Abstract
Background
Recruitment and participation in clinical trials by minorities, particularly African Americans and rural underserved populations, are low. This report examines predictors of clinical trial recruitment and participation for adult Marylanders.
Methods
A cross-sectional design was used to survey 5,154 adults (18 years and older) residing in 13 of the 24 jurisdictions in Maryland, including urban Baltimore City, and the rural regions of Western Maryland and the Eastern Shore. The survey, conducted between December 2001 and March 2003, used Computer-Assisted Telephone Interviewing and random-digit dialing procedures. Primary dependent variables included “ever asked to participate” (i.e., recruited) and “participated” in clinical trials.
Results
11.1% of the respondents had been recruited to clinical trials. In addition, 59.4% of the respondents recruited to clinical trials actually participated in a clinical trial. Among respondents recruited to clinical trials, black and middle income respondents were significantly less likely to actually participate in clinical trials; whereas, respondents who received information about clinical trials from their health care provider, who were knowledgeable about clinical trials, and those who had the time commitment were significantly more likely to participate in clinical trials.
Conclusions
These results suggest serious gaps in efforts to recruit racial/ethnic minorities and residents of rural regions into clinical trials. The findings provide the basis for the development and implementation of community-based educational programs for both the general public and health care professionals, and to enhance availability of community-based clinical trials, especially in the rural areas of the state.
Author's Key Words (or terms): clinical trial, cancer disparities, patient selection, research participant recruitment, patient participation, rural or urban population, minority groups, neoplasm, knowledge, attitudes, source of information
Introduction
Clinical trials are a critical resource for the discovery of new prevention, diagnostic and treatment modalities for disease. For cancer, clinical trials have produced advances in treatment as well as prevention. Despite these advances in cancer prevention and patient care, only about 3-5% of cancer patients participate in clinical trials.1, 2
Assuring diversity in clinical trial participation is a national priority. In 1993, the most recent amendment to the National Institutes of Health (NIH) Revitalization Act (Public Law 103-43),3 mandated the inclusion of women and minorities in clinical research and government sponsored human subject research including clinical trials. This Act states that women and minorities must be included in all clinical research studies and must be included in Phase III clinical trials. Trials must also be designed to permit valid subgroup analyses. The Act states that cost is not an allowable reason for excluding minorities and that the NIH will support outreach efforts to fulfill this mandate. However, for communities that experience high cancer burden as demonstrated by elevated cancer incidence and mortality rates, such as African Americans, the uninsured and poor, and rural patients, participation in cancer clinical trials is particularly low.2, 4, 5 Moreover, the proportion of trial participants who are African Americans has declined in recent years.5
Participation in clinical trials is affected by individual (patient and provider) and structural factors. Some evidence suggests that slightly less than one-third (32%) of Americans would be willing to participate in clinical trials if asked, and, an additional 38% would be inclined to participate if asked but had some questions or reservations.1 Therefore, factors other than patient intent or willingness seem to impede participation in clinical trials. Some of the salient factors impeding participation in clinical trials include: being a racial minority;2, 5-7 older age;2, 5, 7-9 lower socioeconomic status;2, 6, 7, 10, 11 lack of appropriate clinical trials and the disqualification of patients;1 the reluctance of physicians to engage in accrual;1, 12 doctor-patient communications regarding clinical trials;7, 10, 13 mistrust of academic institutions, research and the medical system;6, 10, 12, 14-17 fear of negative effects;11 lack of community and physician awareness and knowledge of clinical trials benefits;10, 11, 18 lack of sufficient infrastructure (including oncologists and approved cancer programs) to support trials in community settings;2 lack of researcher training in culturally appropriate patient concerns and communication methods;13, 17, 19 certain historical factors;14 lack of adequate support for community outreach; poor access to care;12 and lack of information on available trials.10, 12
The low participation in cancer trials by African Americans and other minorities may contribute to existing cancer survival and mortality rate disparities. These avoidable disparities in cancer research participation are a public health problem in that access to cutting edge advances in cancer prevention and therapy are not equitably available to those populations experiencing substantially higher cancer incidence, morbidity and mortality rates.20 While numerous studies discuss issues related to the low participation of African Americans in trials, very few discuss possible predictive factors associated with trial participation. The focus of this report is two-fold: first, to present the prevalence of clinical trial recruitment and participation in three regions of Maryland; and second, to determine multivariate socio-demographic, rural/urban geography, source of clinical trial information, enabling factors, and health care access predictors of recruitment and participation. The results from this analysis will be used to document geographic and population-specific barriers and channels for clinical trial education and promotion. Moreover, the report's findings would serve as the basis for the development of educational programs to increase participation in trials and also facilitate the identification of key areas for education and outreach programs to promote greater knowledge and awareness of key facts on trials. Ultimately, it is hoped that greater awareness and intensive educational programs will increase the availability and likelihood of participation in clinical trials by urban, rural and minority underserved patients.
Methods
The data are derived from a larger investigation on the health behavior, health care access, and screening and health status conducted in 13 of the 24 jurisdictions in the state of Maryland between December, 2001 and March, 2003. 21 The 13 jurisdictions studied included urban Baltimore City, the three counties of rural Western Maryland (Garrett, Allegany, and Washington counties), and the nine counties of the rural Eastern shore (Cecil, Kent, Queen Anne's, Talbot, Caroline, Dorchester, Wicomico, Somerset, and Worcester counties). In this report, we focus on questions included in the clinical trials module, one of the eight modules included in this survey that examined the prevalence and predictors of clinical trial participation for residents in the state of Maryland. The survey was conducted by the Center for Health Policy/Health Services Research at the University of Maryland School of Medicine. The Human Subject Institutional Review Board of the University of Maryland Baltimore approved the research protocol and all study participants provided verbal informed consent.
Study Design and Procedures
A cross-sectional design was used to survey 5,154 English-speaking, non-institutionalized men and women aged 18 years or older. The survey sample was selected using random digit dialing methodology along with selection of eligible respondents within households, and it employed Computer-Assisted Telephone Interviewing (CATI) data collection procedures.21 To ensure an adequate sample of men surveyed, the study's a priori sample size estimates and response percentages were set for each county at 55% female and 45% male. Based on these estimates, within each eligible household, interviewers solicited the “most available” male member of the household over the age of 18 for the interview. If there was no male in the household or the “most available” male was not willing to or would never be available to respond to the survey, the “most available” female member in the household was requested for the interview. If the most available male was present for the interview at the time of the telephone call, subsequent call attempt efforts were made to contact the eligible male respondent. More than 95% of the Marylanders have telephones (Genesys Sampling Systems, April 2005, personal communications), and random-digit dialing is unimpeded by the non-listing of telephone numbers. Thus, the study results can be considered generalizable only to English-speaking, non-institutionalized adults reachable by telephone within the survey timeframe through the level of effort described below.
