COVID-19 hospitalization requirements in low-resource settings: mathematical modeling can help!
Using mathematical modelling, we combine theoretical assumptions with real-world data to the study of the ongoing epidemic in a low-resource setting, characterized by marked inequalities in health access.
Following the discovery, in December 2019, that a novel coronavirus (now identified as SARS-CoV-2) was circulating and causing clusters of respiratory disease, scientists around the world have joined efforts to understand the multiple facets of this pathogen: How does it spread?1 What treatments can be used to enhance a patient's survival?2 Can we arrive at a safe and successful vaccine, in record-time, to halt the pandemic?3
Fortunately we, as a global scientific community, are on track to answer these questions and many others regarding SARS-CoV-2 and its associated COVID-19 disease outcomes. Particularly, our group, based in Brazil, has gathered expertise of local researchers across multiple disciplines to study and tackle the epidemic that has claimed more than 200,000 lives in this country so far. Brazil's gigantic dimensions and stark social inequalities makes the fight against COVID-19 an even more daunting challenge. For starters, although an Universal Health System has been in place since 1988, access to health services is still limited, particularly when specialized care is needed. Patients with severe presentations of COVID-19 usually require this type of care, in addition to mechanical ventilation and other equipment available only in intensive care units (ICUs). Motivated by this, in our work4we aimed at answering the following central question: Can a mathematical model be used to accurately predict the need for hospitalization during the unfolding of this epidemic?4
Such a model could be a valuable tool for decision-makers, holding the potential to help inform strategic choices on the need for opening new clinical and ICU beds to support health care needs as the number of COVID-19 cases escalate. Our model is named SEIIHURD, an acronym for the 8 stages ("compartments", in mathematical lingo) that individuals can go through: Susceptible to the virus, Exposed, Infectious (either symptomatic or asymptomatic), Hospitalized in clinical ward, Hospitalized in ICU, Recovered, or Dead due to severe disease, effectively plugging into the model's dynamics the range of clinical outcomes related to COVID-19, hospitalization requirements, and mortality. In order to measure the forces for meeting the targets to protect the local health care system, the model embodies a time varying rate of transmission that accounts for human behaviour influenced by the interventions and local health policies implemented. This means that we recognize that the spread of the disease is deeply influenced by our collective attitudes and behaviors (ie. a composite result of individual mobility reduction; use of face masks; social distancing; among others), which in turn is dictated by the set of governmental actions in place, such as stay-at-home orders–or lack thereof5,6. Increased population adherence to social distancing and more strict measures may reduce the transmission rate quicker, while the opposite is expected to occur when no restrictions are enacted.
Taken together, we have shown that a simple mathematical model can largely anticipate the needs for COVID-19-related hospitalizations and could be used as a tool to inform the actions of people involved in decision processes. This is particularly important in low-resource settings, as the already very limited health resources must be efficiently organized. Our model is one of many that have been produced and applied to the study of this unprecedented crisis (eg.7-9), and we hope that the utility and applications of such methods to save lives will be regarded as one of the positive legacies of this pandemic.
1 Anfinrud et al. Visualizing Speech-Generated Oral Fluid Droplets with Laser Light Scattering. N Engl J Med 2020; 382:2061-2063.
2 WHO Solidarity Trial Consortium. Repurposed Antiviral Drugs for Covid-19 — Interim WHO Solidarity Trial Results. DOI: 10.1056/NEJMoa2023184. 2020.
3 Poland et al. SARS-CoV-2 immunity: review and applications to phase 3 vaccine candidates. Lancet. 2020; 396(10262):1595-1606.
4 Oliveira, J.F. et al. Mathematical modeling of COVID-19 in 14.8 million individuals in Bahia, Brazil. Nat Commun 12, 333 (2021).
5 Jorge, D.C.P., Rodrigues, M.S., Silva, M.S., et al. Assessing the nationwide impact of COVID-19 mitigation policies on the transmission rate of SARS-CoV-2 in Brazil. bioRxiv (2020). Pre-print.
6 Haug, N., Geyrhofer, L., Londei, A. et al. Ranking the effectiveness of worldwide COVID-19 government interventions. Nat Hum Behav 2020; 4:1303-1312.
7 Li, R., Pei, S., Chen B. et al. Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV-2). Science 2020; 368(6490):489-493.
8 Nature Human Behavior "COVID-19 and human behavior collection" available at https://www.nature.com/collections/gdjdibibfg
9 Britton et al. A mathematical model reveals the influence of population heterogeneity on herd immunity to SARS-CoV-2. Science 20202; 369(6505):846-849.