Abstract
Background
Clinical trials (CTs) are the mechanism by which research is translated into standards of care. Low recruitment among underserved and minority populations may result in inequity in access to the latest technology and treatments, compromise the generalizability, and lead to failure in identification of important positive or negative treatment effects among under-represented populations.
Methods
Data were collected over a 39-month period on patient eligibility for available therapeutic cancer CTs. Reasons for ineligibility and refusal were collected. The data were captured using an automated software tool for tracking eligibility pre-enrollment. We examined characteristics associated with being evaluated for a trial, and reasons for ineligibility and refusal, overall and by patient race.
Results
African-Americans (AAs) were more likely than Whites to be ineligible (odds ratio, (OR) = 1.26, 95% confidence interval (CI) = 1.0–1.58) and if eligible, to refuse participation (OR = 1.79, 95% CI = 1.27–2.52), even after adjusting for insurance, age, gender, study phase, and cancer type. White patients were more likely to be ineligible due to study-specific or cancer characteristics. AAs were more likely to be ineligible due to mental status or perceived noncompliance. Whites were more likely to refuse due to extra burden, due to concerns with randomization and toxicity, or because they express a positive treatment preference. AAs were more likely to refuse because they were not interested in CTs, because of family pressures, or they felt overwhelmed (NS)).
Discussion
This study is the first to directly compare ineligibility and refusal rates and reasons captured prospectively in AA and White cancer patients. The data are consistent with earlier studies that indicated that AA patients more often are deemed ineligible and, when eligible, more often refuse participation. However, differences in reasons for ineligibility and refusal by race have implications for a cancer center to participate in CTs appropriate for the population of patients served. On a broader scale, consideration should be given to modifying eligibility criteria and other design aspects to permit broader participation of minority and other underserved groups.
Introduction
Clinical trials (CTs) are used to evaluate new therapies and are the mechanism through which research is translated into standards of care. The effectiveness of this translational process is greatly dependent on the number and representativeness of participants enrolled in trials, yet less than 5% of all adult cancer patients enter CTs. Despite a nearly 20-year effort by the National Institutes of Health (NIH) to enhance CT accrual, these rates are not improving, and even lower participation rates are reported in minority and underserved populations [1–7]. Low representation of these populations in CTs results in inequity in access to the latest technologies and cancer treatments, compromises the generalizability and external validity of the CT results [4–8], and may fail to identify important positive or negative treatment effects among under-represented populations [7–10].
Barriers to CT accrual that are related to patient and physician characteristics and the health care system have been identified [11–17]. Eligible patients may refuse to participate for many reasons, including concerns regarding experimentation and loss of control over treatment decisions [11–13,18] and failure to understand important trial procedures, such as randomization [19,20]. Physician barriers may represent a significant component of nonaccrual, due to difficulty initiating CT discussions and reconciling the dual roles of treating physician and clinical researcher [15,21–26].
Patient race has been shown to be associated with trial eligibility and refusal [27,28]. Despite the number of studies exploring barriers resulting in underrepresentation of minorities in CTs [28–34], there has been limited research that explores racial differences in reasons for ineligibility and for refusal to enroll in cancer CTs at the time of recruitment [28,33,34]. Those that did collect prospective information often lacked adequate representation of minorities in the study samples to permit comparisons of different racial groups [3,35]. Of the 36 studies identified in a systematic review of barriers to minority participation in trials, none reported the barriers according to racial groups [28].
To examine reasons for ineligibility and refusal among cancer patients in real time, we prospectively collected data on patients evaluated for CTs at an urban cancer center that treats significant numbers of both African-American (AA) and White patients. The aims of this study were to examine characteristics associated with being designated as ineligible for a trial and characteristics associated with refusing trial participation. The focus of this analysis is on comparison of reasons for ineligibility and refusal among AA and White cancer patients.
Methods
Study patient identification
Patient data for this study were captured and reported through the Clinical Trials Eligibility Database (CTED). CTED is an automated software system developed at Virginia Commonwealth University’s Massey Cancer Center that facilitates the clinical research evaluation process by linking longitudinal patient data from multiple electronic sources that include clinical reports (e.g., pathology and radiology reports, inpatient and outpatient clinical notes), clinical laboratory results, as well as billing records and scheduled visits. Clinical research staff (clinical research associates and research nurses) evaluate a patient’s eligibility for a specific CT through review of the linked data in CTED and/or by review of additional records in the Electronic Health Record (EHR) or from reports obtained from outside providers. For each evaluation, clinical research staff manually record a variety of parameters relevant to the patient’s status with respect to a specific CT or set of trials. These details include eligibility, specific reasons for ineligibility and refusal status and reasons for refusal, and referral source for the specific combination of patient and study. A detailed description of CTED functionality has been published [36].
During the development period for CTED, potential reasons for trial ineligibility and refusal were obtained by interviewing clinical research staff, by performing a comprehensive literature review, and then revising the initial ineligibility and refusal list to include additional reasons identified during the first 6 months of usage that had not been captured through the initial processes. Reasons were subjected further to qualitative thematic analysis techniques [37,38], and the resulting discrete categories were programmed as pull-down menus to be used by research staff to select pertinent reasons for ineligibility of and refusal by the individual patients evaluated. Reasons not included in the pull-down list could be specified by research staff. Thus, the CTED application provides a repository for clinical research staff to easily enter information for every patient evaluated for a CT. This system feeds data to the Clinical Trials Data Management System (OnCore©) that is used to track the patient once enrolled.
