Findings: SARS-CoV-2 analysis 🔍
1️⃣ Age-free methodology: Indicated a 12-17% increase in risk of being symptomatic if infected by someone symptomatic.
2️⃣ Age-dependent methodology: Positive estimates for the strength of symptom propagation persisted.
[NB: Preprint, not yet peer-reviewed]
Table 1: Epidemiological parameter values for age-free data.
4/ 📊 Synthetic data analysis 📊
✅ To verify the impact of different assumptions regarding data availability (e.g. volume of data available).
🔍 Investigate the robustness of estimates to severity-dependent biases in case reporting.
[NB: Preprint, not yet peer-reviewed]
Table 2: Summary case data used for parameter estimation stratified by secondary case age group. We provide the number of primary-secondary case pairs (n) from each study used in our parameter estimation, stratified by the age group of the secondary case, along with the symptom breakdown of all the pairs. Columns headed X -> Y give the number of pairs where the primary case has symptom status X and the secondary case has symptom status Y (e.g. A ->A is the number of pairs where both the primary and secondary cases were asymptomatic).
Table 3: Details of the vaccination status, age group and variant for the England and Israel household data sets. Note that the Norway contact tracing data set is omitted here as we did not have access to this level of demographic information.
3/ 🌍 Real-world data analysis: SARS-CoV-2 🌍
💻 Estimated strength of symptom propagation from 3️⃣ public datasets collected during the COVID-19 pandemic with data on symptom presence/absence.
🏴 England households
🇮🇱 Israel households
🇳🇴 Norway contact tracing data
[NB: Preprint, not yet peer-reviewed]
Figure 1: Schematic showing how symptom severity is determined in the symptom propagation model according to the two parameters characterising symptom propagation, alpha and nu. White-shaded individuals correspond to those susceptible to infection, yellow-shaded individuals correspond to infectious cases with mild symptoms, and red-shaded individuals correspond to infectious cases with severe symptoms. The values on the arrows show the corresponding probability. An infected individual has probability alpha of copying the symptom severity of their infector and a probability 1-alpha of reverting to the baseline probability of having severe disease, i.e. they develop severe disease with probability nu.
2/ 🔬 Scientific overview 🔬
⚠️ Methods gap: Limited previous attempts to quantify presence of and/or strength of symptom propagation.
🚧 Need to account for case severity-dependent biases in household & contact tracing studies: Due to self‐selecting volunteers & participants self-reporting contacts.
1/ ❓ Interested in a methodology to estimate the strength of symptom propagation from a relatively small number of primary-secondary case pair data? 🧪 #IDSky #EpiSky
📄 See our preprint “Estimating the strength of symptom propagation from primary-secondary case pair data"
🔗: doi.org/10.64898/202...
Can routinely collected antibody data from an acute febrile illness clinical surveillance platform serve as a proxy for estimating population immunity?
New paper with @enilles.bsky.social et al suggests surveillance data can indeed be useful alternative approach: wwwnc.cdc.gov/eid/article/...
Image of a duck farm with a person in the background.
The US Department of Agriculture’s Animal and Plant Health Inspection Service (APHIS) reported more avian flu activity at commercial poultry facilities in Indiana, which has seen high levels of H5N1 activity this spring.
Read more: ow.ly/vXnG50YGLS5
