Adults with a higher cardiometabolic disease staging score who also experience social determinants of health have increased risks for severe COVID-19 outcomes, according to study findings published in Obesity.
In a retrospective analysis of patient electronic medical record data, researchers found prediction models that factored in cardiometabolic health history from the 3 years before COVID-19 infection as well as social determinants of health could most accurately predict severe COVID-19 outcomes and identify those most at risk.
“We were pleasantly surprised that including social determinants of health improved the prediction model and confirmed our hypothesis, given the limited measures we had due to restrictions of electronic medical record data,” Carrie R. Howell, PhD, assistant professor in the department of medicine, division of preventive medicine at the University of Alabama at Birmingham, told Healio. “Our work here, as well as other current work and work among other researchers, is beginning to justify the need to look at these measures and to pair them with clinical measures to better capture the patient’s unique experience and improve risk stratification.”
Howell and colleagues conducted a retrospective study using data on cardiometabolic markers and social determinants of health to predict COVID-19 outcomes. Adults at the University of Alabama at Birmingham health system aged 35 years and older who had a positive COVID-19 test and complete clinical data available from 2017 to 2020 were included. Cardiometabolic disease staging score was calculated for each patient using BMI, blood pressure and levels of glucose, HDL cholesterol and triglycerides. Insurance status was collected from EMRs. Neighborhood level data were merged by census tract to describe characteristics of each patient’s residence. The social vulnerability index uses census tract data to describe social conditions that influence human suffering and financial hardship, the researchers wrote, with a higher score indicating more social vulnerabilities. Census tract data were linked to data from the U.S. Health Resources and Services Administration (HRSA) Data Warehouse – Primary Care Service Area Data to determine access to health care providers. COVID-19 outcomes were defined as hospitalization, ICU admission or death during hospitalization.
There were 2,873 patients included in the analysis (mean age, 58 years; 59% women; 45% non-Hispanic Black). Of the study cohort, 33% were hospitalized with COVID-19, 13.6% were admitted to the ICU and 4.2% died. About 35% of the cohort lived in an area with high social vulnerability, and 31.5% lived in an area with a health care provider shortage.
Every 1 standard deviation unit increase in cardiometabolic disease staging score was associated with an increased risk for COVID-19 hospitalization (OR = 2; 95% CI, 1.83-2.2), ICU admission (OR = 1.88; 95% CI, 1.67-2.11) and death (OR = 1.69; 95% CI, 1.4-2.04). Compared with those with private insurance, patients with no insurance had an increased risk for hospitalization (OR = 3.35; 95% CI, 1.88-5.98), ICU admission (OR = 2.99; 95% CI, 1.53-5.56) and death (OR = 7.27; 95% CI, 2.65-19.94). Similar findings were observed for adults with public insurance. Residents in census tracts with high social vulnerability had a higher likelihood for hospitalization (OR = 1.57; 95% CI, 1.21-2.03) and ICU admission (OR = 1.66; 95% CI, 1.21-2.28) than those living in census tracts with low social vulnerability.
A model using cardiometabolic disease staging score parameters to predict COVID-19 hospitalization had an area under the curve of 0.776. When individual and neighborhood social determinants of health were added, the AUC increased to 0.819. Similar findings were observed when ICU admission or mortality were the primary outcome.
“When we included social determinants of health, our prediction model significantly improved,” Howell said. “Clinically, we should be including these measures and concepts in our risk prediction models that are being used to guide care.”
Howell said future research should focus on validating models using social determinants of health to assess cardiometabolic outcomes. She added that more common acknowledgement of social determinants of health in clinical care models is also needed.