Trained interviewers conducted the interviews during optimal calling times, as described by Aday.22 Interviewers were routinely monitored by staff for quality control, and all interviews were recorded using a digital voice logger. Respondents were called without prior notification and were not paid for their participation. A minimum of 8 attempts were made before a final disposition code was assigned to a telephone number.
A bank of telephone numbers was obtained from Genesys Sampling Systems. Overall, 33,130 numbers were called and assigned a final disposition code: complete, active or passive refusal, unable to contact (non-contact), inaccessible household, or ineligible household or number. Active refusal is defined as eligible respondents who were forthright in refusing to be interviewed and those who started an interview but interrupted it and were not willing to continue. Passive refusal is defined as households where answering machines were repeatedly encountered. Non-contact is defined as households where the eligible respondent interrupted the survey and was willing to continue but did not schedule and complete the survey during the survey timeframe, and households where the potential respondent repeatedly scheduled call attempts but never conducted the interview during the survey timeframe. Inaccessible households were defined as those where there were repeated no answer and where the telephone line was constantly busy. Ineligible households or numbers were defined as business or non-private residence, telephone numbers with inconsistencies (i.e., numbers not in service or disconnected lines), fax or computer lines, wrong numbers, household where the telephone number had changed, households outside the state or without adults (18 years and older), and households with non-English-speaking adults.
Out of the 33,130 telephone numbers used, 9,297 were ineligible. Of the remaining 23,833 eligible numbers, 202 telephone numbers (0.85%) could not be contacted, 11,663 telephone numbers (48.94%) actively refused participation, 1,987 telephone numbers (8.34%) were categorized as passive refusals, and 4,827 telephone numbers (20.25%) were inaccessible. The remaining 5,154 eligible numbers resulted in a completed survey by the designated respondent. The completion rate, defined as the number of completed interviews divided by the sum of completed interviews and active refusals,23 for the survey was 30.6%. The completion rates for urban and rural jurisdictions were 39.0% for Baltimore City, 31.0% for rural Western Maryland, and 29.3% for rural Eastern Shore.
Measures
The overall survey instrument included 212 questions adapted from the Centers for Disease Control and Prevention's (CDC) Behavioral Risk Factor Surveillance Survey (BRFSS), the Commonwealth Fund Comparative Minority Health Survey, and original survey questions developed from prior research conducted through the Center for Health Policy/Health Services Research at the University of Maryland School of Medicine.21 The survey instrument was conceptualized on eight general dimensions or modules: health status; health care coverage and satisfaction; lifestyle factors, such as nutrition, exercise, weight control, alcohol and tobacco use (including smokeless tobacco); gender-specific questions on cancers of the breast, cervix, and prostate, as well as colorectal cancer, including utilization of screening and early detection examinations for these cancers; preventive behaviors for cancers of the skin and mouth; other health issues such as hypertension, cholesterol, and cardiovascular disease; clinical trials knowledge, attitudes, sources of information, and potential barriers to participation; and socio-demographics including race/ethnicity, gender, income, education, occupation, and military status. The average length of the survey varied by region: 19.1 minutes for interviews in Baltimore City; 19.6 minutes for the Eastern Shore region; and 23.3 minutes in the Western Maryland region. In this report, we focus on socio-demographics, health status and health insurance coverage, and the module on clinical trials.
The focus on the clinical trial module was to determine attitudes, awareness and knowledge of clinical trial aspects, previous participation, barriers, most likely sources of information on clinical trials, and predictive factors associated with willingness to participate for residents in 13 of Maryland's underserved geographic areas: urban Baltimore City, rural three county Western Maryland, and rural nine county Eastern Shore region. Independent variables included several socio-demographic measures such as age, sex, race/ethnicity, region (county) of residence, education level, and income level; health status; and health insurance coverage. Race/ethnicity was categorized as white or Caucasian, black or African American, or other. Age and health status variables were dichotomized as less than 65 years and 65 years or older for age, and poor or non-poor (including fair, good, very good, or excellent) health status. Education categories reflected the highest grade/level of school completed. Primary health insurance coverage was reported as either through work or union, through someone else's work or union, purchased directly, Medicare, Medicaid, other group, Veterans' Administration (VA), or no insurance. For the multivariate analysis, health insurance coverage was categorized as: no insurance, public insurance, or private insurance. Annual income was categorized as low income (less than $15,000), middle income ($15,000-$50,000), and high income (above $50,000). Finally, respondents were classified as living in Baltimore City, Western Maryland, or the Eastern Shore according to their county of residence. Other independent variables included factors that may influence participation in clinical trials (or “enabling factors”). Respondents were presented with nine separate factors and asked whether or not each would make them more likely to participate in clinical trials. These nine factors included reimbursement, insurance coverage, transportation, childcare, increased knowledge, time commitment, anonymity, medical follow-up, and additional health care. Lastly, respondents were asked whether or not they had received information concerning clinical trials from any of seven possible sources: printed literature, their doctor, the internet, the church, over the radio, from a community group, or via the television.
Dependent variables included being “ever asked to participate” (or recruited) in clinical trials and “participated” in clinical trials. All respondents were asked the question, “Have you ever been asked to participate in a clinical trial?” Response to this question provided the primary information covering efforts to recruit different demographic groups for trials. Respondents who indicated that they had been asked to participate in a clinical trial were then asked, “Did you participate?”
Analysis
Bivariate and multivariate analyses were conducted to determine relationships between the independent variables and the binary outcome variables. Chi-square tests were used to assess differences between each of the two outcome variables (recruited and participated) and independent variables such as socio-demographics (i.e., gender, race/ethnicity, age, income, education level), health status, access to care (i.e., health insurance status), geography of residence, the seven sources of information about clinical trials, and the nine enabling factors. Independent multivariate stepwise, logistic regression models were constructed for the two outcome variables to further explore the effects of the independent variables (i.e., socio-demographics, geography of residence, health status, and access to care, seven sources of information, and nine enabling factors) on being recruited to clinical trials and participating in clinical trials. The independent variables sources of information and enabling factors were only included in the logistic regression model for the outcome variable “participation” in clinical trials. The sample for the logistic regression models included respondents who self-identified themselves as either black or white for race, and those who responded either “yes” or “no” on the individual predictor and outcome variables. Respondents who self-identified themselves as being of “other” race and those who answered “do not know/unsure” or “refused” for the predictor and outcome variables in the model were excluded from the multivariate stepwise logistic regression analyses. For each categorical variable in the model, the referent category had an odds ratio (OR) of 1.0. The logistic regression results appear as odds ratios and 95% confidence interval (CI).24 Hosmer-Lemeshow test statistics was used to assess the goodness-of-fit.
