Novel Index Identifies Social Vulnerability Barriers to Cancer Clinical Trial Enrollment

High levels of social vulnerability, including in terms of education, transportation, and neighborhood resources, are associated with lower odds of enrollment in cancer clinical trials, according to findings from a multivariable analysis presented during the 2024 American Society of Clinical Oncology Annual Meeting.1

“Unlike the reporting of [sex], race, and ethnicity for federally funded clinical trials, there is no mandate to collect and report social determinants of health,” Mariam F. Eskander, MD, MPH, the study’s senior author, wrote in an email to OncologyLive. “A major hurdle is simply collecting those data to understand [whether] a particular trial is enrolling a socioeconomically diverse population and what those barriers to enrollment might be. This is the first step and one of the major messages of this study––that we cannot fix what we do not measure. These social determinants data can be collected at the individual level or through a publicly available area-level index such as the Vizient Vulnerability Index [VVI].”

Eskander is also a surgical oncologist in the Gastrointestinal/Hepatobiliary Oncology Program at Rutgers Cancer Institute and RWJBarnabas Health, as well as an assistant professor in the Department of Surgery, Division of Surgical Oncology, Section of Gastrointestinal Oncology, at the Rutgers Robert Wood Johnson Medical School in New Brunswick, New Jersey.

The investigators used the Vizient Clinical Database to determine the VVIs of subgroups within the study population. VVI is a novel index that aims to characterize social drivers of health at the census tract level. The Vizient Clinical Database contains inpatient and outpatient clinical claims data encompassing 98% of academic medical centers, more than 100 cancer hospitals, and more than 700 community hospitals in the US. The database captures information including patient demographics, billing data, clinical outcomes, social needs, and obstacles to care that may influence a person’s overall health.

The VVI is made up of 9 domains: clean environment, economic, education, social environment, transportation, housing, neighborhood resources, health care access, and public safety. The domains include 43 components that elucidate factors such as lower median income, unemployment, no college degree, lower rates of voting participation, and no car or public transportation access, among others.

Study authors used International Classification of Diseases, Tenth Revision (ICD-10) codes to identify patients with breast, prostate, lung, colorectal, or pancreatic cancer; the first quartile of patients by VVI was deemed to be the least vulnerable, and the fourth quartile the most vulnerable. Investigators performed a cross-sectional analysis of diagnosis of cancer with a subsequent clinical trial code from 2022 to 2023 as well as a sensitivity analysis of clinical trial enrollment within 1 year of a new diagnosis of cancer in 2022. They also conducted a univariate analysis of clinical trial enrollment, multivariable analyses predicting clinical trial enrollment adjusted for demographic factors and comorbidities with separate models for overall VVI and the 9 VVI domains, and an interaction analysis of Black race and VVI.

The primary outcome of the study was participation in a clinical trial as measured by the ICD-10-CM (Clinical Modification) code Z-00.6 for “examination of participant in a clinical trial.”

Study authors identified 2,660,566 patients during the study period, 1.5% of whom were enrolled in a clinical trial; 28.6%, 26.3%, 22.8%, 12.8%, 9.5% of patients with breast, prostate, lung, colorectal, and pancreatic cancer, respectively, were enrolled in clinical trials. Young patients (< 55 years) were more likely to be enrolled in a clinical trial compared with older patients at respective rates of 19% vs 16% (P < .0001), as were men (50%) vs women (46%; P < .0001), White patients (79%) vs other patients (78%; P < .0001), those with private insurance (33%) vs other (32%; P < .0001), those with metastatic (58%) vs nonmetastatic (25%) disease (P < .0001), and those in the first VVI quartile (27%) vs the fourth VVI quartile (23%; P < .0001).

Additional findings from the analysis revealed that being older than 75 years (adjusted OR, 0.61; 95% CI, 0.58-0.65; P< .0001), female (adjusted OR, 0.97; 95% CI, 0.94-0.99; P=.0352), Black (adjusted OR, 0.78; 95% CI, 0.75-0.81; P<.0001); having Medicaid (adjusted OR, 0.76; 95% CI, 0.73-0.80; P<.0001); or being diagnosed with cancer at a community hospital (adjusted OR, 0.30; 95% CI, 0.29-.031; P<.0001) was associated with decreased odds of clinical trial enrollment. Being classified in VVI quartile 2 (adjusted OR, 0.85; 95% CI, 0.83-0.88; P=.0002), VVI quartile 3 (adjusted OR, 0.85; 95% CI, 0.81-0.88; P<.0001), or VVI quartile 4 (adjusted OR, 0.86; 95% CI, 0.81-0.91; P=.0442) was also associated with decreased odds of clinical trial enrollment. In terms of the VVI domains, high vulnerability in terms of education (adjusted OR, 0.82; 95% CI, 0.78-0.86; P<.0001), neighborhood resources (adjusted OR, 0.87; 95% CI, 0.83-0.91; P<.0001), or transportation (adjusted OR, 0.89; 95% CI, 0.85-0.94; P<.0001) was associated with decreased odds of clinical trial enrollment.

Being classified in the fourth VVI quartile decreased the odds of clinical trial enrollment for Black (OR, 0.80; 95% CI, 0.68-0.88; P<.0001) and White patients (OR, 0.88; 95% CI, 0.89-0.96; P=.0028). However, findings from an interaction analysis revealed that high social vulnerability decreased the enrollment odds more for Black patients than White patients (P for interaction=.0054).

“We found that living in a socially vulnerable census tract lowers the likelihood that a patient with lung, breast, prostate, colorectal, or pancreas cancer will enroll in a clinical trial,” Eskander said. “This applies to the entire cohort but impacts Black patients more than it does White patients, meaning a Black patient living in a socially vulnerable neighborhood is less likely to enroll in a clinical trial than a White patient from a socially vulnerable neighborhood. This is the most contemporary and nationally representative analysis of the association between social vulnerability and clinical trial enrollment and the first to explore the intersection of social vulnerability and race.”

Data from the sensitivity analysis showed that 1.0% of 1,820,247 total patients enrolled in a clinical trial within 1 year of diagnosis in 2022. Living in the most vulnerable census tract was associated with a 19% decrease in the odds of enrolling in a clinical trial, and high social vulnerability decreased the odds of enrollment more for Black patients than for White patients (P for interaction<.0001).

Eskander and her coauthors noted that their study was limited by challenges in quantifying patients who are eligible for clinical trials, heterogeneity of cancer types and disease biology, unmeasured confounding factors, and inaccurate/incomplete coding. Eskander noted that the next steps for this research are collecting social determinants of health data for all patients, taking action to bring clinical trial sites closer to socially vulnerable neighborhoods, investing in clinical trial education for patients and providers, and advocating for government funding to address social and financial barriers to clinical trial enrollment. The study authors will place an emphasis on educational, transportation, and neighborhood domain barriers.

“We are working to understand the role of social vulnerability in different populations and settings as well as tease out the underlying barriers that most impact clinical trial enrollment,” Eskander said. “We encourage institutions to look at their own data to assess barriers for socially vulnerable patients and design interventions at the system and individual level to meet community need.”

Reference

Perati SR, Mohayya SM, Shippey E, et al. Social vulnerability and clinical trial enrollment: the next frontier of health equity. J Clin Oncol. 2024;42(suppl 16):1508. doi:10.1200/JCO.2024.42.16_suppl.1508