Study sample
The initial sample for this study was derived from all the patients with a scheduled visit to a National Cancer Institute (NCI)–designated cancer center in Richmond, Virginia, from June 2006 to March 2010 and who had an eligibility evaluation record completed in the CTED system. The analytic sample included patients with scheduled visits who also had a billing diagnosis for a specific cancer. The billing diagnosis for cancer in populations with a high prevalence of cancer patients has been demonstrated to have a high sensitivity, specificity, and positive predictive value [39,40].
Exclusion criteria for the analytic sample included age less than 21 years, incarceration, and race other than AA or White. Also excluded were patients who were evaluated only for nontherapeutic, companion, or ancillary studies.
Measures
Several data domains were captured and used in these analyses. Each is described below:
Demographic characteristics: Demographic information included gender, race (race determined either from the insurance record or entered by a registration clerk), age, and insurance status (commercial/health maintenance organization (HMO), Medicaid, Medicare, uninsured) available from patient scheduling or billing records and were used to represent characteristics that have been reported to be associated with CT enrollment.
Cancer diagnosis: Patients were classified as having a primary diagnosis of either a solid tumor cancer (e.g., breast, colorectal, lung) or a hematological malignancy (e.g., lymphoma, leukemia).
Trial characteristics: The trial phase (I, II, III, or IV) was obtained from the linked Clinical Trials Data Management system (OnCore) used to track patients post CTs enrollment.
-
Outcomes variables:
Eligibility status: Trial eligibility status (eligible or ineligible) was recorded in CTED by the clinical research staff who completed the assessment according to the eligibility criteria for an individual trial protocol. Those records with an eligibility ascertainment noted as pending (including records where the full evaluation had not been completed or eligibility status had not been determined at the time of data capture for this analysis) were coded separately as ‘final eligibility status not determined’.
Refusal status: Patients who were trial eligible were followed to ascertain whether they chose to participate in the offered CT and were classified as refusing or not refusing a trial.
CT ineligibility reasons. Reasons for patient ineligibility were coded for each patient. (A list of these reasons is provided in Table 1.) Reasons for ineligibility that were determined by the research nurse based on an interview with the patient or discussion with the attending physician, which included noncompliance and inappropriate mental status. Noncompliance included (1) consistent ‘no-shows’ for clinic appointments, (2) active substance abuse, and (3) ‘perceived instability’. Ineligibility due to inappropriate mental status was recorded whenever a patient was deemed incapable of understanding and providing informed consent.
CT refusal reasons. Reasons for refusal of participation were coded for each patient (a list of these reasons is provided in Table 2). The refusal category labeled ‘perceived extra burden’ included the following reasons: additional out-of-pocket expenses, lack of insurance coverage, extra clinic visits, and travel.
Table 1.
Number and percent of patients ineligible for clinical trials, overall and by race
| African-American, n = 341
|
White, n = 507
|
Total, N = 848
|
|
|---|---|---|---|
| n (%) | n (%) | N (%) | |
| Comorbidity | 68 (19.8) | 91 (18.0) | 159 (18.8) |
| Prior treatment* | 48 (14.0) | 99 (19.6) | 147 (17.3) |
| Extent of disease | 53 (15.4) | 76 (15.0) | 129 (15.2) |
| Histopathology nonmatch | 33 (9.6) | 45 (8.9) | 78 (9.2) |
| Second/multiple primary | 16 (4.7) | 27 (5.3) | 43 (5.1) |
| No available protocol | 10 (2.9) | 25 (4.9) | 35 (4.1) |
| Performance status | 15 (4.4) | 15 (3.0) | 30 (3.5) |
| Interval since diagnosis or treatmenta | 8 (2.3) | 20 (4.0) | 28 (3.3) |
| Stage of disease | 10 (2.9) | 18 (3.6) | 28 (3.3) |
| Does not meet trial-specific eligibility requirements | 9 (2.6) | 19 (3.8) | 28 (3.3) |
| Physician decisionb | 12 (3.5) | 15 (3.0) | 27 (3.2) |
| Anticipated noncompliance* | 21 (6.1) | 5 (0.99) | 26 (3.1) |
| Cancer characteristics | 7 (2.0) | 16 (3.2) | 23 (2.7) |
| Mental status/cognitive impairmentc** | 13 (3.8) | 3 (0.59) | 16 (1.9) |
| Age exclusion | 2 (0.58) | 5 (0.99) | 7 (0.8) |
| Enrolled in competing study | 1 (0.29) | 4 (0.79) | 5 (0.6) |
| Language barrier | 0 (0.00) | 1 (0.20) | 1 (0.1) |
| Missing reason for ineligibility | 17 (5.0) | 21 (4.2) | 38 (4.5) |
Includes evaluations for patients who had insufficient time or too much time elapsed to meet a particular eligibility criterion.
Physician determined that a patient was not a good candidate for a trial or physician preferred standard therapy for the patient.
Judged by research staff or attending physician as patient having an inability to comprehend informed consent.
Chi-squared test, 0.001 <p > 0.0001.
Chi-squared test, p <0.0001.
Table 2.