Results
Sample Characteristics
The sample consisted of 5,154 adults surveyed in urban Baltimore City (n=681), rural Western Maryland (n=1122), and rural Eastern Shore (n=3351). Overall, the majority of adults surveyed were less than 65 years of age (74.7%), white (79.2%), female (53.1%), considered themselves in “not poor” health (94.6%), high school graduate (34.7%) or with some college (25.4%) level of education, with health insurance coverage (90.6%), and with an annual income of less than $50,000 (59.7%). There were significant differences in socio-demographic characteristics of the adults surveyed in the three regions. Urban Baltimore City respondents were more likely to be less than 65 years of age (80.5% vs. 74.0% in Eastern Shore and 73.3% in Western Maryland, p<.001), African American (56.3% vs. 16.3% in Eastern Shore and 1.9% in Western Maryland, p<.001) males or females, and have health insurance coverage provided through their work or union (49.1% vs. 40.9% in Eastern Shore and 38.6% in Western Maryland, p<.001). Furthermore, the sample in urban Baltimore City was less likely to have graduated from high school or have some college education (50.7% vs. 59.7 in Eastern Shore and 67.0% in Western Maryland, p<.001) and have health insurance coverage provided through Medicare (16.5% vs. 24.0% in Eastern Shore and 26.3% in Western Maryland, p<.001).
Prevalence of Clinical Trial Recruitment and Participation
Overall, out of the 5,154 respondents, 574 respondents (11.1%) surveyed in the 13 jurisdictions in Maryland were recruited in clinical trials. Moreover, among the 574 respondents who were recruited, 341 respondents (59.4%) actually participated in clinical trials (data not shown). Table 2 presents the prevalence of recruitment to and participation by socio-demographic characteristics, health status, and insurance coverage. Overall, respondents more likely (p<.001) to be recruited to clinical trials were 65 years or older (14.4%), had poor health status (17.7%), had some college or higher level of education (63.4%), had either private (purchased directly or through work or union, 32.0%) or public health insurance coverage (VA, Medicaid, or Medicare, 51.7%), and were residents of urban Baltimore City (19.7%) followed by rural Western Maryland (13.6%). Among the respondents who were recruited, white females (64.8%), white males (61.1%), male (100.0%) respondents from an other race, and those residing in rural Western Maryland (68.9%) followed by rural Eastern Shore (60.6%) vs. those residing in urban Baltimore City (47.0%) were significantly (p<.001) more likely to actually participate in clinical trials.
Table 2. Recruitment to and Participation in Clinical Trials.
| Recruited N=574 |
Participated N=341 |
|||
|---|---|---|---|---|
| n | % | n | % | |
| Race/Gender | *** | |||
| White male | 204 | 10.5 | 124 | 61.1 |
| White female | 262 | 12.6 | 169 | 64.8 |
| Black male | 33 | 8.9 | 14 | 42.4 |
| Black female | 54 | 9.7 | 21 | 38.9 |
| Other male | 6 | 8.5 | 6 | 100.0 |
| Other female | 8 | 16.7 | 4 | 50.0 |
| Age Group | *** | |||
| Under 65 | 386 | 10.1 | 232 | 60.3 |
| 65 and over | 185 | 14.4 | 107 | 58.2 |
| Health Status | *** | |||
| Poor | 49 | 17.7 | 30 | 61.2 |
| Not poor | 524 | 10.8 | 311 | 59.6 |
| Education | *** | |||
| Never attended | 0 | 0.0 | ||
| Elementary | 14 | 9.0 | 6 | 42.9 |
| Some high school | 33 | 7.0 | 18 | 54.5 |
| High school grad | 133 | 7.6 | 87 | 65.9 |
| Some college | 165 | 12.8 | 93 | 56.4 |
| College grad | 113 | 15.2 | 70 | 62.5 |
| Some graduate school | 26 | 18.1 | 13 | 50.0 |
| Graduate degree | 87 | 17.3 | 53 | 60.9 |
| Insurance Status | *** | |||
| No insurance | 28 | 5.8 | 18 | 64.3 |
| Through work or union | 216 | 10.2 | 131 | 60.6 |
| Someone else's work or union | 65 | 9.1 | 37 | 56.9 |
| Purchased directly | 41 | 12.7 | 24 | 58.5 |
| Medicare | 176 | 14.8 | 98 | 56.0 |
| Medicaid | 20 | 16.3 | 13 | 65.0 |
| Other group | 6 | 8.2 | 5 | 83.3 |
| Veterans' Administration | 14 | 20.6 | 10 | 71.4 |
| Income Level | ||||
| Less than $7,500 | 26 | 11.9 | 16 | 61.5 |
| $7,500-$14,999 | 31 | 8.8 | 15 | 50.0 |
| $15,000-$24,999 | 54 | 11.2 | 32 | 59.3 |
| $25,000-$34,999 | 68 | 10.9 | 35 | 51.5 |
| $35,000-$49,999 | 59 | 9.0 | 31 | 52.5 |
| $50,000-$74,999 | 98 | 11.9 | 58 | 59.2 |
| $75,000-$99,999 | 50 | 13.5 | 37 | 74.0 |
| $100,000 or more | 49 | 12.7 | 35 | 71.4 |
| Geographical Area | *** | *** | ||
| Urban Baltimore City | 133 | 19.7 | 62 | 47.0 |
| Rural Eastern Shore | 289 | 8.7 | 175 | 60.6 |
| Rural Western Maryland | 152 | 13.6 | 104 | 68.9 |
p<.05,
p<.01,
p<.001 by the χ2 test.
Sources of Clinical Trial Information
Table 3 contains data on the seven different sources of information about clinical trials. The most common source of information was television, followed by print media, and radio. The church was the least common source of information about clinical trials. With some exceptions, the television followed by the print media, were the most common sources of information for all race/gender groups, age groups, education level, income groups, and urban or rural region of residence. African American males were more likely to receive information about clinical trials from the television followed by the radio, and African American females more likely to receive information about clinical trials from the television followed by the print media. In addition, increased levels of education and income determined the primary source for clinical trial information; for example, respondents who were college graduates or who had higher levels of education, as well as those reporting incomes of $100,000 or more, were proportionally more likely to receive information about clinical trials from the print media, followed by the television.
