A recent study by a team of engineers from the Whiting School of Engineering’s Department of Civil and Systems Engineering (CaSE) offers new insights on the changing patterns of COVID-19 mortality across the U.S. from August 2020 to March 2023. Published in PLOS Global Public Health in September, the study, “Evolving patterns of COVID-19 mortality in US counties: A longitudinal study of healthcare, socioeconomic, and vaccination associations”, examines how various factors such as health care access, socioeconomic status, and vaccination rates influenced mortality at different stages of the pandemic.
COVID-19 studies have been plentiful since the pandemic began, but most focused on its early stages. Setting this study apart is its comprehensive approach tracking the evolving impact of the virus over three years through changing variants, public health measures, and access to treatments and vaccines; its capture of local variations at the county level; and its incorporation of the Social Vulnerability Index and other factors that make communities more susceptible to negative effects during public health emergencies.
“COVID-19 was a moving target,” said CaSE doctoral student, Fardin Ganjkhanloo. “The virus changed, our policies changed, and our behavior changed. We wanted to know: Did the risk factors for mortality change, as well?”
The team found that while some risk factors consistently showed a significant association with mortality outcomes, others shifted over time.
Socioeconomic status, access to health care, education level, and income were persistent factors, meaning they were consistent predictors of mortality rates through all stages of the pandemic. Disparities in these factors also consistently correlated with higher mortality rates, highlighting fundamental inequities that remained influential despite changes during the pandemic.
Similarly, counties with fewer hospital beds experienced worse outcomes, especially as hospital capacity became strained in later waves, underscoring the pivotal role of hospital capacity management.
Other factors, however, were dynamic. Race, for example, showed different correlations at different stages of the pandemic. Early on, minority communities were disproportionately affected, but this association weakened as vaccination campaigns and targeted public health interventions took effect. By late 2020, the correlation between racial minority status and mortality had largely reversed.
“Race was a big surprise for us,” Ganjkhanloo said. “What started as a significant risk factor early in the pandemic shifted over time, likely due to a combination of health care interventions and behavioral changes.”
The study’s use of county-level data, the most detailed available for nationwide analysis, provided insights into how local factors influenced mortality. While national and state-level studies can overlook local disparities, this study captured changing trends in the relationship between urban and rural areas, showing varied effects at different stages.
While urban areas generally faced higher mortality rates throughout most of the pandemic, even when multiple socioeconomic and healthcare factors were considered, this trend was not consistent, particularly during the early stages.
As expected, vaccination proved effective in reducing mortality, as counties with higher vaccination rates saw fewer deaths. However, the researchers found that the vaccine’s protective effect was more pronounced in some counties, reflecting disparities in health care infrastructure and public health messaging.
“Understanding both persistent and dynamic factors is necessary for developing targeted and effective public health strategies, especially for vulnerable populations,” said Kimia Ghobadi, the John C. Malone Assistant Professor of Civil and Systems Engineering and the study’s principal investigator. “Persistent factors highlight areas requiring long-term, systemic interventions, while dynamic factors demonstrate the need for adaptive, flexible response strategies. These insights are imperative for improving future preparedness.”
Looking forward, the team’s methodology could be valuable for studying other infectious diseases, chronic conditions, or even the health impacts of environmental changes over time.
“With a comprehensive, longitudinal, and multi-level methodology, there is potential to advance the way we analyze and understand the dynamic nature of public health crises,” said Ganjkhanloo.
Additional study collaborators include doctoral student Farzin Ahmadi, recent CaSE graduate Ensheng Dong, Engr ’24 (PhD), and Lauren Gardner, the Alton and Sandra Cleveland Professor in the Department of Civil and Systems Engineering.