Why location matters: associations between county-level characteristics and availability of National Cancer Oncology Research Program and National Cancer Institute sites

Abstract

Background

The majority of patients with cancer seek care at community oncology sites; however, most clinical trials are available at National Cancer Institute (NCI)–designated sites. Although the NCI National Cancer Oncology Research Program (NCORP) was designed to address this problem, little is known about the county-level characteristics of NCORP site locations.

Methods

This cross-sectional analysis determined the association between availability of NCORP or NCI sites and county-level characteristic theme percentile scores from the Center for Disease Control and Prevention’s Social Vulnerability Index themes. Health Resources and Services Administration’s Area Health Resource Files were used to determine contiguous counties. We estimated risk ratios and 95% confidence intervals (CIs) using modified Poisson regression models to evaluate the association between county-level characteristics and site availability within singular and singular and contiguous counties.

Results

Of 3141 included counties, 14% had an NCORP, 2% had an NCI, and 1% had both sites. Among singular counties, for a standard deviation increase in the racial and ethnic theme score, there was a 22% higher likelihood of NCORP site availability (95% CI = 1.10 to 1.36); for a standard deviation increase in the socioeconomic status theme score, there was a 24% lower likelihood of NCORP site availability (95% CI = 0.67 to 0.87). Associations were of smaller magnitude when including contiguous counties. NCI sites were located in more vulnerable counties.

Conclusions

NCORP sites were more often in racially diverse counties and less often in socioeconomically vulnerable counties. Research is needed to understand how clinical trial representation will increase if NCORP sites strategically increase their locations in more vulnerable counties.