a Application Case 10.5 Agent-Based Simulation Helps Analyze Spread of a Pandemic Outbreak Knowledge about the spread of a disease plays an mitigation strategies. The simulation models each state important role in both preparing for and respond of an individual in each time unit, based on the indi- ing to a pandemic outbreak. Previous models for vidual probabilities to transition from susceptible state such analyses are mostly homogenous and make to infected stage and then to recovered state and use of simplistic assumptions about transmission hack to susceptible state. The simulation model also and the infection rates. These models assume that uses an individual's duration of contact with infected each individual in the population is identical and individuals. The model also accounts for the rate of typically has the same number of potential contacts disease transmission per time unit based on the type with an infected individual in the same time period of contact between individuals and for behavioral Also each infected individual is assumed to have the changes of individuals in a disease progression sime probability to transmit the disease. Using these (being quarantined or treated or recovered). It is models, implementing any mitigation strategies to flexible enough to consider several factors affecting vaccinate the susceptible individuals and treating the mitigation strategy, such as an individual's age, the infected indivicals become extremely difficult residence, level of general interaction with other under limited resources members of population, number of individuals in each In order to effectively choose and implement a household, distribution of households, and behav. mitigation strategy, modeling of the disease spread oral aspects involving daily commutes, attendance at has to be done across the specific set of individuals, schools, and asymptotic time period of disease. which enables researchers to prioritize the selection The simulation model was tested to measure of individuals to be treated first and also gauge the the effectiveness of a mitigation strategy involving effectiveness of mitigation strategy an advertising campaign that urged individuals who Although nonhomogenous models for spread have symptoms of disease to stay at home rather of a disease can be built based on individual charac than commute to work or school. The model was teristics using the interactions in a contact network, based on a pandemic influenza outbreak in the such individual levels of infectivity and vulnerability greater Toronto area. Each individual agent, gener require complex mathematics to obtain the informaated from the population, was sequentially assigned tion needed for such models to households. Individuals were also assigned to Simulation techniques can be used to generate different ages based on census age distribution; hypothetical outcomes of discase spread by simu all other pertinent demographic and behavioral lating events on the basis of hourly daily, or other attributes were assigned to the individuals. periods and tailying the outcomes throughout the The model considered two types of contact: simulation. A nonhomogenous agent-based simula dose contact, which involved members of the same tion approach allows each member of the population household or commuters on the public transport: to be simulated individually, considering the unique and causal contact, which involved random individual characteristics that affect the transmission individuals among the same census tract. Influenza and infection probabilities. Furthermore, individual pandemic records provided past disease transmis behaviors that affect the type and length of contact sion data, including transmission rates and contact between individuals, and the possibility of infected time for both close and causal contacts. The effect individuals recovering and becoming immune, can of public transportation was simplified with an also be simulated via agent-based models assumption that every individual of working age One such simulation model, built for the used the nearest subway line to travel. An initial Ontario Agency for Health Protection and Promotion outbreak of infection was fed into the model. A total (OAHPP) following the global outbreak of severe of 1,000 such simulations was conducted. acute respiratory syndrome (SARS) in 2002-2003. The results from the simulation indicated that simulated the spread of disease by applying various there was a significant decrease in the levels of infected (Continued Application Case 10.5 (Continued) and deceased persons as an increasing number of 2. List the various factors that were fed into the agent- infected individuals followed the mitigation strategy based simulation model described in the case. of staying at home. The results were also analyzed by 3. Elaborate on the benefits of using agent-based answering questions that sought to verify issues such simulation models. as the impact of 20 percent of infected individuals 4. Besides disease prevention, in which other situa- staying at home versus 10 percent staying at home. tions could agent-based simulation be employed The results from each of the simulation outputs were fed into geographic information system software, What We Can Learn from This Application ESRI ArcGIS, and detailed shaded maps of the Case greater Toronto area, showing the spread of disease based on the average number of cumulative infected Advancements in computing technology allow individuals. This helped to determine the effectiveness for building advanced simulation models that are of a particular mitigation strategy. This agent-based nonhomogeneous in nature and factor for many simulation model provides a what-if analysis tool that socio-demographic and behavioral factors. These can be used to compare relative outcomes of differ simulation models further enhance the support for ent disease scenarios and mitigation strategies and policy decision making by hypothetically simulating help in choosing the effective mitigation strategy many real-time complex problem situations. QUESTIONS FOR DISCUSSION Source: D. M. Aleman, TG. Wibisono, and B. Schwartz, "A Nonhomogeneous Agent-Based Simulation Approach to 1. What are the characteristics of an agent-based Modeling the spread of Disease in a Pandemic Outbreak." simulation model? Interfaces, Vol. 41, No. 3, 2011, pp. 301-315