COVID-19 Utilization and Resource Visualization Engine (CURVE) to Forecast In-Hospital Resources

Abstract

Background

The emergence of COVID-19 has created an urgent threat to public health worldwide. With rapidly evolving demands on healthcare resources, it is imperative that healthcare systems have the ability to access real-time local data to predict, plan, and effectively manage resources.

Objective

To develop an interactive COVID-19 Utilization and Resource Visualization Engine (CURVE) as a data visualization tool to inform decision making and guide a large health system’s proactive pandemic response.

Methods

We designed and implemented CURVE using R Shiny to display real-time parameters of healthcare utilization at Atrium Health with projections based upon locally derived models for the COVID-19 pandemic. We used the CURVE app to compare predictions from two of our models –one created before and one after the statewide stay-at-home and social distancing orders (denoted before- and after-SAH-order model). We established parameter settings for best-, moderate-, and worst-case scenarios for pandemic spread and resource use, leveraging two locally developed forecasting models to determine peak date trajectory, resource use, and root mean square error (RMSE) between observed and predicted results.

Results

CURVE predicts and monitors utilization of hospital beds, ICU beds, and number of ventilators in the context of up-to-date local resources and provides Atrium Health leadership with timely, actionable insights to guide decision-making during the COVID-19 pandemic. The after-SAH-order model demonstrated the lowest RMSE in total bed, ICU bed, and patients on ventilators.

Conclusions

CURVE provides a powerful, interactive interface that provides locally relevant, dynamic, timely information to guide health system decision making and pandemic preparedness.

SciScore for 10.1101/2020.05.01.20087973: (What is this?)

Please note, not all rigor criteria are appropriate for all manuscripts.

Table 1: Rigor

Institutional Review Board Statement	not detected.
Randomization	not detected.
Blinding	not detected.
Power Analysis	not detected.
Sex as a biological variable	not detected.

Table 2: Resources

No key resources detected.

Results from OddPub: We did not detect open data. We also did not detect open code. Researchers are encouraged to share open data when possible (see Nature blog).

Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:

There are some limitations and scope conditions that should be noted. First, although the CURVE app was developed for healthcare systems to insert local data using either csv files or through direct connection to their enterprise data warehouse, users still require R and R Shiny proficiency to modify or to add more tables, charts, or graphs. Second, in order to provide reliable, locally relevant forecasts, users must have access to local data and modeling expertise with frequent recalibration of the model to local context. If such changes are not monitored and accounted for, projections could be taken out of context, leading to erroneous conclusions.

Results from TrialIdentifier: No clinical trial numbers were referenced.

Results from Barzooka: We did not find any issues relating to the usage of bar graphs.

Results from JetFighter: We did not find any issues relating to colormaps.

Results from rtransparent:

Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
Thank you for including a funding statement. Authors are encouraged to include this statement when submitting to a journal.
No protocol registration statement was detected.

Read the original source