Number and percent of patients who refused clinical trial participation by reasons for refusal, overall and by race
| African-American, n = 130
|
White, n = 216
|
Total, N = 346
|
|
|---|---|---|---|
| n (%) | n (%) | N (%) | |
| Extra financial or logistical burden* | 13 (10.0) | 48 (22.2) | 61 (17.6) |
| Lack of interest in trial* | 33 (25.4) | 26 (12.0) | 59 (17.0) |
| Avoidance of specific treatment | 15 (11.5) | 24 (11.1) | 39 (11.3) |
| Preference for specific treatment** | 10 (7.7) | 28 (13.0) | 38 (11.0) |
| Fear of toxicity/side effects | 9 (6.9) | 21 (9.7) | 30 (8.7) |
| Discomfort with randomization*** | 6 (4.6) | 20 (9.3) | 26 (7.5) |
| Physician barrier | 6 (4.6) | 16 (7.4) | 22 (6.4) |
| Overwhelmed physically or emotionally | 8 (6.1) | 6 (2.8) | 14 (4.0) |
| Pressure from family/cultural factors*** | 9 (6.9) | 5 (2.3) | 14 (4.0) |
| Seen only for a second opinion | 0 (0.0) | 1 (0.46) | 1 (0.29) |
| No reason provided** | 21 (16.1) | 21 (9.7) | 42 (12.1) |
Chi-squared test, p <0.0001.
Chi-squared test, 0.001 > p <0.0001.
Chi-squared test 0.05 > p <0.001.
Statistical approach
Descriptive statistics were computed for gender, age, insurance status, and cancer type for the total sample, by race, by phase of trial, and by eligibility status. Frequencies for the reasons for trial ineligibility and for refusing the index CT (index CT = CT for which the patient was judged eligible) were also computed for the total sample and by race. Either a t-test or chi-square test was used to compare distributions among AAs and Whites. Probabilities of 0.05 or less were defined as statistically significant, without accounting for multiple comparisons of AAs and Whites.
A logistic regression model with eligibility status as the dependent variable was used to identify characteristics associated with ineligibility for a therapeutic cancer CT. A second logistic regression model with refusal status as the dependent variable was used for the subset of patients eligible for a CT to identify factors associated with refusal to participate in a therapeutic CT.
Covariates were entered simultaneously into each logistic regression model; these included race, age (grouped into quartiles), gender, insurance status (using commercial insurance as the reference category), type of cancer (using hematologic cancers as the reference group for diagnosis), and phase of trial (using either phase I or III, alternately, as the reference for other phases). The data from the patients for whom evaluation records had no study phase indicated (N = 207) were excluded from the regression analyses (Typically, no study phase was specified whenever a patient had been considered for a group of CTs but was determined to be ineligible for any of them before eligibility for a specific trial was considered.) Statistical significance of regression parameters was based on the probability that the odds ratio (OR) was not unity (p < 0.05) and the 95% confidence interval (CI) on the OR did not include unity.
Results
Patient sample
Table 3 presents by race and for the total sample the characteristics for 1955 patients who had at least one evaluation for a therapeutic CT during the study interval. The majority of these patients were White (64%) and female (68%); the mean age was 56 years. Approximately half (49.8%) had private insurance, while 28.6% had Medicare and 11.2% had Medicaid; the remaining patients were uninsured. More than half of the patients in the sample (56.5%) were assessed for a phase III trial. Rates of ineligibility by trial phase were as follows: phase I = 55.2%, phase II = 49.0%, phase III = 43.8%, and phase IV = 35.7%. A total of 38% were eligible for a therapeutic study; of these 742 patients, 46.6% refused to participate in a CT for which they were eligible.
Table 3.
Characteristics of patients evaluated for clinical trials by racial classification
| African-American, n = 713
|
White, n = 1242
|
Total, N = 1955
|
||
|---|---|---|---|---|
| n (%) | n (%) | N (%) | ||
| Gender | Women | 485 (68.0) | 844 (68.0) | 1329 (68.0) |
| Men | 228 (32.0) | 398 (32.0) | 626 (32.0) | |
| Age quartile | 21–48 years | 194 (27.2) | 280 (22.5) | 474 (24.2) |
| 49–57 years | 195 (27.3) | 319 (25.7) | 514 (26.3) | |
| 58–64 years | 166 (23.3) | 340 (27.4) | 506 (25.9) | |
| ≥65 years | 158 (22.2) | 303 (24.4) | 461 (23.6) | |
| Insurance | Private commercial | 230 (32.3) | 744 (59.9) | 974 (49.8) |
| Medicaid | 140 (19.6) | 79 (6.4) | 219 (11.2) | |
| Medicare | 229 (32.1) | 330 (26.6) | 559 (28.6) | |
| Uninsured | 114 (16.0) | 89 (7.2) | 203 (10.38) | |
| Cancer type | Solid tumor | 615 (86.2) | 983 (79.1) | 1598 (81.7) |
| Hematological | 82 (11.5) | 217 (17.5) | 299 (15.3) | |
| Missing | 16 (2.3) | 42 (3.4) | 58 (3.0) | |
| Trial phase | I | 58 (8.1) | 154 (12.4) | 212 (10.8) |
| II | 156 (21.9) | 262 (21.1) | 418 (21.4) | |
| III | 426 (59.7) | 678 (54.6) | 1104 (56.5) | |
| IV | 4 (0.6) | 10 (0.8) | 14 (0.7) | |
| None specifieda | 69 (9.7) | 138 (11.1) | 207 (10.6) | |
| Eligibility status | Eligible | 238 (33.4) | 504 (40.6) | 742 (37.9) |
| Ineligible | 341 (47.8) | 507 (40.8) | 848 (43.4) | |
| Not yet determined | 134 (18.8) | 231 (18.6) | 365 (18.7) | |
| Refusalsb | 130 (54.6) | 216 (42.8) | 346 (46.6) |
Patients considered for a generic set of studies (e.g., breast cancer) who were determined to be ineligible for any clinical trial.
