George Mason University professor Elise Miller-Hooks and her team have been studying and modeling the flow of patients through American hospitals in times of crisis since 2014.
Now, in the face of the COVID-19 pandemic, they are collaborating with Mersedeh Tariverdi who is a Senior Data Scientist in the Health, Nutrition, and Population Group at the World Bank to launch a web portal that will support models that aid hospital responses to the pandemic.
The team's work supports evidence-based decision making, informed by the models, to rethink and facilitate hospital operations when utilizing limited critical resources as demand surges. While working as a Ph.D. student under Miller-Hooks' supervision, Tariverdi developed key models that now serve as the backbone for the MASH-Pandemics portal.
Only hours after President Donald Trump declared a national emergency on March 13, graduate students working with Miller-Hooks began incorporating personal protective equipment, ventilators, and COVID-19 patients with their care paths into the models.
Disposable equipment such as face masks and permanent equipment like ventilators are treated differently. The models allow the team to assess alternative capacity enhancement and demand management strategies, crisis standards of care options, and hospital collaboration measures in a dynamically changing environment.
We're working as fast as we can because the need is so urgent and the crisis is enormous. This effort is about health care worker's safety and helping the hospitals cope with the pandemic. It's about saving lives, as well as the efficient and equitable allocation of resources."
Elise Miller-Hooks, Professor, George Mason University
The work can help hospitals to best cope with surge demand in spite of limited resources. It can aid decision-makers in regional response and hospital collaboration planning with health care facilities offering various levels of care, e.g. primary health care.
It can inform decisions on the mobilization of critical supplies, supplemental space, including tents and ships, and teams of frontline health care workers and other first responders to where they are needed the most.
The models provide general recommendations that are made based on findings that assume a generic urban hospital design and crisis standards of care approaches to treating patients in this pandemic. Hospital administrators and others working in an official capacity can request runs of the MASH-Pandemics models through the portal to help them with decision-making tailored for their hospitals or group of hospitals working in collaboration.
This effort builds on previously developed sophisticated discrete-event simulation models to replicate operations in the ongoing COVID-19 pandemic. MASH-Pandemics was developed on a patient-based, resource-constrained, multi-unit hospital model originally designed for assessing hospital preparedness to serve emergency patients in a surge, pandemic, mass casualty incident, and disaster events.
"Our models are unique for several reasons," says Miller-Hooks. "If you look at most other hospital models, and there are many out there, they generally focus on single key units and are designed for routine operations," she says. "Sometimes they'll include two units. Our models track patients along their entire care paths, along with physical resources and staff, through ten essential units. The models are very detailed as is needed to capture the complexities of their operations. We do not know of other models that have considered mass casualty incidents at this level.
"Our insights to aid recommendations rely on feedback from collaborating disaster and emergency medicine experts from Johns Hopkins University including Lauren Sauer and Scott Levin, as well as Thomas Kirsch," says Miller-Hooks. "World Bank Group's engagement brings a global perspective to the collaboration, and this perspective is critical in aiding hospitals at the frontlines of this world-wide emergency in resource-scarce settings."
COVID-19 pandemic led to significant decline in cardiac arrest survival rates