UTLDR: an agent-based framework for modeling infectious diseases and public interventions

Abstract

Due to the SARS-CoV-2 pandemic, epidemic modeling is now experiencing a constantly growing interest from researchers of heterogeneous study fields. Indeed, due to such an increased attention, several software libraries and scientific tools have been developed to ease the access to epidemic modeling. However, only a handful of such resources were designed with the aim of providing a simple proxy for the study of the potential effects of public interventions (e.g., lockdown, testing, contact tracing). In this work, we introduce UTLDR, a framework that, overcoming such limitations, allows to generate "what if" epidemic scenarios incorporating several public interventions (and their combinations). UTLDR is designed to be easy to use and capable to leverage information provided by stratified populations of agents (e.g., age, gender, geographical allocation, and mobility patterns…). Moreover, the proposed framework is generic and not tailored for a specific epidemic phenomena: it aims to provide a qualitative support to understanding the effects of restrictions, rather than produce forecasts/explanation of specific data-driven phenomena.

Keywords: Activity driven networks; Agent-based modelling; Compartmental models; Epidemics.

Fig. 1

(a) In the SEIR model an individual can be in one of four states: (S)usceptible, (E)xposed, (I)nfected or (R)ecovered. Arrows indicate transitions among compartments. (b) The UTR model extends SEIR by introducing the Tested meta-compartment (blocks in green). Testing can be applied to both Exposed and Infected populations and results in transitions to Quarantine compartments (E_T and I_T)

Fig. 2

(a) The UTLR extension includes lockdown compartments (blocks in orange). Restrictions can be applied to Susceptible, Exposed and Infected populations and result in transitions to locked compartments (S_L, E_L and I_L) (b) The UTLDR compartment adds the possibility of differentiating between recovered and immunized population (R), and dead (D)

Fig. 3

(a) The UTLDR module with limited ICU availability includes the severe hospitalized population H_T, that is differentiated to the one with mild symptoms I_T; moreover, population in H_T that can not be treated adequately is placed in F. (b) The UTLDR module with corpse disposal allows the contact of population in S with an infected corpse in D

Fig. 4

(a) The UTLDR model with partial immunity considers the possibility that population in S can be reinfected again. (b) The UTLDR model with vaccination includes a new sub-population V that has (successfully) being vaccinated

Fig. 5

Experiments on the BA model. (a) The simplest SEIR + (b) re-infection allowed transition; (c) Testing and (d) Lockdown scenarios + (e) dead-recovered distinction + (f) ICU availability

Fig. 6

Experiments on the ER model: (a) The simplest SEIR + (b) re-infection allowed transition; (c) Testing and (d) Lockdown scenarios + (e) dead-recovered distinction + (f) ICU availability

Fig. 7

Diffusion trends for the Tuscany case study. All scenarios start with a setup phase of 30 iterations, during which only UTDR compartments are active. (a) A single lockdown of 90 iterations is activated; (b) two consecutive lockdown of length 60 and 30 iterations respectively are activated - separated by 30 iterations of UTDR; (c) The same setting of the previous scenario but the separation among consecutive lockdown is set to 60 iterations; (d) Same setting of (b) but lockdown lengths are switched