All refusals are considered to be among eligible patients. Final eligibility determination may not have been completed for all refused patients as the workup stopped. Denominators for percentages are number eligible for that column.
Compared to AA patients, White patients were slightly older (57.6 vs. 55.9 years, p < 0.004) and more often had private, commercial insurance (59.9% vs. 32.3%, p < 0.0001). AA patients more often had been deemed ineligible compared to White patients (47.8% vs. 40.8%, p < 0.004). A higher proportion of eligible AA patients refused trial enrollment than did eligible White patients (54.6% vs. 42.8%, p <0.003).
Reasons for ineligibility and refusal
We identified 18 reasons for ineligibility for a therapeutic CT among the analytic sample (Table 1). The most common reason was comorbidity (18.8% of the total number of ineligible patients); the second most common reason was that the patient had received prior treatment that excluded them from the index trial (17.3%). Other important reasons for ineligibility included extent of the disease and histo-pathologic findings that did not satisfy requirements for the study.
Among the differences in reasons for ineligibility among AAs and Whites, AA patients more often than White patients had been designated as ineligible due to ‘mental status’, that is, perceived or documented inability to provide informed consent, and expected noncompliance based on a history of missed appointments, drug abuse, and so on. White patients were more frequently ineligible than AA cancer patients due to a history of prior treatment. The proportions of AA and White patients were similar with respect to comorbidity and cancer-related characteristics (Table 1). Few patients in this sample were excluded due to age.
Among the 11 categories of reasons for refusal to participate in the index trial, the most common category overall was the extra financial or logistical burden associated with CT participation (patient costs of participation, extra visits, travel time, etc.), accounting for 17.6% of the 346 refusals (Table 2). Other reasons for refusal were lack of interest in the trial, not wanting a specific treatment (primarily chemotherapy) provided as part of the trial, and preference for a specific treatment. No reason for refusal was recorded for 12.1% of the evaluations for which the patient refused participation.
As with reasons for ineligibility, some reasons given for refusing trial participation differed by patient race (Table 2). Whites refused about twice as often as AAs due to additional financial or logistical burden associated with CT participation (22.2% vs. 10.0%), because they had a preference for a specific treatment (13.0% vs. 7.7%), and due to discomfort with randomization (9.3% vs. 4.6%). Eligible AA patients about twice as often as Whites refused due to a lack of interest in the index CT (25.4% vs. 12.0%). AA patients also declined to provide a reason for refusal more often than White patients (16.1% vs. 9.7%). Eligible AA patients cited family pressures or cultural factors as reasons for refusal more frequently than White patients.
Logistic regression model for eligibility status
The first logistic regression model used ineligibility for a therapeutic cancer CT as the dependent (outcome) variable and sociodemographics, disease, and trial characteristics as the independent variables. Adjusted ORs for ineligibility among AAs compared to Whites and their 95% CIs are shown in Table 4. AAs were more likely to be ineligible for a trial but the lower bound of the CI was 1 (OR = 1.26, 95% CI = 1.0–1.58). Women were more likely to be ineligible than men (OR = 1.77, 95% CI = 1.41–2.25). Medicaid, Medicare, and uninsured patients were each more likely to be ineligible than commercially insured patients with ORs of 1.64 (95% CI = 1.14–2.35) for Medicaid, 2.35 (95% CI = 1.64–3.36) for Medicare, and 2.02 (95% CI = 1.38–2.94) for the uninsured patients. Patients were more likely to be ineligible when evaluated for a phase II or III trials (OR = 1.66, 95% CI = 1.14–2.43) compared with patients evaluated for a phase I study. Conversely, in a second model, in which phase I was replaced by phase III as the reference category, patients were less likely to be ineligible for phase III trials (OR = 0.76, 95% CI = 0.60–0.95), and there were no substantive changes to the other model parameters (data not presented). Neither age nor cancer type was associated with ineligibility after accounting for other covariates. In a third regression model, we removed age and cancer type from the independent variables; race, gender, and type of insurance remained associated with ineligibility (data not shown).
Table 4.
