Healthcare worker absenteeism, child care costs, and COVID-19 school closures: a simulation analysis

Elizabeth T Chin; Benjamin Q Huynh; Nathan C Lo; Trevor Hastie; Sanjay Basu

We assume an incubation period of 5.1 days and an infectious period of 6.5 days 26, 27 . The R 0 of SARS-CoV-2 is estimated to be between 2.0 and 6.0, and we examine values within that range (R 0 =2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, 5.5, 6.0). Since the COVID-19 outbreak curve is over a short duration, we ignore births, death, and immigration. We assume that 86% of infections are mild or asymptomatic, with asymptomatic individuals as 50% less infectious than symptomatic individuals 28 . Symptomatic individuals are assumed to reduce contact by 75%. Age-stratified hospitalization rates and infection fatality ratios were obtained from Verity et al 29 . We choose to only apply the infection fatality ratios to symptomatic individuals to obtain a conservative estimate. We assume that individuals develop immunity after recovering from COVID-19 in the short term. To estimate the demand on the healthcare system, we assume that 30% of hospitalizations will require critical care (invasive mechanical ventilation, vasopressor support, or further intensive care-level intervention), and that individuals requiring hospitalization will stay for 8 days and individuals requiring critical care will stay in the ICU for 10.4 days 26 . To simulate the effects on a particular county, we seed the simulation for county age demographics. We estimate the effectiveness of school closures by calculating the reduction in peak ICU bed demand between social distancing and social distancing plus school closure conditions.