A systematic review of COVID-19 epidemic models with endogenous human behaviour. What's next?
Pith reviewed 2026-06-27 10:55 UTC · model grok-4.3
The pith
COVID-19 models expanded their use of data but rarely captured how behavior changes in response to the epidemic itself.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
The COVID-19 pandemic saw great strides in terms of the expanded use of empirical data in epi-behavioural modelling. However, it also showed shortcomings with respect to limited use of behavioural empirical data, lack of innovation in model structure, and limited engagement with other disciplines and decision-makers. Overall, our results suggest that identifying priorities in model design and behavioural data, building an adequate data collection infrastructure, leveraging on AI advancements, and fostering interdisciplinarity are strategies of utmost importance for pandemic preparedness.
What carries the argument
Systematic review of epidemic models that endogenously incorporate human behaviour responding to epidemic dynamics.
If this is right
- Model design priorities should focus on better integration of behavioural observations.
- Dedicated infrastructure for collecting behavioural data would support more accurate models.
- AI tools can help process behavioural data and improve model performance.
- Greater exchange across disciplines and with decision-makers would make models more usable.
Where Pith is reading between the lines
- Real-time mobility or survey data streams could be linked directly to model parameters to reduce reliance on static assumptions.
- Standard compartmental structures may need replacement by network or agent-based forms that allow behaviour to alter contact patterns dynamically.
- Policy users might test model outputs against observed behaviour during past waves to identify which modelling choices matter most for decisions.
Load-bearing premise
The review's search terms and inclusion rules identified a representative set of models rather than a biased or incomplete subset of the literature.
What would settle it
A follow-up review that locates and includes a large body of models relying on detailed behavioural surveys and introducing new structural forms would contradict the reported shortcomings.
read the original abstract
Human behaviour and epidemic dynamics are intertwined, yet accounting for this feedback remains one of the key challenges of epidemiological modelling. The COVID-19 pandemic was an opportunity to overcome the traditional limitations of the field, raising expectations that data-informed endogenous approaches to behaviour modelling would advance substantially. To quantify the progresses made, we conducted a systematic review of SARS-CoV-2 transmission models endogenously including human behaviour in response to epidemic dynamics. The COVID-19 pandemic saw great strides in terms of the expanded use of empirical data in epi-behavioural modelling. However, it also showed shortcomings with respect to limited use of behavioural empirical data, lack of innovation in model structure, and limited engagement with other disciplines and decision-makers. Overall, our results suggest that identifying priorities in model design and behavioural data, building an adequate data collection infrastructure, leveraging on AI advancements, and fostering interdisciplinarity are strategies of utmost importance for pandemic preparedness.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The manuscript is a systematic review of SARS-CoV-2 transmission models that endogenously incorporate human behaviour in response to epidemic dynamics. It reports expanded use of empirical data during the COVID-19 pandemic but identifies shortcomings in the limited incorporation of behavioural empirical data, lack of innovation in model structures, and limited engagement with other disciplines and decision-makers. It concludes by recommending priorities in model design and data collection, infrastructure building, AI leverage, and interdisciplinarity for pandemic preparedness.
Significance. If the review's sample is representative and its synthesis accurate, the work would usefully document field-wide patterns in epi-behavioural modelling and highlight actionable gaps for future preparedness. The explicit call for data infrastructure and interdisciplinarity could inform funding and collaboration priorities, provided the underlying evidence base is robust.
major comments (1)
- [Methods] Methods section: the search strategy (databases, keywords, date ranges), operational definition of 'endogenous human behaviour', inclusion/exclusion criteria, and any sensitivity or inter-rater analyses are not described with sufficient detail or PRISMA-style reporting. This directly affects the ability to evaluate whether the reviewed models constitute a representative sample or whether classes such as reinforcement-learning, network-based, or non-compartmental structures were systematically omitted, which is load-bearing for the claims of limited innovation and limited behavioural-data use.
Simulated Author's Rebuttal
We thank the referee for their detailed and constructive review. We address the major comment on the methods section below and will revise the manuscript accordingly to improve transparency and reproducibility.
read point-by-point responses
-
Referee: [Methods] Methods section: the search strategy (databases, keywords, date ranges), operational definition of 'endogenous human behaviour', inclusion/exclusion criteria, and any sensitivity or inter-rater analyses are not described with sufficient detail or PRISMA-style reporting. This directly affects the ability to evaluate whether the reviewed models constitute a representative sample or whether classes such as reinforcement-learning, network-based, or non-compartmental structures were systematically omitted, which is load-bearing for the claims of limited innovation and limited behavioural-data use.
Authors: We agree that the methods section requires substantially more detail and PRISMA-style reporting to allow proper evaluation of sample representativeness and to support the claims regarding limited innovation and behavioural data use. In the revised version we will add: (i) a complete PRISMA flow diagram with numbers at each stage, (ii) explicit search strings, databases searched, date ranges, and inclusion/exclusion criteria, (iii) the precise operational definition used for 'endogenous human behaviour', and (iv) any inter-rater agreement or sensitivity checks performed. These additions will directly address concerns about omitted model classes (e.g., reinforcement-learning or network-based structures) and strengthen the evidential basis for our conclusions. revision: yes
Circularity Check
Systematic review exhibits no circularity
full rationale
This paper is a systematic review synthesizing existing SARS-CoV-2 transmission models that incorporate endogenous human behaviour. It presents no mathematical derivations, equations, fitted parameters, predictions, or model structures of its own. All claims are framed as synthesis of cited prior literature, with no self-definitional reductions, fitted-input predictions, or load-bearing self-citation chains that could collapse a derivation to its inputs by construction. The central findings on strides and shortcomings are presented as observational summaries rather than derived results.
Axiom & Free-Parameter Ledger
Reference graph
Works this paper leans on
-
[1]
“Irrational Behaviour-Induced Suppression, Enhancement and Oscillations of Voluntary Vaccination: A Stochastic Model with Delays.” Physica D: Nonlinear Phenomena 489 (May): 135143. https://doi.org/10.1016/j.physd.2026.135143. Capasso, Vincenzo, and Gabriella Serio. 1978. “A Generalization of the Kermack- McKendrick Deterministic Epidemic Model.” Mathemati...
-
[2]
Coupled Contagion Dynamics of Fear and Disease: Mathematical and Computational Explorations
https://doi.org/10.1093/rfs/hhab040. Epstein, Joshua M., Jon Parker, Derek Cummings, and Ross A. Hammond. 2008. “Coupled Contagion Dynamics of Fear and Disease: Mathematical and Computational Explorations.” PLOS ONE 3 (12): e3955. https://doi.org/10.1371/journal.pone.0003955. Funk, Sebastian, Erez Gilad, Chris Watkins, and Vincent A. A. Jansen. 2009. “The...
-
[3]
infection function
with governmental action and individual reaction 𝛽(𝑡) = 𝛽0(1 − 𝜌)𝑒𝑥𝑝 (1 − D N) 𝜅 𝛽(𝑡), 𝛽0 ≔ perceived risk-varying and baseline transmission rate, respectively 𝐷 ≔ public risk awareness evolving according to 𝐷𝛼𝐷 = 𝑑𝛾𝐼 − 𝜆𝐷 𝜅 ≔ strength of public response 𝜌 ≔ governmental action strength Das et al. 2022 A Fractional Order Model to Study the Effectiveness o...
2022
-
[4]
outbreak in Wuhan, China with individual reaction and governmental action 𝛽(𝑡) = 𝛽0(1 − 𝛼)(1 − 𝐷 𝑁) 𝜅 𝛽(𝑡), 𝛽0 ≔ perceived risk-varying transmission and baseline transmission rate, respectively 𝐷 ≔ public risk awareness evolving according to 𝐷̇ = 𝑑𝛾𝐼 − 𝜆𝐷 𝜅 ≔ strength of public response 𝛼 ≔ governmental action strength 𝐼 ≔ infectious individuals Paper Tit...
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.