Table 3. Sources of Clinical Trial Information.
| Church (N=123) |
Community (N=393) |
Doctor (N=663) |
Print (N=2155) |
Internet (N=707) |
Radio (N=1843) |
TV (N=2650) |
|
|---|---|---|---|---|---|---|---|
| % | % | % | % | % | % | % | |
| Race/Gender | *** | *** | *** | * | *** | ** | |
| White Male | 1.7 | 6.6 | 13.8 | 41.5 | 14.1 | 40.4 | 52.5 |
| White Female | 1.5 | 8.0 | 14.5 | 48.1 | 14.6 | 34.8 | 54.6 |
| Black Male | 5.1 | 10.2 | 8.9 | 24.9 | 11.5 | 31.0 | 44.2 |
| Black Female | 4.3 | 8.5 | 8.4 | 36.0 | 10.4 | 31.3 | 49.5 |
| Other Male | 2.8 | 7.0 | 5.6 | 39.4 | 19.7 | 35.2 | 49.3 |
| Other Female | 6.3 | 10.4 | 12.5 | 52.1 | 18.8 | 45.8 | 58.3 |
| Age Group | ** | *** | *** | *** | *** | ||
| Under 65 | 2.1 | 7.8 | 12.9 | 43.6 | 16.2 | 40.0 | 55.3 |
| 65 and over | 3.4 | 7.5 | 13.1 | 38.2 | 6.5 | 25.2 | 43.0 |
| Health Status | *** | * | *** | ||||
| Poor | 3.2 | 9.5 | 19.9 | 35.3 | 10.2 | 20.9 | 49.6 |
| Not poor | 2.4 | 7.6 | 12.6 | 42.7 | 14.0 | 37.1 | 52.4 |
| Education | *** | *** | *** | *** | *** | *** | |
| Never attended | 0.0 | 0.0 | 9.1 | 36.4 | 27.3 | 36.4 | 36.4 |
| Elementary | 2.6 | 5.2 | 6.4 | 15.0 | 2.5 | 13.0 | 26.5 |
| Some high school | 3.4 | 5.8 | 6.0 | 18.6 | 7.0 | 17.5 | 39.4 |
| High school graduate | 2.3 | 5.2 | 10.2 | 31.4 | 8.0 | 29.8 | 47.5 |
| Some college | 2.0 | 8.2 | 14.0 | 47.7 | 15.8 | 41.0 | 57.9 |
| College graduate | 2.3 | 9.8 | 17.2 | 59.8 | 21.7 | 46.9 | 57.8 |
| Some graduate school | 2.1 | 11.1 | 19.6 | 60.1 | 21.1 | 53.8 | 58.9 |
| Graduate degree | 1.8 | 13.8 | 21.5 | 69.3 | 25.5 | 51.9 | 65.2 |
| Insurance Status | * | ** | *** | *** | *** | *** | |
| No insurance | 2.5 | 5.4 | 7.1 | 30.1 | 11.9 | 31.9 | 47.9 |
| through work or union | 1.6 | 8.3 | 12.8 | 47.2 | 16.8 | 41.8 | 56.7 |
| Someone else's work or union | 2.4 | 7.7 | 14.7 | 47.2 | 17.1 | 41.5 | 56.1 |
| Purchased directly | 2.2 | 8.1 | 14.5 | 42.5 | 16.1 | 41.4 | 53.3 |
| Medicare | 3.5 | 7.5 | 14.0 | 37.5 | 7.1 | 25.2 | 43.6 |
| Medicaid | 4.0 | 8.9 | 18.0 | 29.2 | 12.2 | 28.5 | 54.5 |
| Other group | 5.5 | 6.9 | 8.3 | 44.4 | 13.7 | 30.6 | 53.5 |
| Veterans' Administration | 1.5 | 4.5 | 13.4 | 32.8 | 13.2 | 29.4 | 45.6 |
| Income Level | * | *** | *** | *** | *** | *** | |
| Less than $7,500 | 4.1 | 6.9 | 11.1 | 29.8 | 10.0 | 24.1 | 44.3 |
| $7,500-$14,999 | 3.4 | 8.2 | 9.0 | 29.7 | 8.1 | 23.6 | 45.1 |
| $15,000-$24,999 | 3.9 | 5.6 | 11.3 | 34.2 | 9.3 | 31.2 | 48.5 |
| $25,000-$34,999 | 1.6 | 6.6 | 11.1 | 35.3 | 13.1 | 31.9 | 50.7 |
| $35,000-$49,999 | 2.1 | 7.0 | 12.3 | 45.2 | 13.6 | 39.8 | 54.2 |
| $50,000-$74,999 | 1.9 | 9.0 | 12.4 | 51.6 | 18.0 | 42.7 | 57.3 |
| $75,000-$99,999 | 1.3 | 9.0 | 16.8 | 55.6 | 21.6 | 54.2 | 63.0 |
| $100,000 or more | 0.8 | 9.6 | 22.3 | 61.5 | 22.0 | 52.4 | 60.2 |
| Geographic Area | ** | *** | *** | ** | |||
| Urban Baltimore City | 3.4 | 9.9 | 12.4 | 49.6 | 15.4 | 42.3 | 52.2 |
| Rural Eastern Shore | 2.2 | 6.8 | 12.6 | 42.2 | 13.5 | 38.2 | 53.4 |
| Rural Western Maryland | 2.2 | 8.9 | 14.2 | 38.2 | 13.6 | 26.6 | 48.5 |
p<.05,
p<.01,
p<.001 by the χ2 test.
Enabling Factors for Clinical Trial Participation
The distribution of the nine factors that enable participation in clinical trials is presented in Table 4. African American respondents were less likely than their white counterparts to be influenced to participate by seven of the nine enabling factors (reimbursement, insurance coverage, greater knowledge, time commitment, follow-up care, and additional medical care). Important determinants for participation in clinical trials included childcare, not having to provide names, and transportation (especially for African American females).
Table 4. Factor Enabling Participation in Clinical Trials.
