Evaluating vaccine allocation strategies using simulation-assisted causal modeling

• Published in: Patterns (New York, N.Y.) Vol. 4; no. 6; p. 100739
• Main Authors: Kekić, Armin, Dehning, Jonas, Gresele, Luigi, von Kügelgen, Julius, Priesemann, Viola, Schölkopf, Bernhard
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 09.06.2023; The Author(s); Elsevier
• Subjects: causality; COVID-19; DSML 2: Proof-of-concept: Data science output has been formulated, implemented, and tested for one domain/problem; modeling; SEIR model; vaccine; vaccine allocation; COVID-19; SEIR model; vaccine; modeling; vaccine allocation; DSML 2: Proof-of-concept: Data science output has been formulated, implemented, and tested for one domain/problem; causality
• ISSN: 2666-3899; 2666-3899
• DOI: 10.1016/j.patter.2023.100739

We develop a model to retrospectively evaluate age-dependent counterfactual vaccine allocation strategies against the coronavirus disease 2019 (COVID-19) pandemic. To estimate the effect of allocation on the expected severe-case incidence, we employ a simulation-assisted causal modeling approach that combines a compartmental infection-dynamics simulation, a coarse-grained causal model, and literature estimates for immunity waning. We compare Israel’s strategy, implemented in 2021, with counterfactual strategies such as no prioritization, prioritization of younger age groups, or a strict risk-ranked approach; we find that Israel’s implemented strategy was indeed highly effective. We also study the impact of increasing vaccine uptake for given age groups. Because of its modular structure, our model can easily be adapted to study future pandemics. We demonstrate this by simulating a pandemic with characteristics of the Spanish flu. Our approach helps evaluate vaccination strategies under the complex interplay of core epidemic factors, including age-dependent risk profiles, immunity waning, vaccine availability, and spreading rates. •Evaluated vaccination strategies using causality and infection dynamics simulations•Compared Israel’s implemented strategy for COVID-19 with counterfactual scenarios•Different age groups need prioritization based on target (infections/severe cases)•The modular structure enables easy adaptation of the model to future pandemics Learning about a complex system and simulating alternative scenarios under changed conditions or dynamics is a challenging problem. Consider the time evolution of COVID-19 cases, which depends on a combination of contact patterns, demographics, and vaccination rates. How many severe cases could have been prevented had a different vaccine allocation strategy been implemented? To answer such counterfactual questions, we propose an approach that merges (1) coarse-grained causal modeling, (2) ordinary-differential-equation-based simulation, and (3) domain knowledge, combining the advantages of different modeling paradigms. The resulting hybrid model can be viewed as a “causal digital twin” of the underlying complex system; it captures relevant features thereof and allows reasoning about novel scenarios and interventions. We hope that our hybrid causal approach can inspire modeling for other domains where causal reasoning about a complex system is of interest. Kekić et al. combine causal modeling, infection dynamics simulations, and literature estimates of immunity waning to evaluate vaccine allocation strategies during pandemics. Their study of COVID-19 shows that prioritizing middle-aged groups reduces infections, while prioritizing elderly groups minimizes severe cases most effectively. This work also highlights the impact of waning immunity and timing of vaccination campaigns on vaccine allocation effectiveness, providing insights for pandemic response.