New paper: https://t.co/GrhgSDjBf7. We find that under-ascertainment and delays in reporting new cases have a surprisingly asymmetric impact on epidemic decision-making. It is much harder to know when to intervene than when to relax an intervention.

NEW EPISODE #ScienceInContext! This week, Dr @SabineLvE speaks with Dr @krisparag1 about methods to estimate the instantaneous reproduction number and approaches to estimate uncertainty 👇 @imperialcollege @UniofOxford @ImperialSPH

A model-predictive control approach using noisy real-time case data can optimize timing of epidemic interventions, reducing both outbreak peaks and #Intervention costs compared to preset or threshold-based methods. https://t.co/mTTxBnz235

@AmJEpi @NicSteyn Open access here https://t.co/vCrW5o85SD with an informative visualisation of various approaches (Fig S1 to appear)

New in @AmJEpi, led by @NicSteyn: we show how neglected smoothing parameters in popular Rt estimators can delay outbreak detection or inflate confidence. We derive a new likelihood & marginalisation approach to make Rt estimates more robust and better quantify uncertainty.

Lots of fun working on this study. We take observed disease incidence and infer (well, refute) causal relations. @UriObolski @krisparag1

"A primer on inference and prediction with epidemic renewal models and sequential Monte Carlo" (by Nicholas Steyn, Kris V. Parag, Robin N. Thompson, Christl A. Donnelly):

CLOSING SOON: Postdoc (London, UK) 2 Postdocs in Pathogen Genomics and Virus Genomic Epidemiology. with Oliver Pybus, Sarah Hill, Jayna Raghwani at @RoyalVetCollege More details:

The December issue of SIAM News is now live! In this edition, @science_eye overviews a recent study that derived a simple #mathematical equation to succinctly capture the wing flapping frequency a flying #animal must maintain to remain airborne. Read more: https://t.co/79RnM5SbvR

Oldies but goldies: Isaac Newton, De analysi per aequationes numero terminorum infinitas, 1669. The standard second-order method for root finding. Understanding its global convergence is hard. https://t.co/aoWeI50FB8

This study was inspired by the seminal work on factors behind controllability https://t.co/hBF7kfQK3P and attempts to generalise those factors by leveraging tools from control theory that are not often used in epidemiology.

With this framework we can explore the impact of pre-symptomatic spread, surveillance biases, heterogeneities and differing generation times. We hope this approach can help inform on when targeted vs broader NPIs are needed and highlight the value of modelling feedback loops.

We reframe the epidemic as a positive feedback loop between past and new infections. Targeted NPIs (e.g. quarantines) disrupt that loop. The gain margin generalises R, the delay margin considers lags from surveillance or NPIs and we find that R only works in limited conditions.

Now in @PhysRevX https://t.co/qoFvzZMOtz. We challenge the common use of R for assessing epidemic controllability. Using control theory, we derive epidemic gain and delay margins to better reflect the effort needed to stabilise outbreaks and pinpoint when targeted NPIs fail.

We have Kris Parag speaking on “Closing the feedback loop between infectious disease dynamics and real-time interventions” at the epinowcast seminar tomorrow at 3pm. https://t.co/407FBeVs2P

Seven basic rules for causal inference https://t.co/szjYNFJQzO

Great collab with @RobinNThompson. We develop an event level R0, which includes importation patterns that can be behaviour driven. Builds on interesting work by @CarolineColijn and colleagues

We show how differences in perceived infection risks can boost epidemic superspreading. When less cautious individuals attend riskier events, spread is faster than expected. Data on behaviour and event attendance are integral to assessing outbreak risks.

New preprint led by @sandor_beregi thinking about how we can decide when to relax or implement interventions using optimal control algorithms. We expose how delays, underreporting and policy frequencies can limit achievable epidemic control. https://t.co/uKN5K7msCK

Please RT, the Spatial Epidemiology Lab ( https://t.co/9SrgXZJfJe, @ULBRecherche) is hiring: I am looking for a post-doc to work on spatial models helping to target under-immunised communities during vaccination activities in the DRC: