In fcrawford/covid19_icu: COVID-19 ICU dynamics

---

title: "Scenario Report: COVID-19 Hospital Capacity" date: "Generated on r format(Sys.time(), '%d %B, %Y')" output: pdf_document: df_print: kable fontsize: 11pt geometry: margin=1in params: t: NA young: NA medium: NA I_init: NA I_final: NA distribution: NA doublingtime: NA rampslope: NA M: NA L: NA L_occupied: NA M_occupied: NA Lfinal: NA Lramp: NA Mfinal: NA Mramp: NA avg_LOS_ICU: NA avg_LOS_Floor: NA p_death_ICU2: NA p_death_ICU3: NA p_death_floor2: NA p_death_floor3: NA doprotocols: NA dynamicModel: NA ed_visits_timeseries: NA L_final: NA M_final: NA

---

This report includes the parameters values, plot outputs, and summary of key results from the model, given user-defined inputs. Model outputs are completely dependent on user-defined inputs and depend on assumptions encoded in the model structure. Please familiarize yourself with the model structure and inputs for accurate interpretation of these outputs. The model structure can be found on the "About" tab of the COVID-19 Hospital Capacity online webapp (https://forrestcrawford.shinyapps.io/covid19_icu/).

Key points

params_rmd =update_inputs(t=params$t,
                    young= params$young,
                    medium =params$medium,
                    #######################
                    I_init= params$I_init,
                    I_final=params$I_final,
                    distribution =params$distribution,
                    doublingtime=params$doublingtime,
                    rampslope=params$rampslope,
                    #######################
                    M=params$M,
                    L=params$L,
                    L_occupied=params$L_occupied,
                    M_occupied=params$M_occupied,
                    Lramp=params$Lramp,
                    Mramp=params$Mramp,
                    ######################
                    avg_LOS_ICU=params$avg_LOS_ICU,
                    avg_LOS_Floor=params$avg_LOS_Floor,
                    #####################
                    p_death_ICU2= params$p_death_ICU2,
                    p_death_ICU3= params$p_death_ICU3,
                    p_death_floor2=params$p_death_floor2,
                    p_death_floor3=params$p_death_floor3,
                    #####################
                    ed_visits_timeseries=  params$ed_visits_timeseries,
                    #####################
                    L_final=params$L_final,
                    M_final=params$M_final)

dynamicModel = 0

txt <- text_hospital(doprotocols=params$doprotocols,
                     params = params_rmd,
                     dynamicModel=dynamicModel)



    rownames(txt) = txt$Variable

    txt$Value = as.character(txt$Value)

    txt2 = txt

      if (txt2["Days to floor overflow","Value"] == "No shortage"){
      txt2["Days to floor overflow","Value"] = "a time beyond the simulation"
    } else {txt2["Days to floor overflow","Value"] = paste(txt2["Days to floor overflow","Value"], " days")}

    if (txt2["Days to ICU overflow","Value"] == "No shortage"){
      txt2["Days to ICU overflow","Value"] = "a time beyond the simulation"
    } else {txt2["Days to ICU overflow","Value"] = paste(txt2["Days to ICU overflow","Value"], " days")}

Under specified parameters, capacities, and expansion strategies, the model predicts:

• ICU beds will reach capacity at r txt2["Days to ICU overflow", "Value"].

• Floor beds will reach capacity at r txt2["Days to floor overflow", "Value"].

• The total number of deaths will be r txt2["Total deaths of COVID19+ patients", "Value"].

• A hospital case-fatality rate of r txt2["Case fatality ratio for COVID19+ patients", "Value"].

• Accommodating all COVID19+ patients will require r txt2["Extra floor beds needed for COVID19+ patients", "Value"] extra floor beds and r txt["Extra ICU beds needed for COVID19+ patients", "Value"] extra ICU beds.

Parameter Values

These are the values set by the user to describe the number of COVID19+ patients presenting to the health care system, health system capacity, and features of the patient population.

txt_params = text_parameters(doprotocols = params$doprotocols,
                             params = params_rmd,
                             dynamicModel = dynamicModel)

rownames(txt_params) <- NULL

knitr::kable(txt_params)

Plots

plots<-plot_hospital(doprotocols = params$doprotocols,
                             params = params_rmd,
                             dynamicModel = dynamicModel)

This is a plot of user-defined inputs to the simulation: the number of COVID19+ cases presenting to the health system each day during the simulation.

    plot(plots[[1]])

\newpage

This is a plot of cumulative COVID19+ presentations and deaths at each time point during the simulation.

    plot(plots[[2]])

This plot illustrates the number of deaths which will occur during the simulation. The total number of deaths is broken down by location in the hospital.

    plot(plots[[3]])

\newpage

This plot illustrates the number of patients who are projected to utilize ICU or floor beds during the simulation. If floor and ICU capacity are exceeded, a label appears which indicates the day on which this occurs.

    plot(plots[[4]])

Summary

    rownames(txt) = c()
    names(txt) = c()
    knitr::kable(txt)

\newpage

About the application

This app implements a model for utilization of ICU beds, floor beds, and patient deaths due to COVID-19. The model was developed by the COVID-19 Statistics, Policy modeling and Epidemiology Collective (C-SPEC). The app was written by Soheil Eshghi, Margret Erlendsdottir, Maile Thayer Phillips, Suzan Iloglu, Christian Testa and Forrest W. Crawford using the R shiny framework.

We are especially grateful to Gregg Gonsalves, David Paltiel, Hanna Ehrlich, Raphael Sherak, Melanie Chitwood, Thomas Thornhill, Nicole Swartwood, and Stephanie Perniciaro for advice and comments.

Direct queries to [e-mail].

Source code at https://github.com/fcrawford/covid19_icu.

Report issues to the Github issues page.

fcrawford/covid19_icu documentation built on Nov. 21, 2020, 12:14 p.m.