In COVID-19 Era, Modeling Shows Promise in Preventing Health Care–Associated Infections

An uptick in modeling and genomic data caused by the COVID-19 pandemic could lead to important infection-prevention measures in health care settings.

The COVID-19 pandemic has led to new attempts to better understand how pathogens spread in health care settings, but several hurdles remain that limit incorporation of the latest modeling technology into everyday clinical practice.

In a new article, investigators review what the latest models suggest regarding health care–setting transmission of the SARS-CoV-2 virus, as well as other pathogens, like methicillin-resistant Staphylococcus aureus and Clostridioides difficile.

Writing in the journal Current Opinion in Infectious Diseases, the authors noted that close to 35,000 Americans are believed to die each year as a result of antimicrobial-resistant infections. Many of those infections originate in a health care setting, with an estimated 4% of hospitalized patients developing health care–associated infections (HAIs), they wrote.

However, stopping the threat of HAIs is a complex problem, given the wide range of people, equipment, and pharmaceuticals interacting in a health care setting. One way to “disentangle” those factors, the investigators said, is using mathematical, statistical, and computational modeling. Modeling can also be used to find and assess mitigation measures, they said.

One strategy for reducing HAIs is targeted vaccinations, including a potential vaccine against C difficile infection (CDI).

The authors noted that findings from one study, which investigated a potential vaccine against CDI, “suggested that between 3 and 16 CDI cases per 1000 vaccinated patients could be averted, including 1 to 5 cases among unvaccinated individuals.”

Another opportunity would be decolonizing organisms that can facilitate infection, the investigators said. This strategy holds promise, but they said it requires additional assessment, including on the basis of cost-effectiveness.

The authors also said one of the most prominent HAI-mitigation efforts, antimicrobial stewardship, remains important and continues to perform well in models and the real world.

“For example, by curtailing nonspecific use of antibiotics, cases of CDI have been lowered by over 30% in some hospitals,” they wrote.

Looking forward, they authors said genomic data and analysis are opening up new avenues to understand how transmission takes place in hospitals. Such data can be used to develop models of transmission and clusters within a specific hospital setting. In one study, genomic and epidemiological data from hospitalized patients was used to locate clusters of SARS-CoV-2 infection.

“The results helped clinical, infection control, and hospital management teams to improve interventions and patient safety,” they wrote.

Additionally, machine learning is emerging as a potent tool, one that can potentially help clinicians understand patients at high risk of contracting an infection, they said.

The authors closed their paper by looking at some of the key gaps and challenges remaining in HAI prevention. Among them are racial and socioeconomic gaps in outcomes, a problem that has come into stark relief amid the COVID-19 pandemic.

“Predictive models for HAI should continue to explore the specific impact of demographic factors such as sex and race,” the authors concluded. “Although a number of studies include demographic variables in the model building process, causal dependencies can be challenging to illicit.”

Other challenges include the limitation that risk modeling is based on conditions at a specific location at a specific point in time, the authors noted, meaning models may need to be adjusted or retrained as conditions change.

Even when models are high performing, a gap remains between the validation of the models and implementation of their insights. After all, if those insights never make it into wide use, the benefits of modeling will be significantly blunted. Yet the authors said the COVID-19 experience shows the inverse is also true.

“In particular, the rapid, high-stakes collaborative effort that brought modelers, hospital administrators, and public health officials together to manage the COVID-19 pandemic has shown that greater emphasis on the deployment process is needed to efficiently and effectively address future outbreaks,” they said.

Reference

Stachel A, Keegan LT, Blumberg S; CDC MInD Healthcare Program. Modeling transmission of pathogens in healthcare settings. Curr Opin Infect Dis. 2021;34(4):333-338. doi:10.1097/QCO.0000000000000742