New Study Highlights Importance of Population Density in the Outcomes of Cancer Patients with COVID-19

An examination of data from a nationwide registry of patients with cancer and COVID-19 in the prevaccination era revealed no differences in outcomes by region but substantial heterogeneity in outcomes across cancer care delivery centers. Overall, adjusted mortality rates improved significantly across all regions over time. Regional variability in outcomes was absent despite vastly disparate COVID-19 treatments being used across different areas of the country.

Published in the JAMA Network Open, the study is part of the COVID-19 and Cancer Consortium [CCC19], which involves cancer centers all over the U.S. The CCC19 is the largest multicenter registry to examine the clinical characteristics, course of illness, and outcomes among patients with cancer and COVID-19.

A previous examination of the CCC19 registry revealed that patients with cancer and COVID-19 were at much higher risk of death or major adverse clinical outcomes, including intensive care unit admission (ICU) and intubation, than COVID-19 patients without cancer.

“We wanted to potentially identify best practices by evaluating whether there were significant temperospatial differences in outcomes or differences in outcomes by region of the country,” explains study co-author Nathan Pennell, MD, PhD, an oncologist at Cleveland Clinic’s Taussig Cancer Institute, one of the 83 centers included in the analysis.

The researchers hypothesized that major clinical outcomes would vary significantly across the U.S. given fluctuating case rates, resource disparities, and rationing of care. Overall, 32.9% of the study cohort resided in the Northeast, 34.5% in the Midwest, 18.8% in the South, and 13.8% in the West.

Key findings

The spatiotemporal pattern of COVID-19 cases across the U.S. was similarly present in patients with cancer. In the Northeast, more than three-fourths of cases (77.5%) were diagnosed from March to May, whereas from June to August, the South and the West reported their highest proportion of cases (48.8% and 51.3%, respectively). The West had the highest proportion of cases from September to November (18.5%).

Differences in the use of COVID-19 treatments were observed across the regions: Remdesivir was used more often in the West compared with other regions (13.3% vs. a range of 7.0% to 10.6%), whereas hydroxychloroquine was used to a greater extent in the Northeast (23.7%) compared with the other regions (range: 2.9% vs. 17.9%). These statistics may possibly reflect the earlier surge of COVID-19 in the Northeast – a time when the effectiveness of hydroxychloroquine was unclear, explains Dr. Pennell.

Although unadjusted 30-day mortality rates varied from 6.1% in the West to 19.6% in the Northeast, possibly reflecting a higher rate of ICU admission in the Northeast, no significant difference was found in adjusted 30-day mortality between regions. Across the nation, adjusted 30-day mortality rates improved over time, and this temporal trend may have been responsible for the lack of significant differences between regions, the study authors postulated. Dr. Pennell notes that investigators found substantial heterogeneity in outcomes depending on population density.

Importance of population density

“Patients treated at a center located in an urban area that was less densely populated (<250,000 people) had significantly better outcomes than those seen at centers in more densely populated city centers,” says Dr. Pennell. “It’s unclear exactly what led to that difference, as none of the measurable variables seemed to influence 30-day mortality. However, we do know that that an urban area with a population density under 250,000 is synonymous with the suburbs. The idea is that there may have been some unmeasured social determinants of health that we couldn’t quite pinpoint, but they may have been related to the better socioeconomic status of people who lived in those areas.”

Across the country, patient care was consistent throughout the first year of the pandemic, he emphasizes. “Patients were getting good care that was not measurably different based upon where they lived,” he explains. “The differences in 30-day death between centers, in what seemed to be a similar adjusted patient population, may be worth examining in more granular detail.”

In addition to the study being conducted prior to the availability of COVID-19 vaccines, the data reflect outcomes when the original variant was predominant and may not apply to subsequent variants, adds Dr. Pennell.