Multivariable logistic regression models predicting ineligibility and refusal
| Variable | Ineligibility
|
Refusal
|
|||
|---|---|---|---|---|---|
| OR | 95% CI | OR | 95% CI | ||
| Race | White | 1.00 | – | 1.00 | – |
| AA | 1.26 | 1.00–1.58* | 1.79 | 1.27–2.52*** | |
| Gender | Men | 1.00 | – | 1.00 | – |
| Women | 1.77 | 1.41–2.25 | 0.939 | 0.641–1.375 | |
| Age | 21–48 | 1.00 | – | 1.00 | – |
| 49–57 | 1.21 | 0.90–1.63 | 1.08 | 0.71–1.65 | |
| 58–65 | 1.38 | 1.02–1.87 | 1.19 | 0.78–1.83 | |
| ≥65 | 1.09 | 0.70–1.70 | 1.83 | 0.91–3.66 | |
| Insurance | Commercial | 1.00 | – | 1.00 | – |
| Medicaid | 1.64 | 1.14–2.35*** | 0.89 | 0.52–1.55 | |
| Medicare | 2.35 | 1.64–3.36 | 0.61 | 0.34–1.09 | |
| Uninsured | 2.02 | 1.38–2.94 | 0.81 | 0.44–1.47 | |
| Cancer type | Hematologic | 1.00 | 1.00 | ||
| Solid tumor | 1.19 | 0.85–1.69 | 1.92 | 1.16–3.18** | |
| Study phase | Phase I | 1.00 | 1.00 | ||
| Phases II and III | 1.66 | 1.14–2.43** | 1.18 | 0.65–2.13 | |
OR: odds ratio; CI: confidence interval; AA: African-American.
p <0.05,
p <0.01,
p <0.001.
Logistic regression model for trial refusal
In a separate analysis of factors associated with trial refusal among patients eligible for a therapeutic CT, the logistic regression model used refusal status as the dependent variable and sociodemographics, disease, and trial characteristics as independent variables (Table 4). AA patients were 1.8 times as likely as White patients to refuse participation in a therapeutic CT (OR = 1.79, 95% CI = 1.27–2.52). Cancer type was the only other covariate in this model that was significantly associated with refusal; patients with solid tumors were nearly twice as likely to refuse trial participation as patients with hematologic malignancies (OR = 1.92, 95% CI = 1.16–3.18). When the phase III replaced phase I as the reference phase in this regression model for comparison to all other phases, patients were nearly twice as likely to refuse phase III trials (OR = 1.91, 95% CI = 1.33–2.75) compared to trials of all other phases. In this model, cancer type was no longer significantly associated with refusal, but there were no other substantive changes to parameter estimates (data not shown). We repeated the analysis using a model without age and cancer type; the association of refusal with race remained.
Discussion
This study is the first to compare directly ineligibility and refusal reasons captured prospectively in AA and White cancer patients. The data are consistent with earlier studies that indicated that AA patients more often are deemed ineligible [15,33,41] and when eligible, more often refuse participation [27,28]. This prospective analysis adds to our current understanding of why there are racial differences in CT enrollment and provides data to suggest interventions to address barriers pertaining to the patient, the physician, and the health-care system [42].
Earlier studies that captured data prospectively either did not focus on cancer [14] or had insufficient numbers of AA patients to provide meaningful results for racial comparisons [3,35]. Studies that retrospectively capture reasons for refusal may suffer from recall bias [43]. Capture of reasons at the time of a patient’s evaluation and decision regarding participation in a CT reduces such biases [9,16,44–46].
In comparing differences in reasons for ineligibility among AA and White cancer patients, some reasons for ineligibility were based on subjective judgment on the part of the nurse or physician. These reasons were recorded more often for AA than White patients. While a relatively small proportion of all reasons, impaired mental status and expected noncompliance have been identified in other studies as reasons for ineligibility [15,33,41,47]. These reasons assigned by the health-care provider may reflect social or economic barriers for the patient, such as time or cost of travel, alternate child-care arrangements, or other factors that impact the ability or willingness to attend scheduled visits [47]. At our urban cancer center, AA patients are more often uninsured and have less social support than our White patients. These logistical issues may be viewed by research staff as impediments to successful trial participation [47]. Nevertheless, identification of these reasons for ineligibility offers an opportunity to expand access by designing trials that reimburse expenses related to participation, that require fewer clinic visits, or that provide for assessments at locations more convenient to participants.
Reasons for refusal to participate in a trial for which a patient had been judged to be eligible differed between AA and White patients. AA cancer patients more often expressed family or cultural barriers as reasons for nonparticipation. These barriers emphasize the need to elicit family participation in the decision-making process. A higher proportion of eligible AAs than Whites refused participation because they were overwhelmed with the decision-making process, suggesting the need for additional support infrastructure designed to serve minority and under-served populations (or any patients in need). Identifying and addressing these cultural differences have been shown to have a positive impact on enrollment of AA patients in at least one study [48].
Methods to address exclusionary criteria, such as limited insurance coverage of CT-related costs, may reduce disparities in enrollment and simultaneously increase generalizability of CT results as studies have shown that providing financial coverage of the CT-related costs eliminated insurance as a predictor of trial participation [49,50]. Nevertheless, disparate enrollment and limited generalizability among AA and patients with lower socioeconomic status likely will continue until other barriers, such as the additional direct costs that are unlikely to be covered by insurance (out-of-pocket expenses, extra visit co-payments), indirect costs (travel, child care), and lack of social support for CT participation, are eliminated [25,48,50].
The proportion of AA patients who cited fear of randomization or extra burden as reasons for refusal of trial participation was only half that for Whites similar to findings by Gadegbeku et al. [12]. He identified health-related factors and other burden-related factors rather than psychosocial factors as influencing CT participation by AAs [12]. In an analysis of refusers who cited ‘extra burden’ as the reason for refusal, we found that AA patients more often reported the higher costs and additional visits associated with CT participation as the primary reason for refusal, while White patients in this refusal category more often stated that they preferred another more convenient location for treatment.
