metadata-UVA-EpiHiper.txt

team_name: UVA Biocomplexity Institute 
model_name: EpiHiper
model_abbr: UVA-EpiHiper
model_version: "2021-12-19"
model_contributors: Jiangzhuo Chen (UVA) <chenj@virginia.edu>, Stefan Hoops (UVA) <shoops@virginia.edu>, Parantapa Bhattacharya (UVA) <parantapa@virginia.edu>, Dustin Machi (UVA) <dm8qs@virginia.edu>, Bryan Lewis (UVA), Madhav Marathe (UVA) 
website_url: Not available
license: cc-by-4.0

methods: We use our agent-based model, called EpiHiper, which computes stochastic transmissions of a disease in a synthetic contact network between individuals and stochastic state transitions within each individual following a disease model. Our disease model is created based on the "COVID-19 Pandemic Planning Scenarios" document prepared by CDC (https://www.cdc.gov/coronavirus/2019-ncov/hcp/planning-scenarios.html) and is stratified by age group.
The disease model is extended to have waning immunity. The immune waning has an exponential distribution with mean of 0.5 year for the time to transition to a partially susceptible state; the age-specific protection on nodes in the partially susceptible state is as specified in the scenario description.
Our simulations are initialized with county level data of prior infections (part of which have waned immunity) and recent daily confirmed case counts and state level prior vaccinations (part of which have waned immunity). Prior infections are derived from confirmed cases with different ascertainment rates for different age groups (smaller for children than for adults).
We run EpiHiper simulations for each state and combine the output to get results of the whole US. The simulations produce daily infections, hospitalizations, and deaths; and each simulation runs for multiple replicates. We aggregate daily data to get weekly data and compute quantiles for each target from the multiple replicates.

modeling_NPI: We have modeled the following NPIs: (i) A fraction of the population chooses to reduce non-essential (shopping, religion, and other) activities. (ii) All K-12 schools are closed from late-December 2021 to the beginning of 2022 and again from mid-June 2022 to late-August 2022; and face masks are used in school while schools are open. (iii) A fraction of symptomatic people choose to self-isolate themselves at home.
compliance_NPI: Compliances are 15% on average for (i), 100% for (ii), and 75% for (iii).
contact_tracing: "Not applicable".
testing: "Not applicable".
vaccine_efficacy_transmission: same as vaccine efficacy against symptoms specified in the scenario description.
vaccine_efficacy_delay: as specified in the scenario description.
vaccine_hesitancy: as specified in the scenario description, configured per state.
vaccine_immunity_duration: vaccinated people go back to a partially immune state after 0.5 years on average.
natural_immunity_duration: recovered (from infections) people go back to a partially immune state after 0.5 years on average.
case_fatality_rate: age stratified, as suggested in the CDC "COVID-19 Pandemic Planning Scenarios" document but reduced for people with immunity according to the specified protection against symptoms.
infection_fatality_rate: age stratified, as suggested in the CDC "COVID-19 Pandemic Planning Scenarios" document but reduced for people with immunity according to the specified protection against symptoms.
asymptomatics: 35% for naively susceptible people (without immunity), increased for people with immunity (from natural infection and/or vaccination).
age_groups: preschool (0-4 years), students (5-17), adults (18-49), older adults (50-64) and seniors (65+).
importations: "Not applicable".
confidence_interval_method: estimated from multiple replicates of simulations.
calibration: For each state, we calibrate the transmissibility parameter in our disease model targeting the state level esimtated effective reproduction number for the last date of fitting data.
spatial_structure: embedded in our synthetic contact network model.