| Reimburse (n=2612) |
Insurance (n=3030) |
Transport (n=2022) |
Childcare (n=1144) |
Knowledge (n=3313) |
Time Comt. (n=2345) |
No Name (n=1927) |
Follow- up (n=3418) |
Added Care (n=3302) |
|
|---|---|---|---|---|---|---|---|---|---|
| % | % | % | % | % | % | % | % | % | |
| Race/Gender | *** | *** | *** | * | * | *** | ** | ||
| White male | 58.0 | 64.2 | 40.0 | 20.4 | 66.0 | 48.5 | 35.7 | 69.1 | 66.9 |
| White female | 51.3 | 63.1 | 39.2 | 22.1 | 68.8 | 48.5 | 40.9 | 71.3 | 68.5 |
| Black male | 42.8 | 49.7 | 38.6 | 28.2 | 64.2 | 48.4 | 40.2 | 59.8 | 61.4 |
| Black female | 44.0 | 51.3 | 44.3 | 26.9 | 62.2 | 44.6 | 41.0 | 59.5 | 60.5 |
| Other male | 48.5 | 57.1 | 47.8 | 35.3 | 73.5 | 43.9 | 45.6 | 67.6 | 67.2 |
| Other female | 54.5 | 47.8 | 52.2 | 29.8 | 68.2 | 52.2 | 39.1 | 67.4 | 67.4 |
| Age Group | *** | *** | *** | *** | *** | *** | *** | *** | |
| Under 65 | 57.3 | 64.8 | 40.6 | 26.8 | 69.9 | 49.9 | 41.8 | 70.6 | 69.6 |
| 65 and over | 37.0 | 48.5 | 39.1 | 10.0 | 57.2 | 42.2 | 29.6 | 60.5 | 56.2 |
| Health Status | *** | ** | |||||||
| Poor | 50.9 | 62.8 | 55.1 | 22.5 | 64.0 | 49.0 | 29.7 | 64.8 | 68.6 |
| Not poor* | 52.4 | 60.7 | 39.4 | 22.6 | 66.8 | 47.9 | 39.3 | 68.3 | 66.1 |
| Education | *** | *** | *** | *** | *** | ** | ** | *** | *** |
| Never attended | 63.6 | 45.5 | 36.4 | 36.4 | 63.6 | 45.5 | 45.5 | 63.6 | 63.6 |
| Elementary | 36.6 | 47.1 | 41.6 | 17.3 | 49.3 | 38.7 | 29.7 | 48.6 | 50.7 |
| Some high school | 45.1 | 51.6 | 45.1 | 29.0 | 58.5 | 45.8 | 32.6 | 55.0 | 53.8 |
| High school grad | 48.0 | 59.4 | 43.4 | 25.3 | 66.3 | 47.4 | 38.3 | 64.3 | 64.1 |
| Some college | 60.0 | 66.3 | 41.2 | 23.9 | 72.2 | 52.2 | 42.7 | 74.2 | 72.9 |
| College grad | 53.7 | 61.8 | 33.0 | 16.4 | 66.2 | 43.7 | 38.8 | 72.2 | 65.6 |
| Some graduate school | 60.3 | 68.3 | 41.1 | 16.5 | 65.5 | 54.9 | 36.2 | 80.4 | 77.5 |
| Graduate degree | 55.4 | 62.3 | 32.0 | 15.6 | 68.5 | 48.8 | 39.2 | 75.3 | 72.1 |
| Insurance Status | *** | *** | *** | *** | *** | ** | *** | *** | *** |
| No insurance | 57.3 | 63.3 | 42.8 | 28.7 | 68.8 | 47.9 | 43.0 | 70.8 | 70.4 |
| through work or union | 56.4 | 64.5 | 38.0 | 24.4 | 69.1 | 50.9 | 41.7 | 70.2 | 68.8 |
| Another's work or union | 55.8 | 62.3 | 38.6 | 25.9 | 68.6 | 45.7 | 40.7 | 69.8 | 66.0 |
| Purchased directly | 52.4 | 62.5 | 34.2 | 22.0 | 67.7 | 47.7 | 38.5 | 66.9 | 69.4 |
| Medicare | 41.0 | 53.5 | 44.3 | 13.4 | 60.5 | 44.5 | 31.1 | 64.0 | 60.2 |
| Medicaid | 58.5 | 64.2 | 57.1 | 44.3 | 73.9 | 57.1 | 38.1 | 72.5 | 73.1 |
| Other group | 49.3 | 51.5 | 40.8 | 18.3 | 64.8 | 44.3 | 40.0 | 65.2 | 61.4 |
| Veterans' Administration | 45.3 | 41.5 | 43.9 | 17.9 | 53.7 | 35.4 | 29.9 | 50.0 | 55.2 |
| Income Level | *** | *** | * | * | ** | ||||
| Less than $7,500 | 57.4 | 62.5 | 54.5 | 31.3 | 73.4 | 52.2 | 41.8 | 69.0 | 69.8 |
| $7,500-$14,999 | 52.3 | 61.1 | 51.6 | 26.9 | 63.5 | 49.3 | 39.8 | 65.7 | 67.4 |
| $15,000-$24,999 | 57.6 | 64.9 | 50.0 | 28.9 | 70.3 | 50.3 | 44.8 | 70.4 | 69.1 |
| $25,000-$34,999 | 54.6 | 64.8 | 44.7 | 26.5 | 71.0 | 50.8 | 41.6 | 68.8 | 68.8 |
| $35,000-$49,999 | 57.9 | 69.5 | 43.0 | 26.3 | 73.8 | 52.3 | 41.7 | 77.2 | 75.3 |
| $50,000-$74,999 | 58.1 | 65.4 | 37.3 | 21.1 | 70.0 | 51.0 | 40.4 | 73.8 | 69.6 |
| $75,000-$99,999 | 58.9 | 64.9 | 32.2 | 21.6 | 69.7 | 49.2 | 40.7 | 73.8 | 70.5 |
| $100,000 or more | 52.8 | 60.8 | 25.1 | 16.0 | 66.0 | 52.9 | 31.4 | 74.3 | 66.7 |
| Geographic Area | * | *** | *** | ** | ** | *** | *** | ||
| Urban Baltimore City | 49.2 | 52.4 | 39.1 | 22.9 | 61.7 | 45.3 | 38.3 | 60.6 | 61.2 |
| Rural Eastern Shore | 51.7 | 60.4 | 38.4 | 21.9 | 66.8 | 47.4 | 38.6 | 68.8 | 65.9 |
| Rural Western Maryland | 55.6 | 66.7 | 46.1 | 24.4 | 68.9 | 51.1 | 39.6 | 70.3 | 70.1 |
p<.05,
p<.01,
p<.001 by the χ2 test.
Respondents under age 65 were more likely to be influenced by five of the nine factors, including reimbursement, insurance coverage, increased knowledge, follow-up, and added medical care. Respondents aged 65 and older were likely to be influenced in their participation by increased knowledge, follow-up, and added care. Factors important for respondents who reported “poor” health status included reimbursement, insurance, transportation, more knowledge, follow-up and provision of additional health care.
Education level was associated with five enabling factors, including reimbursement (except for respondents who reported elementary or some high school education or high school graduate), insurance (especially for respondents with some high school or more education), more knowledge (except for respondents with elementary level education), follow-up care (except for respondents with elementary level education), and provision of additional health care. With a few exceptions, both uninsured and insured respondents were more likely to be influenced in their decision to participate in clinical trials by factors such as reimbursement (except those Medicare, other group insurance, or VA insurance), insurance coverage (except those with VA insurance), more knowledge, follow-up, and provision of additional health care. Geographically, residents of both urban and rural setting were more likely to be influenced by factors such as reimbursement (especially for residents of rural Western Maryland and Eastern Shore), insurance, more knowledge, follow-up, and provision of additional health care in their decision to participate.
