Patient Income Level and Cancer Clinical Trial Participation in a Prospective Survey Study

Cancer clinical trials provide the best evidence for showing the efficacy of new treatments. However, only a small percentage of adult cancer patients participate in clinical trials.¹ The issue of income disparities in clinical trial participation has been poorly addressed; limiting income disparities is important for ensuring rapid enrollment and fair access to trials. We previously found that patients with annual household income <$50,000 were 27% less likely to participate in clinical trials.² This provocative result was derived from one of many analyses of demographic and socioeconomic factors within a single, cross-sectional dataset, and so was considered hypothesis generating. The confirmation of this finding with prospectively collected data is critical for affirming its validity.

Methods

We utilized data from a prospective cooperative group survey study of barriers to participation in clinical trials conducted in eight geographically diverse cancer clinics.³ Adult patients with a new diagnosis (de novo or recurrent disease) of breast, lung, or colorectal cancer under consideration for systemic therapy were enrolled prior to making a treatment decision. Patients were then followed for a maximum of six months to assess whether they participated in a clinical trial. Patient-level baseline characteristics, including income, were collected.

We examined the association of income (<$50,000/year vs. ≥$50,000/year) and trial participation in a multivariable logistic regression model stratified by cancer type. We adjusted for the following factors that could potentially influence participation rates: age, sex, race (self-reported by participants), education, travel distance, and disease stage (initial diagnosis vs. recurrent disease). The analysis was conducted at the alpha=.05 level. We also examined whether there was evidence of an ordinal association of income level and trial participation by categorizing income as <$20,000/year vs. $20,000-$49,999/year vs. ≥$50,000/year.

Results

In total, n=1581 patients were eligible, and n=1262 (80%) with an income value were available for analysis. Patients were predominantly <65 years (71%), female (84%), and non-African American (93%; Table 1). In multivariable regression, patients with annual household income <$50,000 had a 32% lower odds of trial participation than higher income patients (12% vs. 17%; OR=0.68, 95% CI, 0.47-0.99, p=.04). Trial participation decreased as annual household income decreased from ≥$50,000 to $20,000-$49,999 to <$20,000 (17% vs. 13% vs. 11%; OR=0.75, 95% CI, 0.58-0.96, p=.02). Lower income patients were consistently less likely to participate in clinical trials across the key subgroups defined by model covariates (Figure 1).

Table 1. Patient Characteristics (N=1262)^*.

Characteristic	N	%
Age
<65	895	71%
≥65	367	29%
Sex
Female	1061	84%
Male	201	16%
Race
African American	84	7%
All Other	1178	93%
Annual income
<$20,000	276	22%
$20,000-$49,999	381	30%
≥$50,000	605	48%
Education
<2-year college degree	688	55%
≥2-year college degree	572	45%
Unknown	2
Disease stage
First diagnosis	1023	82%
Recurrent	225	18%
Missing	14
Distance patient traveled to receive care at clinic, miles
<13	357	28%
≥13	905	72%
Cancer type
Breast	894	71%
Lung	229	18%
Colorectal	139	11%
Participating clinics
University of Arkansas	160	13%
University of California (Davis)	200	16%
H Lee Moffitt Cancer Center	24	2%
Ozarks Regional CCOP	232	18%
Puget Sound	349	28%
Upstate Carolina	186	15%
Wayne State University	39	3%
Wichita CCOP	72	6%

CCOP = Community Clinical Oncology Program

^*Although n=1581 patients were eligible, an income value was not reported (239), unknown (68), or missing (12) for 319 patients, leaving n=1262 patients available for analysis.

Figure 1.

Forest plot of the association of income and clinical trial participation by each factor included in the multivariable regression model. Each square represents an odds ratio (OR), and each horizontal line is the 95% CI. The vertical line is the line of equal odds. For lower-income individuals, the odds of clinical trial participation were consistently lower (that is, to the left of the line of equal odds) within nearly all subgroups of all the factors included in this analysis. Only some of these findings are statistically significant, likely due to limited power. There was no statistical evidence that the association of income and clinical trial participation differed according to any of the covariates (interaction p>.15 in all cases).

Discussion

In a prospective study of barriers to trial participation, lower income patients were less likely to participate in clinical trials. Lower income patients are likely more sensitive to marginal financial expenditures than higher income patients.^4,5 Incentives or reimbursement may be appropriate, though they should not be coercive to patients. One approach to alleviate the financial risk associated with clinical trial participation would be to cover the excess costs of participation, including copayments and coinsurance. Direct compensation to clinical trial participants has also been recommended. ⁶ Considerations with respect to time off from work, child care, and transportation could also improve access to clinical trials for lower income patients. Future research should investigate how to overcome financial barriers to clinical trial participation.

The identification of patient income level as an independent predictor of trial participation is important for multiple reasons. If income is associated with health status, then improving representativeness of lower income patients on trials would improve the generalizability of study outcomes. Also, greater participation of lower income patients would allow trials to be conducted more quickly, speeding the development of new treatments. Crucially, since clinical trial treatments represent the newest available treatments, access to this vital resource should be available to individuals of all income levels.