Treatment preferences also were reported as a reason for refusal more often among White patients than AA patients. However, overall, nearly 23% of all the patients cited either a positive or negative preference for the treatment under consideration as a reason for refusal to participate in a CT. Patient refusal due to treatment preference may hinder recruitment to phase III trials in particular where standard-of-care alternatives are available outside the trial.
Limitations
While the data reported herein were collected prospectively from patients, they were captured by clinical research staff who entered the data into the system. Thus, reasons for refusal and ineligibility were filtered through the nurses’ perspectives and understanding of the patients. However, the findings are consistent with other prospective studies [3,14,15,23,25,35,47]. The data were drawn from the patient population of a single urban cancer center, which may limit the generalizability of results. Conversely, an important benefit of our concurrent data capture within a single system is that the information directly reflects the cancer patient population seen at that institution so that the findings are locally applicable.
We were not able to assess whether differences in comorbidity that may confound the differences in eligibility were associated with insurance status (i.e., Medicare population) because comorbidity was a reason for ineligibility and therefore could not be assessed as an independent variable.
Finally, our study sample does not contain races other than AAs and Whites, limiting our ability to make statements about other under-represented groups. Because race may not be self-reported, there may be misclassification of patients according to racial group. Furthermore, given the modest sample sizes for AA patients, care should be taken to assess in future research the possible role of variables not found to be significant in the present sample.
Conclusion
In conclusion, there are marked differences in barriers to eligibility and participation in CTs among AA and White patients. Systematic use of the type of information reported here may allow cancer centers to address barriers specific to their patient population, select CTs that most appropriately target their population, and avoid commitments to participate in trials for which accrual may be difficult or impossible [47]. Such data could impact trial development on a broader scale to direct changes in eligibility criteria that restrict the participation of minority patients and help improve the CTs infrastructure [4,5,51]. Ongoing use and additional testing of the CTED system or similar systems in other cancer centers have the potential to provide data that could directly impact CT recruitment.
Acknowledgments
Funding
This project was supported by the Massey Cancer Center. The development and testing of the software application used in this project was approved by the VCU IRB #HM11089.
Footnotes
Conflict of interest
None declared.
References
- 1.Avis NE, Smith KW, Link CI, Hortobagyi GN, Rivera N. Factors associated with participation in breast cancer clinical trials. J Clin Oncol. 2006;24(12):1860–67. doi: 10.1200/JCO.2005.03.8976. [DOI] [PubMed] [Google Scholar]
- 2.Ellis P, Butow PN, Tattersall MHN, Dunn SM, Houssami N. Randomised clinical trials in oncology: Understanding and attitudes predict patients willingness to participate. J Clin Oncol. 2001;19:3554–61. doi: 10.1200/JCO.2001.19.15.3554. [DOI] [PubMed] [Google Scholar]
- 3.Lara PN, Higon N, Lim N, et al. Prospective evaluation of clinical trial accrual patterns: Identifying potential barriers to enrolment. J Clin Oncol. 2001;19:1728–33. doi: 10.1200/JCO.2001.19.6.1728. [DOI] [PubMed] [Google Scholar]
- 4. [accessed 15 April 2010];National Cancer Clinical Trials System for the 21st Century. Available at: http://www.nap.edu/open-book.php?record_id=12879&page=1.
- 5.Young RC. Cancer clinical trials: A chronic but curable crisis. N Engl J Med. 2010;363(4):306–09. doi: 10.1056/NEJMp1005843. [DOI] [PubMed] [Google Scholar]
- 6.Back AL, Arnold RM, Baile WF, et al. Efficacy of communication skills training for giving bad news and discussing transitions to palliative care. Arch Intern Med. 2007;167:453–60. doi: 10.1001/archinte.167.5.453. [DOI] [PubMed] [Google Scholar]
- 7.Bartlett C, Doyal L, Ebrahim S, et al. The causes and effects of socio-demographic exclusions from clinical trials. Health Technol Assess. 2005;9(38):1–152. doi: 10.3310/hta9380. [DOI] [PubMed] [Google Scholar]
- 8.Britton A, McKee M, Black N, et al. Threats to the applicability of randomised trials, exclusions and selective participation. J Health Serv Res Policy. 1999;4(2):112–21. doi: 10.1177/135581969900400210. [DOI] [PubMed] [Google Scholar]
- 9.Branson RD, Davis KJ, Butler KL. African Americans participation in clinical research: Importance, barriers and solutions. Am J Surg. 2007;193(1):32–39. doi: 10.1016/j.amjsurg.2005.11.007. [DOI] [PubMed] [Google Scholar]
