Fapesp

FAPESP and the Sustainable Development Goals


System forecasts hospital demand for personal protective equipment


System forecasts hospital demand for personal protective equipment

The open-access tool was developed in Brazil by researchers at the Center for Mathematical Sciences Applied to Industry to minimize the risk of shortages and overspending (photo: Rovena Rosa / Agência Brasil)

Published on 03/19/2021

By Elton Alisson  |  Agência FAPESP – In addition to having to plan the supply of intensive care beds and mechanical ventilators for the treatment of severe COVID-19 patients, hospital managers face the challenge of procuring personal protective equipment (PPE) for medical teams and other staff. Failure to predict the required amount of PPE with accuracy could lead to shortages or overspending.

To help them, researchers affiliated with the Center for Research in Mathematical Sciences Applied to Industry (CeMEAI), in partnership with Bionexo, a company that specializes in digital solutions for the management of health processes, have developed a system to project hospital demand for PPE during the COVID-19 pandemic in real time.

The open-access system is described in an article published on the platform medRxiv in a preprint version that has not yet been peer-reviewed.

CeMEAI is one of the Research, Innovation and Dissemination Centers (RIDCs) funded by FAPESP and is hosted by the University of São Paulo’s Institute of Mathematics and Computer Sciences (ICMC-USP) in São Carlos.

“The system helps hospitals maintain buffer stocks of PPE at a time when these resources have become scarce and expensive,” Francisco Louzada Neto, a professor at ICMC-USP and one of the CeMEAI researchers participating in the project, said during a webinar titled Focusing on the mathematics of COVID-19 in South America and held by FAPESP on June 4.

The system uses a mathematical model that predicts consumption of specific items of PPE over time based on a projected epidemiological curve for the area covered by the hospital and healthcare variables such as hospitalization rates, emergency room visit frequencies and numbers of beds available, all of which are combined to compute medical staffing requirements.

By integrating the data for these factors, the system forecasts demand for face coverings, gowns, coveralls, and other PPE, showing when consumption will peak, how long the peak will last, and any probable fluctuations.

“Based on these forecasts and projections, a hospital can estimate the amount of PPE to be purchased in order to minimize the risk of shortages or of oversupply potentially causing shortages elsewhere,” Louzada explained.

Hospital managers across Brazil have expressed interest in the system developed by CeMEAI, he added.

Actions to combat the pandemic

Development of the system is only one of several initiatives undertaken by CeMEAI in recent months to contribute to actions to combat the COVID-19 pandemic. When the outbreak began, the RIDC’s researchers mobilized to produce mathematical models that could be used to forecast its progress in Brazil, assess the effects of social isolation, and help cities plan quarantines, among others (read more at: agencia.fapesp.br/33297).

The COVID-19 section of CeMEAI’s website provides access to all the projects.

“Our research group includes mathematicians, statisticians and computer scientists. As a result, we have so far been able to develop nine projects relating to COVID-19,” José Alberto Cuminato, a professor at ICMC-USP and one of CeMEAI’s coordinators, told Agência FAPESP.

The article “Safety stock: predicting demand on the supply chain in Brazilian hospitals during the COVID-19 pandemic” (doi: 10.1101/2020.05.27.20114330) by Oilson Alberto Gonzatto Jr., Diego Carvalho do Nascimento, Cibele Maria Russo, Marcos Jardel Henriques, Caio Paziani Tomazella, Maristela Oliveira Santos, Denis Neves, Diego Assad, Rafaela Guerra, Evelyn Keise Bertazo, Jose Alberto Cuminato and Francisco Louzada can be read at: www.medrxiv.org/content/10.1101/2020.05.27.20114330v4.

 

Source: https://agencia.fapesp.br/33576