Multivariate Predictors of Recruitment to and Participation in Clinical Trials
Multivariate stepwise logistic regression analyses were used to evaluate the relative contributions of socio-demographic characteristics, health status, insurance coverage, and geographic area of residence as predictors of recruitment to and participation in clinical trials. To reiterate, the sample for the multivariate stepwise logistic regression analyses included only the respondents who self-identified themselves as either black or white on race, and responded either “yes” or “no” on the predictor and outcome variables. Respondents excluded from the analyses were those who self-identified themselves as “other” for race and responded “do not know/unsure” or “refused” on the predictor and outcome variables. The multivariate model for participation also included as independent variables the sources of clinical trial information and enabling factors.
The independent predictors of recruitment to clinical trials are presented in Table 5. Respondents in poor health (OR = 1.83, CI = 1.21-2.76), having public health insurance coverage (OR = 1.98, CI = 1.57-2.51), and those having some college or higher level of education (OR = 2.32, CI = 1.84-2.92) were significantly more likely to be recruited. Respondents who were black (OR=0.61, CI= 0.44-0.85), residents of rural Western Maryland (OR=0.46, CI=0.33-0.65), and residents of rural Eastern Shore (OR=0.30, CI=0.22- 0.40) were significant less likely to be recruited. Among respondents recruited to clinical trials, blacks (OR = 0.38, CI = 0.21-0.68) and middle-income respondents (OR = 0.57, CI = 0.37-0.89) were less likely to participate (Table 6). In addition, respondents who were informed about clinical trials by their health care provider (OR = 1.69, CI = 1.08- 2.65), were knowledgeable about clinical trials (OR = 2.09, CI = 1.26-3.46), and who had the time commitment (OR = 1.67, CI = 1.06-2.63) were significantly more likely to actually participate.
Table 5. Multivariate Predictors of Recruited (“Ever Asked”) to Participate in Clinical Trials.
| Recruited to Clinical Trials (“Ever Asked”) | |||
|---|---|---|---|
| Odds Ratio | 95% Conf. | Interval | |
| Black | 0.61 | 0.44 | 0.85 |
| Poor health status | 1.83 | 1.21 | 2.76 |
| Resident of rural Western Maryland | 0.46 | 0.33 | 0.65 |
| Resident of rural Eastern Shore | 0.30 | 0.22 | 0.40 |
| Have public insurance | 1.98 | 1.57 | 2.51 |
| Some college or higher level of education | 2.31 | 1.84 | 2.92 |
Hosmer and Lemeshow Goodness-of-Fit Test p= 0.6746
Table 6. Multivariate Predictors of Participated in Clinical Trials (N=376).
| Participated in Clinical Trials | |||
|---|---|---|---|
| Odds Ratio | 95% Conf. | Interval | |
| Black | 0.38 | 0.21 | 0.68 |
| Middle income | 0.57 | 0.37 | 0.89 |
| Informed about clinical trials by a doctor | 1.69 | 1.08 | 2.65 |
| Knowledgeable about clinical trials | 2.09 | 1.26 | 3.46 |
| Have the time commitment to participate | 1.67 | 1.06 | 2.63 |
Hosmer and Lemeshow Goodness-of-Fit Test p= 0.9280
Discussion
Enhancing recruitment and participation in clinical trials, especially for minorities, is a national priority. Addressing avoidable disparities in participation requires better understanding about the specific factors that may lead to recruitment and participation. The results presented in this report shed light on factors associated with clinical trial recruitment and participation in a population-based, rural and urban sample covering 13 of the 24 jurisdictions in the state of Maryland. Moreover, the results are some of the first comprehensive, population-based estimates on clinical trial recruitment and participation prevalence and predictors.
Findings from this study indicate that in Maryland the significant predictors of being recruited included being African American, in poor health, receiving public health insurance, having some college or higher level of education, and residing in either rural Western Maryland or the Eastern Shore. In addition, significant predictors of participation included race/ethnicity, income level, health care professional providing information about clinical trials, knowledge about clinical trials, and time commitment. The television, radio and print media were the important sources for clinical trial information and the preference for these sources varied by socio-demographic factors. In terms of enabling factors, knowledge about clinical trials, insurance coverage for trials, reimbursement for participation, and time commitment required were some of the more important factors identified.
The predictors of self-reported participation in clinical trials in Maryland corroborate those documented in other studies. Similar to findings reported here, other studies have also reported that the likelihood of reported participation was lower among racial minorities (African Americans)2, 4, 7 and those with less income;2, 6, 7, 10, 11 and, the likelihood of reported participation in clinical trials was higher when health care professionals provided information about clinical trials 7, 10, 13 and participants had knowledge about clinical trials.10, 11, 18 An additional factor that may enhance participation in clinical trials is the time commitment required.
The findings presented in this report have several programmatic implications. The results provide evidence for the development of interventions aimed at the community that can be very specific for gender, race/ethnicity, rural and urban settings, educational attainment, and income levels. Moreover, these attributes coupled with the selective preference and influence of different sources of health information provide for the development of intervention that not only target by socio-demographic and geographical attributes but also by the method of communicating health information. Given the extremely low rates of self-reported recruitment, the results also provide evidence for the development of physician and other health professional awareness and education programs.
Future research on clinical trial recruitment and participation needs to explore the individual and structural dimensions. Individual-level dimensions would include the psychological, cultural, and social attributes that influence the decision-making processes of participation. These studies would be especially important to enhance recruitment to prevention trials where the potential benefit/harm is in the distant future. The structural dimensions would include the provision of clinical trials in rural, non-academic community-settings, and methods to enhance physician and health professional awareness and education about available trials. Other areas of research include the evaluation of innovative recruitment strategies in population-based settings.
Policy implications of this research are several. The data reported in this study underscore the importance of population-based studies covering urban and rural settings to document needs and behaviors relevant to specific communities. In addition, the data suggest the scarcity of recruitment efforts in rural settings, among racial/ethnicity minorities, and among lower socioeconomic groups. These data lend themselves to the definition of policy initiatives at the local (i.e., institutional), state, and national levels to address the prevailing disparities in clinical trial recruitment and participation.
Limitations of this study are those inherent in telephone survey research. The data, being self-reported, are subject to recall bias and social desirability bias. These biases may skew the estimations of the true prevalence of population-based recruitment and participation. Although the study utilized rigorous call attempt efforts to encourage potential participants to complete the survey and to contact potential participants, there remains the sample bias due to several factors including the repeated encounter with answering machines, refusal to participate in the survey, and the lack of telephones in households. There is some evidence to suggest that potential bias due to non-completion of telephone health surveys may be minimal.23
In summary, this study presents systematically documented population-based prevalence data and identified predictors for recruitment to and participation in clinical trials among Marylanders residing in urban and rural settings. The reported rate of recruitment to clinical trials is very low. It is likely that the reported low rate of recruitment and, subsequently, participation in clinical trials may explain, at least in part, the disparities in health observed among minorities, the underserved, and rural communities in Maryland. The data provide a foundation for targeted educational interventions aimed at both the community and health care professionals.
