Scenario variables

low
moderate
high
Hospitalization rate
ICU admit rate
Ventilator rate
Mortality rate
ICU share of days
AAC LOS
ICU LOS
Days til hosp AAC
Days til hosp ICU ICU

Interventions

Other variables


This app forecasts the number of admitted patients for a hospital at given point of time (hospital census).

Variables on the left can be customized in three scenarios: low, moderate, and high. Users should enter values where low <= moderate <= high. A "low" scenario is one where the pandemic threat is lower than a "moderate" scenario. The Intensive Care Unit (ICU) census is expected to be lower in a "low" scenario than a "moderate" scenario and a "high" scenario is one where the pandemic threat is greatest.

The scenarios are depicted in the plot on the "Hospital Census" tab. The plot shows the ICU census, AAC census, and Total census which is the sum of ICU and AAC census By default, only the Total utilization line (red) is displayed. To view the other trends, click the names in the legend to toggle the desired lines.

The blue line represents the ICU census over time. The green line represents the AAC census over time. The "moderate" scenario is the solid line in the middle and the "low" and "high" scenarios are the edges of the transparent ribbon. The outcome of a scenario that falls between the "low" and "high" scenarios should be somewhere within that ribbon. The ribbon is not a confidence interval in a statistic sense.

Scenario variables:

Hospitalization rate: percent of newly infected patients who are admitted to this hospital
ICU admit rate: percent of newly infected patients who are admitted to the ICU
Ventilator rate: percent of newly infected patients who are admitted to the hospital and require use of a ventilator
Mortality rate: percent of newly infected patients whose outcome is death
ICU share of days: percent of days spent in the ICU
AAC LOS: average length of stay (in days) in Adult Acute Care (AAC) units such as Med/Surg.
ICU LOS: average length of stay (in days) for patients who are admitted to the ICU
Days til hosp AAC: number of days after infection that patient is admitted to AAC
Days til hosp ICU: number of days after infection that patient is admitted to the ICU

Interventions:

This section allows users to customize the dates and estimated impact factor of different policy interventions. Each intervention is assumed to reduce the amount of person-to-person contact and reduce the rate of new infections. For each intervention, users should enter a number between 0 and 1.

For example, if the state instituted social distancing on Feb 28, 2020 and it is estimated that this policy will reduce the infection rate by 15%, then users should enter the date 2020-02-28 with an impact factor of 0.15. As new interventions are introduced, the cumulative impact is expected to increase. Smoothing is applied between intervention dates to make transitions from one intervention to another less abrupt. This smoothing is shown for the "moderate" scenario in the "Intervention Smoothing" tab.

The default scenario assumes that the intial infection took place on 2020-02-01 and that the first intervention took place on 2020-01-28.

Other variables:

Start date: The date of the first infection in the hospital catchment area
End date: The latest date for the prediction time frame
Number of beds: number of beds in the hospital or health system being analyzed, shown on the plot as a dashed gray line
Population size: number of people in the hospital's catchment area
Hospital share of market: percent of people in the region that are served by this hospital
Days til death from infection: for patients whose outcome is death, it is the number of days between infection and death
Quarantine days: number of days to quarantine after exposure to the virus, per the CDC
Expected doubling days: number of days it takes to double the number of new cases (AHA says to expect a doubling time of 7-10 days)


AAC = Adult Acute Care, ICU = Intensive Care Unit, Total = AAC + ICU

Click on the names in the legend to toggle the lines. Bold lines represent the "moderate" scenario and the semitransparent lines represent the "low" and "high" scenarios. The shaded region represents the possible outcomes between the "low" and "high" scenarios.