diff --git a/data-processed/PSI-PROF/2024-04-28-PSI-PROF.gz.parquet b/data-processed/PSI-PROF/2024-04-28-PSI-PROF.gz.parquet new file mode 100644 index 0000000..7fb77ef Binary files /dev/null and b/data-processed/PSI-PROF/2024-04-28-PSI-PROF.gz.parquet differ diff --git a/data-processed/PSI-PROF/metadata-PSI-PROF.txt b/data-processed/PSI-PROF/metadata-PSI-PROF.txt new file mode 100644 index 0000000..3f94422 --- /dev/null +++ b/data-processed/PSI-PROF/metadata-PSI-PROF.txt @@ -0,0 +1,35 @@ +team_name: PSI +model_name: PROF +model_abbr: PSI-PROF +model_contributors: Ben-Nun M (Predictive Science) , Turtle J (Predictive Science) , Riley P (Predictive Science) +website_url: https://predsci.com/usa-flu-hosp/ +model_version: "1.0" +methods: For each state our scenario projections are generated using a mechanistic model with S[Sv]E[Ev]I[Iv]HR compartments, where 'v' subscripts indicate vaccinated. The model is calibrated to the 2023-24 data, and extrapolated through the 2024-25 season. +license: cc-by-4.0 +modeling_NPI: "Not applicable" +compliance_NPI: "Not applicable" +contact_tracing: "Not applicable" +testing: "Not applicable" +vaccine_efficacy_transmission: "One half VE against hospitalization" +vaccine_efficacy_delay: "Not applicable" +vaccine_hesitancy: "Not applicable" +vaccine_immunity_duration: "Waning against hospitalization: 270 days (mean). Immunity escape: scenario specified" +natural_immunity_duration: "Calibrated to each state" +case_fatality_rate: "Not applicable" +infection_fatality_rate: "Calibrated to each state" +asymptomatics: "Not applicable" +age_groups: "0-17, 18-64 low risk, 18-64 high risk, 65+" +importations: We assumed 10 importations per day per state seeded randomly. +confidence_interval_method: "Not applicable" +calibration: "Not applicable" +spatial_structure: "Not applicable" +methods_long: "We use the population and vaccine time series provided by the Scenario Hub and + \ assume that the infection-susceptibility of vaccinated individuals accounts for one half \ + \ of vaccine effectiveness against severe disease. The model is calibrated to 2023-24 season adult \ + \ and pediatric COVID hospital admissions. Calibration assumes that last season fell \ + \ somewhere between scenarios E and F. Additionally, calibration includes a preference for \ + \ periodic/quasi-periodic trajectories. Calibration is a two-stage process. The first stage \ + \ uses a traditional optimizer to find a centroid for the hyper-cube used in the second stage. \ + \ The national projection profiles are calculated as \ + \ the sum of states using pseudo-randomly ordered state level profiles. "\ +