Maintenance Notice

Due to necessary scheduled maintenance, the JMIR Publications website will be unavailable from Wednesday, July 01, 2020 at 8:00 PM to 10:00 PM EST. We apologize in advance for any inconvenience this may cause you.

Who will be affected?

Accepted for/Published in: JMIR Public Health and Surveillance

Date Submitted: Feb 24, 2021
Date Accepted: Jun 29, 2021

The final, peer-reviewed published version of this preprint can be found here:

Forecasting COVID-19 Hospital Census: A Multivariate Time-Series Model Based on Local Infection Incidence

Nguyen HM, Turk P, McWilliams A

Forecasting COVID-19 Hospital Census: A Multivariate Time-Series Model Based on Local Infection Incidence

JMIR Public Health Surveill 2021;7(8):e28195

DOI: 10.2196/28195

PMID: 34346897

PMCID: 8341089

Forecasting COVID-19 Hospital Census: A Multivariate Time-series Model Based on Local Infection Incidence

  • Hieu Minh Nguyen; 
  • Philip Turk; 
  • Andrew McWilliams

ABSTRACT

Background:

COVID-19 has been one of the most serious global health crises in world history. During the pandemic, healthcare systems require accurate forecasts for key resources to guide preparation for patient surges. Forecasting the COVID-19 hospital census is among the most important planning decisions to ensure adequate staffing, number of beds, intensive care units, and vital equipment.

Objective:

The goal of this study is to explore the potential utility of local COVID-19 infection incidence data in developing a forecasting model for the COVID-19 hospital census.

Methods:

A multivariate time-series framework, called Vector Error Correction model, is used to simultaneously incorporate the COVID-19 hospital census and the local COVID-19 infection incidence and account for their possible long-run relationship. Hypothesis tests and model diagnostics are performed to test for the long-run relationship and examine model goodness-of-fit. 7-day-ahead forecast performance is measured by Mean Absolute Percentage Error, with time-series cross-validation. Based on different scenarios of the pandemic, the fitted model is leveraged to produce 60-day-ahead forecasts.

Results:

There is sufficient evidence that the two time-series have a stable long-run relationship, which serves as an important predictor. The model has very good fit to the data. The typical (median) out-of-sample Mean Absolute Percentage Error is 5.9%, which is lower than 6.6% from a univariate Autoregressive Integrated Moving Average model. Scenario-based 60-day-ahead forecasts exhibit concave trajectories with peaks lagging 2-3 weeks later than the peak infection incidence.

Conclusions:

Our findings show that the local COVID-19 infection incidence can be successfully incorporated into a multivariate time-series framework with the COVID-19 hospital census to improve upon existing forecast models, and to deliver accurate short-term forecasts and realistic scenario-based long-term trajectories to help healthcare systems leaders in their decision making.


 Citation

Please cite as:

Nguyen HM, Turk P, McWilliams A

Forecasting COVID-19 Hospital Census: A Multivariate Time-Series Model Based on Local Infection Incidence

JMIR Public Health Surveill 2021;7(8):e28195

DOI: 10.2196/28195

PMID: 34346897

PMCID: 8341089

Download PDF


Request queued. Please wait while the file is being generated. It may take some time.

© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.