- 10.Swanson GM, Bailer JC. Selection and description of cancer clinical trials participants: Science or happenstance. Cancer. 2002;95(5):950–59. doi: 10.1002/cncr.10785. [DOI] [PubMed] [Google Scholar]
- 11.Jenkins V, Farewell D, Batt L, et al. The attitudes of 1066 patients with cancer towards participation in randomised clinical trials. Br J Cancer. 2010;103(12):1801–07. doi: 10.1038/sj.bjc.6606004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Gadegbeku CA, Stillman PK, Huffman MD, et al. Factors associated with enrollment of African Americans into a clinical trial: Results from the African American study of kidney disease and hypertension. Contemp Clin Trials. 2008;29(6):837–42. doi: 10.1016/j.cct.2008.06.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Payne JK, Hendrix CC. Clinical trial recruitment challenges with older adults with cancer. Appl Nurs Res. 2010;23(4):233–37. doi: 10.1016/j.apnr.2008.12.004. [DOI] [PubMed] [Google Scholar]
- 14.Corbie-Smith G, Viscoli CM, Kernan WN, et al. Influence of race, clinical and other sociodemographical, features on trial participation. J Clin Epidemiol. 2003;56(4):304–09. doi: 10.1016/s0895-4356(03)00046-5. [DOI] [PubMed] [Google Scholar]
- 15.Ulrich CM, James JL, Walker EM, et al. RTOG physician and research associate attitudes, beliefs and practices regarding clinical trials: Implications for improving patient recruitment. Contemp Clin Trials. 2010;31(3):221–28. doi: 10.1016/j.cct.2010.03.002. [DOI] [PubMed] [Google Scholar]
- 16.Advani AS, Atkeson B, Brown CL, et al. Barriers to the participation of African American patients with cancer in clinical trials: A pilot study. Cancer. 2003;97(6):1499–506. doi: 10.1002/cncr.11213. [DOI] [PubMed] [Google Scholar]
- 17.Siminoff LA, Zhang AY, Colabianchi N, Saunders-Sturm C, Shen QS. Factors that predict the referral of breast cancer patients onto clinical trials by their surgeons and medical oncologists. J Clin Oncol. 2000;18(6):1203–11. doi: 10.1200/JCO.2000.18.6.1203. [DOI] [PubMed] [Google Scholar]
- 18.Hancock BW, Aitken M, Radstone C, Vaughan Hudson G. Why don’t cancer patients get entered into clinical trials? Experience of the Sheffield Lymphoma Group’s collaboration in British National Lymphoma Investigation studies. BMJ. 1997;314:36–37. doi: 10.1136/bmj.314.7073.36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Ellis P, Dowsett SM, Butow PN, Tattersall MH. Attitudes to randomised clinical trials among outpatients attending a medical oncology clinic. Health Expectat. 1999;2:33–43. doi: 10.1046/j.1369-6513.1999.00028.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Cunny KA, Miller HW. Participation in clinical drug studies: Motivations and barriers. Clin Ther. 1994;16:273–82. [PubMed] [Google Scholar]
- 21.Fallowfield L. Can we improve the professional personal fulfilment of doctors in cancer medicine. Br J Cancer. 1995;71:1132–33. doi: 10.1038/bjc.1995.220. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Lee JY, Breaux SR. Accrual of radiotherapy patients to clinical trials. Cancer. 1983;52:1014–16. doi: 10.1002/1097-0142(19830915)52:6<1014::aid-cncr2820520614>3.0.co;2-j. [DOI] [PubMed] [Google Scholar]
- 23.Meropol NJ, Buzaglo JS, Millard J, et al. Barriers to clinical trial participation as perceived by oncologists and patients. J Natl Compr Canc Netw. 2007;5(8):655–64. doi: 10.6004/jnccn.2007.0067. [DOI] [PubMed] [Google Scholar]
- 24.Albrecht TL, Eggly SS, Gleeson MEJ, et al. Influence of clinical communication on patients’ decision making on participation in clinical trials. J Clin Oncol. 2008;26:2666–73. doi: 10.1200/JCO.2007.14.8114. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Joseph G, Dohan D. Recruiting minorities where they receive care: Institutional barriers to cancer clinical trials recruitment in a safety-net hospital. Contemp Clin Trials. 2009;30(6):552–59. doi: 10.1016/j.cct.2009.06.009. [DOI] [PubMed] [Google Scholar]
- 26.Martin JF, Henderson WG, Zacharski LG, et al. Accrual of patients into a multi hospital cancer clinical trial and its implications on planning future studies. Am J Clin Oncol. 1984;7:173–82. doi: 10.1097/00000421-198404000-00011. [DOI] [PubMed] [Google Scholar]
- 27.Barofsky I, Sugarbaker PH. Determinants of patient non participation in randomised clinical trials for treatment of sarcomas. Cancer Clin Trials. 1979;2:237–46. [Google Scholar]
- 28.Ford GF, Howerton MW, Lai GY, et al. Barriers to recruiting underrepresented populations in cancer clinical trials: A systematic review. Cancer. 2007;112(2):228–42. doi: 10.1002/cncr.23157. [DOI] [PubMed] [Google Scholar]
- 29.Gross CP, Filardo G, Mayne ST, Krumholz HM. The impact of sociodemographic status and race on trial participation in older women with breast cancer. Cancer. 2005;103(3):483–91. doi: 10.1002/cncr.20792. [DOI] [PubMed] [Google Scholar]