Table 1. Sample Characteristics of the 5,154 Survey Respondents by Study Region.
| Characteristics | All | Baltimore City | Eastern Shore |
Western Maryland |
||||
|---|---|---|---|---|---|---|---|---|
| n | % | n | % | n | % | n | % | |
| Age group, in years*** | ||||||||
| Less than 65 | 3830 | 74.7 | 546 | 80.5 | 2463 | 74.0 | 821 | 73.3 |
| 65 and over | 1298 | 25.3 | 132 | 19.5 | 867 | 26.0 | 299 | 26.7 |
| Race/ethnicity*** | ||||||||
| White | 4024 | 79.2 | 262 | 39.1 | 2691 | 81.7 | 1071 | 96.2 |
| Black | 935 | 18.4 | 377 | 56.3 | 537 | 16.3 | 21 | 1.9 |
| Other | 119 | 2.3 | 31 | 4.6 | 67 | 2.0 | 21 | 1.9 |
| Gender* | ||||||||
| Male | 2416 | 46.9 | 288 | 42.3 | 1572 | 46.9 | 556 | 49.6 |
| Female | 2738 | 53.1 | 393 | 57.7 | 1779 | 53.1 | 566 | 50.5 |
| Race/gender*** | ||||||||
| White male | 1943 | 38.3 | 130 | 19.4 | 1280 | 38.9 | 533 | 47.9 |
| White female | 2081 | 41.0 | 132 | 19.7 | 1411 | 42.8 | 538 | 48.3 |
| Black male | 374 | 7.4 | 137 | 20.5 | 230 | 7.0 | 7 | 0.6 |
| Black female | 561 | 11.1 | 240 | 35.8 | 307 | 9.3 | 14 | 1.3 |
| Other male | 71 | 1.4 | 16 | 2.4 | 41 | 1.2 | 14 | 1.3 |
| Other female | 48 | 1.0 | 15 | 2.2 | 26 | 0.8 | 7 | 0.6 |
| Health Status | ||||||||
| Poor | 279 | 5.4 | 31 | 4.6 | 175 | 5.2 | 73 | 6.5 |
| Not poor | 4867 | 94.6 | 648 | 95.4 | 3172 | 94.8 | 1047 | 93.5 |
| Education*** | ||||||||
| Never attended | 11 | 0.2 | 2 | 0.3 | 8 | 0.2 | 1 | 0.1 |
| Elementary | 158 | 3.1 | 27 | 4.0 | 89 | 2.7 | 42 | 3.8 |
| Some high school | 473 | 9.3 | 89 | 13.2 | 294 | 8.9 | 90 | 8.1 |
| High school graduate | 1767 | 34.7 | 195 | 29.0 | 1142 | 34.6 | 430 | 38.5 |
| Some college | 1291 | 25.4 | 146 | 21.7 | 827 | 25.1 | 318 | 28.5 |
| College graduate | 743 | 14.6 | 99 | 14.7 | 512 | 15.5 | 132 | 11.8 |
| Some graduate school | 144 | 2.8 | 31 | 4.6 | 94 | 2.9 | 19 | 1.7 |
| Graduate degree | 504 | 9.9 | 84 | 12.5 | 335 | 10.2 | 85 | 7.6 |
| Insurance Status*** | ||||||||
| Uninsured | 482 | 9.4 | 72 | 10.6 | 288 | 8.7 | 122 | 11.0 |
| Insured | ||||||||
| Through work or union | 2117 | 41.5 | 333 | 49.1 | 1355 | 40.9 | 429 | 38.6 |
| Someone else's work or union | 717 | 14.0 | 74 | 10.9 | 500 | 15.1 | 143 | 12.9 |
| Purchased directly | 324 | 6.4 | 39 | 5.8 | 225 | 6.8 | 60 | 5.4 |
| Medicare | 1200 | 2.5 | 112 | 16.5 | 796 | 24.0 | 292 | 26.3 |
| Medicaid | 125 | 2.5 | 29 | 4.3 | 64 | 1.9 | 32 | 2.9 |
| Other group | 73 | 1.4 | 9 | 1.3 | 45 | 1.4 | 19 | 1.7 |
| VA | 68 | 1.3 | 10 | 1.5 | 44 | 1.3 | 14 | 1.3 |
| Income Level*** | ||||||||
| Less than $7,500 | 220 | 5.6 | 43 | 8.2 | 115 | 4.6 | 62 | 7.0 |
| $7,500-$14,999 | 357 | 9.1 | 50 | 8.6 | 203 | 8.1 | 104 | 11.8 |
| $15,000-$24,999 | 486 | 12.4 | 65 | 12.5 | 294 | 11.7 | 127 | 14.4 |
| $25,000-$34,999 | 625 | 15.9 | 89 | 17.1 | 377 | 15.0 | 159 | 18.0 |
| $35,000-$49,999 | 656 | 16.7 | 83 | 15.9 | 421 | 16.7 | 152 | 17.2 |
| $50,000-$74,999 | 826 | 21.0 | 102 | 19.5 | 565 | 22.4 | 159 | 18.0 |
| $75,000-$99,999 | 372 | 9.5 | 42 | 8.1 | 259 | 10.3 | 71 | 8.0 |
| $100,000 or more | 386 | 9.8 | 48 | 9.2 | 288 | 11.4 | 50 | 5.7 |
Note: the sample sizes per variable may not add up to 5,154 due to missing values.
p<.05,
p<.01,
p<.001 by the χ2 test.
Acknowledgments
This research was supported by grants from the National Cancer Institute (1UO1CA86249-01), the National Center for Minority Health and Health Disparities (P60MD000532), and the Maryland Cigarette Restitution Fund Program. The contents of the article are solely the responsibility of the authors and do not necessarily represent the views of the funding agencies. The authors acknowledge the outstanding biostatistical contributions provided by Fanlun Meng.
