THE TIME-DEPENDENT COSTS AND BENEFITS OF SCHOOL CLOSURE DURING AN INFLUENZA PANDEMIC

Sunday, October 24, 2010
Sheraton Hall E/F (Sheraton Centre Toronto Hotel)
Yiting Xue, M.phil, Ivar Sønbø Kristiansen, MD, PhD and Birgitte Freiesleben de Blasio, PHD, University of Oslo, Oslo, Norway

Purpose: To estimate the costs and benefits of school closure during an influenza pandemic and to study the trade-off between the gains and the losses with regards to the timing and duration of the school closure intervention.

Method: We developed two models: a dynamic disease model based on simulation to capture the spread of influenza with or without school closure and an economic model which captures the costs and the benefits of the school closure. The disease model is an SEIR (susceptible, exposed, infectious (symptomatically or asymptomatically) and recovered) compartment model based on the age-specific population of Oslo, the capital of Norway. Since we can not predict the nature of the next pandemic, we tested the model with three different basic reproductive rates (R0). The school closure intervention have 18 different scenarios, depending on when it is started, how long it lasts and whether it will bring voluntary social distancing among the people. The model predicts disease outcomes, including the numbers of out-patients, in-patients and deaths, with and without school closure, and with different R0s and intervention scenarios. The costs of school closure included the administration costs, the value of lost learning time for the students, the lost productivity of parents who have to leave work to take care of their children during school closure, minus the costs avoided as a result of fewer infections. The benefits are the quality-adjusted life years gained among those who escaped from the infection or death as a result of school closure.

Result: The disease model indicated that though school closure does postpone the peak of infection by approximately 10 days, reduced the R0 by approximately 6%, its impact on the final disease outcomes is limited, especially when the duration is less than half of the pandemic period. The impact depends on the assumptions in different scenarios. An eight-week school closure in a pandemic of approximate 100 days reduces the disease outcomes by approximately 10%. However, in all scenarios, productivity loss from parents who have to go home during school closure is the dominating cost (more than 80% of the total cost) with almost 42 000 working parents involved.

Conclusion: Though school closure contributes to reducing the disease burden, its disruption to society imposes a potentially huge cost.   

Candidate for the Lee B. Lusted Student Prize Competition