Factors driving interhospital variations in mortality rates for critically ill COVID-19 patients

In critical care, the first step in improving patient outcomes is identifying what drives mortality. Often these drivers are a complex mix of individual patient factors and systems factors, both of which may vary across hospitals, regions and countries.

Since the beginning of the COVID-19 pandemic, emerging data has shown similar interhospital variability in outcomes of critically ill patients with the disease.

Matthew Churpek, MD, MPH, PhD, associate professor, Allergy, Pulmonary and Critical Care Medicine, and colleagues recently published research in the American Journal of Respiratory and Critical Care Medicine that aimed to determine the causes of that variability.

They found that patients’ presenting ICU physiology, the socioeconomic status of a hospital population and hospital strain are the driving factors behind interhospital variation in 28-day mortality rates for critically ill adults with COVID-19.

Analyzing patient- and hospital-level variables

Churpek and colleagues analyzed a nationally representative data set of 4,019 patients with lab-confirmed COVID-19 who were admitted to ICUs at 70 hospitals across the United States between March 4 and June 29, 2020.

Of those 4,019 patients, 1,537 (38 percent) had died within 28 days of admission. However, the mortality rate for those patients varied considerably: from 0 at the lowest-risk hospital to 82 percent at the highest-risk hospital.

To better understand the reasons for this, the team collected 51 patient-level and 29 hospital-level variables and categorized them into the following domains:

Patient-level domains:

  • Acute physiology and severity of illness in the first 48 hours of ICU admission
  • Demographics and comorbidities
  • Treatments provided in the first 48 hours of ICU admission

Hospital-level domains:

  • Socioeconomic factors at the hospital level
  • Hospital strain
  • Hospital quality scores

They then calculated each variable’s contributions to the differences in mortality rates across hospitals, and the degree of increased risk that a patient would have if he or she were admitted to a higher-risk hospital.

They also developed a patient-level model to determine the degree to which each domain contributed to an individual’s mortality risk.

After these adjustments, overall interhospital mortality variation was reduced (odds ratio decline from 2.06 to 1.22). Patients’ acute physiology, hospital socioeconomic status and hospital strain contributed most to the differences. Treatments provided to patients contributed the least.

The patient-level model found that individual mortality was explained mostly by presenting physiology, demographics and comorbidities, hospital socioeconomic status, and strain, with physiology explaining almost half of a patient’s mortality risk.

The role of community socioeconomic status

According to Churpek, this study is the first to investigate both hospital- and individual-level contributors to variation in mortality among critically ill patients with COVID-19.

“We also are the first to show that the socioeconomic status of the community surrounding a hospital is an important contributor to hospital-level variability in outcomes in a geographically representative sample of critically ill patients with COVID-19,” he said.

The team also found that the most important individual variable from the socioeconomic status domain was the percentage of patients at the hospital who traveled more than 45 minutes to work. This variable relates to discrepancies between the locations of low-income neighborhoods and employment opportunities and is an important metric of social risk.

“These findings suggest that COVID-19 may be exacerbating existing healthcare disparities in the United States,” Churpek said.

“For an individual patient critically ill with COVID-19, less than half of their mortality risk can be attributed to their physiology,” he continued. “This highlights the importance of other factors, such as hospital strain, co-morbidities, and socioeconomic status.”

Churpek says the next steps in this research include determining whether the importance of socioeconomic status relates to the individual’s status or to other unmeasured characteristics about a hospital.

“Ultimately, our hope is that by identifying the drivers of mortality variation, we can discover new ways to reduce it and improve outcomes for critically ill patients with COVID-19.”