Sources of support: This research was supported by grants from the National Cancer Institute (1UO1CA86249-01), the National Center for Minority Health and Health Disparities, (P60MD000532), and the Maryland Cigarette Restitution Fund Program. The contents of the article are solely the responsibility of the authors and do not necessarily represent the views of the funding agencies.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Comis RL, Miller JD, Aldige CR, Krebs L, Stoval E. Public attitudes toward participation in cancer clinical trials. J Clin Oncol. 2003 Mar 1;21(5):830–835. doi: 10.1200/JCO.2003.02.105. [DOI] [PubMed] [Google Scholar]
- 2.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 Apr 15;20(8):2109–2117. doi: 10.1200/JCO.2002.08.056. [DOI] [PubMed] [Google Scholar]
- 3.Freedman LS, Simon R, Foulkes MA, et al. Control Clin Trials. 5. Vol. 16. Oct, 1995. Inclusion of women and minorities in clinical trials and the NIH Revitalization Act of 1993--the perspective of NIH clinical trialists; pp. 277–285. discussion 286-279, 293-309. [DOI] [PubMed] [Google Scholar]
- 4.McCaskill-Stevens W, Pinto H, Marcus AC, et al. Recruiting minority cancer patients into cancer clinical trials: a pilot project involving the Eastern Cooperative Oncology Group and the National Medical Association. J Clin Oncol. 1999 Mar;17(3):1029–1039. doi: 10.1200/JCO.1999.17.3.1029. [DOI] [PubMed] [Google Scholar]
- 5.Murthy VH, Krumholz HM, Gross CP. Participation in cancer clinical trials: race-, sex-, and age-based disparities. Jama. 2004 Jun 9;291(22):2720–2726. doi: 10.1001/jama.291.22.2720. [DOI] [PubMed] [Google Scholar]
- 6.Giuliano AR, Mokuau N, Hughes C, et al. Participation of minorities in cancer research: the influence of structural, cultural, and linguistic factors. Ann Epidemiol. 2000 Nov;10(8 Suppl):S22–34. doi: 10.1016/s1047-2797(00)00195-2. [DOI] [PubMed] [Google Scholar]
- 7.Mandelblatt JS, Yabroff KR, Kerner JF. Equitable access to cancer services: A review of barriers to quality care. Cancer. 1999 Dec 1;86(11):2378–2390. [PubMed] [Google Scholar]
- 8.Unson CG, Dunbar N, Curry L, Kenyon L, Prestwood K. The effects of knowledge, attitudes, and significant others on decisions to enroll in a clinical trial on osteoporosis: implications for recruitment of older African-American women. J Natl Med Assoc. 2001 Oct;93(10):392–401. discussion 402-394. [PMC free article] [PubMed] [Google Scholar]
- 9.Zhu K, Hunter S, Bernard LJ, Payne-Wilks K, Roland CL, Levine RS. Recruiting elderly African-American women in cancer prevention and control studies: a multifaceted approach and its effectiveness. J Natl Med Assoc. 2000 Apr;92(4):169–175. [PMC free article] [PubMed] [Google Scholar]
- 10.Harris Y, Gorelick PB, Samuels P, Bempong I. Why African Americans may not be participating in clinical trials. J Natl Med Assoc. 1996 Oct;88(10):630–634. [PMC free article] [PubMed] [Google Scholar]
- 11.Brown DR, Fouad MN, Basen-Engquist K, Tortolero-Luna G. Recruitment and retention of minority women in cancer screening, prevention, and treatment trials. Ann Epidemiol. 2000 Nov;10(8 Suppl):S13–21. doi: 10.1016/s1047-2797(00)00197-6. [DOI] [PubMed] [Google Scholar]
- 12.Shavers-Hornaday VL, Lynch CF, Burmeister LF, Torner JC. Why are African Americans under-represented in medical research studies? Impediments to participation. Ethn Health. 1997 Mar-Jun;2(1-2):31–45. doi: 10.1080/13557858.1997.9961813. [DOI] [PubMed] [Google Scholar]
- 13.Albrecht TL, Blanchard C, Ruckdeschel JC, Coovert M, Strongbow R. Strategic physician communication and oncology clinical trials. J Clin Oncol. 1999 Oct;17(10):3324–3332. doi: 10.1200/JCO.1999.17.10.3324. [DOI] [PubMed] [Google Scholar]
- 14.Shavers VL, Lynch CF, Burmeister LF. Racial differences in factors that influence the willingness to participate in medical research studies. Ann Epidemiol. 2002 May;12(4):248–256. doi: 10.1016/s1047-2797(01)00265-4. [DOI] [PubMed] [Google Scholar]
- 15.Robinson SB, Ashley M, Haynes MA. Attitude of African-Americans regarding prostate cancer clinical trials. J Community Health. 1996 Apr;21(2):77–87. doi: 10.1007/BF01682300. [DOI] [PubMed] [Google Scholar]
- 16.Corbie-Smith G, Thomas SB, Williams MV, Moody-Ayers S. Attitudes and beliefs of African Americans toward participation in medical research. J Gen Intern Med. 1999 Sep;14(9):537–546. doi: 10.1046/j.1525-1497.1999.07048.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Mouton CP, Harris S, Rovi S, Solorzano P, Johnson MS. Barriers to black women's participation in cancer clinical trials. J Natl Med Assoc. 1997 Nov;89(11):721–727. [PMC free article] [PubMed] [Google Scholar]
- 18.Ellis PM, Butow PN, Tattersall MH, Dunn SM, Houssami N. Randomized clinical trials in oncology: understanding and attitudes predict willingness to participate. J Clin Oncol. 2001 Aug 1;19(15):3554–3561. doi: 10.1200/JCO.2001.19.15.3554. [DOI] [PubMed] [Google Scholar]
- 19.Swanson GM, Ward AJ. Recruiting minorities into clinical trials: toward a participant-friendly system. J Natl Cancer Inst. 1995 Dec 6;87(23):1747–1759. doi: 10.1093/jnci/87.23.1747. [DOI] [PubMed] [Google Scholar]
- 20.Baquet CR. Implications for Cancer Prevention and Control through a Statewide Health Network. Paper presented at: Governor's Conference on Cancer Disparities in Maryland; July 2000; Baltimore, MD. [Google Scholar]
- 21.Baquet CR, Commiskey P, Mishra SI, Mullins CD, Meng F. The University of Maryland Center for Health Policy/Health Services Research and University of Maryland Statewide Health Network Baseline Needs Assessment: Disparities in Health Access, Health Status, Health Assessment, and Health Behaviors for Chronic Disease and Clinical Trials Knowledge, Attitudes, and Barriers to Participation in Maryland. Baltimore, MD: University of Maryland School of Medicine, Office of Policy and Planning, University of Maryland Center for Health Policy/Health Services Research; Aug, 2004. [Google Scholar]
- 22.Aday LU. Designing and Conducting Health Surveys. San Francisco, CA: Jossey-Bass; 1989. [Google Scholar]
- 23.Mishra SI, Dooley D, Catalano R, Serxner S. Telephone health surveys: potential bias from noncompletion. Am J Public Health. 1993 Jan;83(1):94–99. doi: 10.2105/ajph.83.1.94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Lemeshow S, Hosmer DW., Jr Estimating odds ratios with categorically scaled covariates in multiple logistic regression analysis. Am J Epidemiol. 1984 Feb;119(2):147–151. doi: 10.1093/oxfordjournals.aje.a113732. [DOI] [PubMed] [Google Scholar]