- 30.Mills EJ, Seely D, Rachlis B, et al. Barriers to participation in clinical trials of cancer: A meta-analysis and systematic review of patient-reported factors. Lancet Oncol. 2006;7(2):141–48. doi: 10.1016/S1470-2045(06)70576-9. [DOI] [PubMed] [Google Scholar]
- 31.McCaskill-Stevens W, Mckinney MM, Whitman CG, Minasian LM. Increasing minority participation in cancer clinical trials: The minority based community clinical oncology program experience. J Clin Oncol. 2005;23(22):5247–54. doi: 10.1200/JCO.2005.22.236. [DOI] [PubMed] [Google Scholar]
- 32.Paskett ED, Katz ML, DeGraffinreid CR, Tatum CM. Participation in cancer trials: Recruiting underserved populations. Clin Adv Hematol Oncol. 2003;1(10):607–13. [PubMed] [Google Scholar]
- 33.Adams-Campbell LL, Ahaghotu C, Gaskins M, et al. Enrollment of African Americans onto clinical treatment trials: Study design barriers. J Clin Oncol. 2004;22(4):730–34. doi: 10.1200/JCO.2004.03.160. [DOI] [PubMed] [Google Scholar]
- 34.Stewart JH, Bertoni AG, Staten JL, Levine EA, Gross CP. Participation in surgical oncology trials: Gender, race/ethnicity and age based disparities. Ann Surg Oncol. 2007;14(12):3328–34. doi: 10.1245/s10434-007-9500-y. [DOI] [PubMed] [Google Scholar]
- 35.Go RS, Frisby KA, Mathiason MA, et al. Clinical trial accrual among new cancer patients at a community based cancer center. Cancer. 2006;106(2):426–33. doi: 10.1002/cncr.21597. [DOI] [PubMed] [Google Scholar]
- 36.Penberthy L, Brown RF, Puma F, Dahman B. Automated matching software for clinical trials eligibility: Measuring efficiency and flexibility. Contemp Clin Trials. 2010;31(3):207–17. doi: 10.1016/j.cct.2010.03.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Bernard HR, Ryan GW. Text analysis: Qualitative and quantitative methods. In: Bernard HR, editor. Handbook of Methods in Cultural Anthropology. AltaMira Press; Thousand Oaks, CA: 1998. pp. 595–646. [Google Scholar]
- 38.Krueger R, Casey M. Focus Groups: A Practical Guide for Applied Research. SAGE; Thousand Oaks, CA: 2000. [Google Scholar]
- 39.Cooper GS, Yuan Z, Stange KC, et al. Agreement of Medicare claims and tumor registry data for assessment of cancer-related treatment. Med Care. 2000;38(4):411–21. doi: 10.1097/00005650-200004000-00008. [DOI] [PubMed] [Google Scholar]
- 40.Penberthy L, McLish DK, Manning C, Retchin S, Smith T. The added value of claims for cancer surveillance: Results of varying case definitions. Med Care. 2005;43(7):705–12. doi: 10.1097/01.mlr.0000167176.41645.c7. [DOI] [PubMed] [Google Scholar]
- 41.King AC, Cao D, Southard CC, Matthews A. Racial differences in eligibility and enrollment in a smoking cessation clinical trial. Health Psychol. 2011;30(1):40–48. doi: 10.1037/a0021649. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Brown RF, Bylund CL, Siminoff LA, Slovin SF. Seeking informed consent to Phase I cancer clinical trial: Identifying oncologists’ communication strategies. Psychooncology. 2010;20:361–68. doi: 10.1002/pon.1748. [DOI] [PubMed] [Google Scholar]
- 43.Smith YR, Johnson AM, Newman LA, et al. Perceptions of clinical research participation among African American women. J Womens Health (Larchmt) 2007;16(3):423–28. doi: 10.1089/jwh.2006.0124. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Comis RL, Miller JD, Aldige CR, Kreps L, Stoval E. Public attitudes towards participation in cancer clinical trials. J Clin Oncol. 2003;21(5):830–35. doi: 10.1200/JCO.2003.02.105. [DOI] [PubMed] [Google Scholar]
- 45.Sateren WB, Trimble EL, Abrams J, et al. How sociodemographics, presence of oncology specialists, and hospital cancer programs affect accrual to cancer treatment trials. J Clin Oncol. 2002;20(8):2109–17. doi: 10.1200/JCO.2002.08.056. [DOI] [PubMed] [Google Scholar]
- 46.Mosenifar Z. Population issues in clinical trials. Proc Am Thorac Soc. 2007;4(2):185–87. doi: 10.1513/pats.200701-009GC. [DOI] [PubMed] [Google Scholar]
- 47.Joseph G, Dohan D. Diversity of participants in clinical trials in an academic medical center: The role of the ‘Good Study Patient?’. Cancer. 2009;115(3):608–15. doi: 10.1002/cncr.24028. [DOI] [PubMed] [Google Scholar]
- 48.Germino BB, Mishel MH, Alexander GR, et al. Engaging African American breast cancer survivors in an intervention trial: Culture, responsiveness and community. J Cancer Surviv. 2011;5(1):82–91. doi: 10.1007/s11764-010-0150-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Martel CL, Li Y, Beckett L, et al. An evaluation of barriers to accrual in the era of legislation requiring insurance coverage of cancer clinical trial costs in California. Cancer J. 2004;10(5):294–300. doi: 10.1097/00130404-200409000-00006. [DOI] [PubMed] [Google Scholar]
- 50.Chun N, Park M. The impact of health insurance mandates on drug innovation: Evidence from the United States. Eur J Health Econ. doi: 10.1007/s10198-012-0379-6. Epub ahead of print 31 January 2012. [DOI] [PubMed] [Google Scholar]
- 51.Fouad M. Enrollment of minorities in clinical trials: Did we overcome the barriers? Contemp Clin Trials. 2009;30:103–04. doi: 10.1016/j.cct.2009.01.009. [DOI] [PubMed] [Google Scholar]