SHIVIR - An Agent-Based Model to assess the transmission of COVID-19 in India

Background COVID-19 has tormented the global health and economy like no other event in the recent past. Researchers and policymakers have been working strenuously to end the pandemic completely. Methodology/Principal Findings Infectious disease dynamics could be well-explained at an individual level with established contact networks and disease models that represent the behaviour of the infection. Hence, an Agent-Based Model, SHIVIR (Susceptible, Infected, Admitted, ICU, Ventilator, Recovered, Immune) that can assess the transmission dynamics of COVID-19 and the effects of Non-Pharmaceutical Interventions (NPI) was developed. Two models were developed using to test the synthetic populations of Rangareddy, a district in Telangana state, and the state itself respectively. NPI such as lockdowns, masks, and social distancing along with the effect of post-recovery immunity were tested across scenarios. The actual and forecast curves were plotted till the unlock phase began in India. The Mean Absolute Percentage Error of scenario MD100I180 was 6.41 percent while those of 3 other scenarios were around 10 percent each. Since the model anticipated lifting of lockdowns that would increase the contact rate proportionately, the forecasts exceeded the actual estimates. Some possible reasons for the difference are discussed. Conclusions Models like SHIVIR that employ a bottom-up Agent-Based Modelling are more suitable to investigate various aspects of infectious diseases owing to their ability to hold details of each individual in the population. Also, the scalability and reproducibility of the model allow modifications to variables, disease model, agent attributes, etc. to provide localized estimates across different places. Author Summary The world has witnessed several infectious disease outbreaks from time to time. COVID-19 is one such event that tormented the life of mankind. Healthcare practitioners, policymakers, and governments struggled enormously to handle the influx of infections and devise suitable interventions. Agent-Based Models that use the population data could cater to these requirements better. Hence, we developed a disease model that represents various states acquired by COVID-19 infected individuals. The contact network among the individuals in the population was defined based on which the simulation progresses. The effect of various Non-Pharmaceutical Interventions such as lockdowns, the use of masks and social distancing along with post-recovery immunity were enacted considering two case studies viz. population of Rangareddy district and Telangana state. The capability of these models to adapt to different input data fields and types make them handy to be tailored based on available inputs and desired outputs. Simulating them using local population data would fetch useful estimates for policymakers. ### Competing Interest Statement The authors have declared no competing interest. ### Funding Statement This study did not receive any funding ### Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines and uploaded the relevant EQUATOR Network research reporting checklist(s) and other pertinent material as supplementary files, if applicable. Yes The synthetic population used, the AnyLogic model along with the input data are available at: https://cloud.anylogic.com/model/7cd10c0c-f1c1-4b8f-9aac-0bf37a45379a?mode=SETTINGS and https://osf.io/utmhg/?view_only=05ac26fc100645be8b1bba6557d606be. The code developed using Python is available at: https://doi.org/10.6084/m9.figshare.19121939.