Abstract
Objectives. We used results generated from the first study of the National Institutes of Health Sentinel Network to understand health concerns and perceptions of research among underrepresented groups such as women, the elderly, racial/ethnic groups, and rural populations.
Methods. Investigators at 5 Sentinel Network sites and 2 community-focused national organizations developed a common assessment tool used by community health workers to assess research perceptions, health concerns, and conditions.
Results. Among 5979 individuals assessed, the top 5 health concerns were hypertension, diabetes, cancer, weight, and heart problems; hypertension was the most common self-reported condition. Levels of interest in research participation ranged from 70.1% among those in the “other” racial/ethnic category to 91.0% among African Americans. Overall, African Americans were more likely than members of other racial/ethnic groups to be interested in studies requiring blood samples (82.6%), genetic samples (76.9%), or medical records (77.2%); staying overnight in a hospital (70.5%); and use of medical equipment (75.4%).
Conclusions. Top health concerns were consistent across geographic areas. African Americans reported more willingness to participate in research even if it required blood samples or genetic testing.
Although more than 80 000 clinical trials are conducted each year in the United States, less than 1% of the population participates in these studies.1 Women, the elderly, members of racial/ethnic minority groups, and rural populations are often underrepresented in research,2,3 leading to findings that do not account for genetic, cultural, linguistic, racial/ethnic, gender, and age differences. A representative population sample in clinical trials is important both from an ethical perspective and to ensure external validity of findings.3–5
The community health worker (CHW) intervention model could help eliminate this discrepancy. CHWs are lay community members who share a common language and culture with the people they serve.6 CHWs primarily deliver interventions, navigate patients through cancer treatment, educate patients about their diverse conditions, and facilitate connections to hidden populations. The value of the CHW in research is increasingly recognized; some states now certify CHWs to further professionalize their role.7,8 In fact, the American Journal of Public Health devoted its December 2011 issue to work associated with CHWs.7,9
Recognizing this potential, the Clinical and Translational Science Award (CTSA) Strategic Goal 4 Committee (Enhancing the Health of Our Communities and the Nation) of the National Institutes of Health (NIH) established the CTSA Community Engagement Resource Development Workgroup. This group subsequently received an American Reinvestment and Recovery Act supplemental grant in 2009 to collaborate across 5 CTSA sites and 2 community-focused national organizations to develop procedures to increase community participation in research, build the capacity of CHWs to expand their role in research by increasing the rigor of health evaluation metrics in the field, and establish a sustainable network, the Sentinel Network, to provide ongoing, real-time assessments of top health and neighborhood needs, concerns, and research perceptions. The data could then be immediately shared with NIH-CTSA sites, and local communities to increase the representativeness and relevance of research by facilitating community participation.
The Sentinel Network is built on an interactive community program called HealthStreet, founded by the lead author (L. B. C.) when she was at Washington University in St. Louis. HealthStreet uses CHWs for 4 aims: to assess the medical problems and health concerns of community residents, to engage in bidirectional sharing of information between the community and research investigators, to link community residents to medical and social services, and to provide community members with opportunities to participate in research.
HealthStreet builds trust between the research community and the general public6,10 by directly engaging individual residents (the program’s first aim) rather than focusing on agencies, providers, and organizations as partners, as is more typical in community engagement research.6 This strategy ensures that HealthStreet’s efforts do not bypass the input of community members or inadvertently privilege the perceptions of community leaders and service providers.11–13
Community members often perceive research as primarily meeting the needs of the researcher.14,15 Recognition of these and other issues has led to an increasingly robust literature on methods to engage the community in research.10,14,16,17 In line with HealthStreet’s second aim to engage in bidirectional sharing of information between communities and researchers, this literature stresses the need for community members to provide meaningful input into study content. With the discrepancy between participants enrolled in research studies and the populations to whom findings relate, efforts are needed that effectively link people to desired medical and social services (the third aim) and that use this exchange to inform individuals of opportunities for relevant research (the fourth aim).
Past studies evaluating barriers to research participation indicate that members of underrepresented groups have concerns with respect to inconvenience of study times, other logistical burdens, and fears about being a guinea pig.12,18,19 That mistrust, directed at researchers and health professionals, is decreasing as a result of new community engaged research efforts.20,21 Programs such as the CTSA Sentinel Network and HealthStreet can help change these perceptions.
Here we report findings generated from the Sentinel Network’s first study on the assessment of health needs and concerns at 5 diverse CTSA sites. We describe the Sentinel Network study protocol, assessment, and results from the first phase of the collaboration.
METHODS
The Sentinel Network builds the capacity of CHWs to engage individuals in their communities by discussing their health concerns and priorities, conferring with them about their research experiences and expectations, and linking them to research opportunities. The Sentinel Network collaboration in place at the time of this study included 5 Resource Development Workgroup sites: Washington University in St. Louis, Missouri (L. B. C.), the University of Rochester, Rochester, New York (N. M. B.), the University of Michigan, Ann Arbor (M. D.), the Albert Einstein College of Medicine, Bronx, New York (H. S.), and the University of California, Davis (S. A.) Sentinel Network sites are geographically and demographically diverse and were in varying stages of their CTSA funding at the time collaboration was initiated.
In addition, 2 community-focused national organizations participated in the training of CHWs and the establishment of guidelines for community engagement efforts: Community-Campus Partnerships for Health (S. D. S.) and Patient Advocates in Research (D. E. C.). Although each CTSA site had unique approaches to engaging its communities, each agreed to this CHW model and contributed to and agreed on all of the methods described.
Recruitment and Training of Community Health Workers
Each site received funds for one full-time-equivalent CHW; additional resources were leveraged from the CTSA, the community engagement program, volunteers, practicum students, and graduate students for additional CHWs. Of the 9 CHWs hired, 8 were female, 5 were African American, 2 were White, 1 was Asian, and 1 was Hispanic. CHWs ranged in age from 22 to 52 years (mean = 35.8), and they had a diversity of experience.
Although there were no set criteria, CHWs were required to be gregarious and outgoing members of their community. Specifically, one CHW managed a student-run free clinic, another had a nursing background, and others were volunteers in their community. CHWs attended a 2-day training session in St. Louis that focused on ethics, privacy, confidentiality, risk management, the intent of each assessment question, protocol methodology, medical terminology, and the methods to be used in recruiting participants and gathering results. Training included vigorous interactive role-playing to solidify each concept.
Participant Recruitment and Assessment
CHWs recruited adults 18 years or older for the Sentinel Network study between March 2010 and September 2011 at places where people congregate, such as barbershops, beauty shops, parks, shelters, bus stops, community agencies, churches, neighborhood associations, health care facilities, sports venues, grocery stores, laundromats, nail salons, fitness centers, colleges, gas stations, check cashing venues, and health fairs. CHWs spoke with community residents individually, providing each with a brief, standardized explanation of the study. A total of 5979 individuals participated in the study.
Sentinel Network partners developed a common CHW-administered assessment that was vigorously pilot tested by CHWs and principal investigators with respect to its acceptability, feasibility, and contextual relevance. A Spanish-translation version was developed by the team at the University of California, Davis. Although each site had the option of including additional items, a core set of questions approximately 15 minutes in length formed the basis for the Sentinel Network intake at all sites.
The CHW assessed the location of the contact (with GPS coordinates), the date and time of contact, and the location’s zip code and recorded the individual’s observed gender and self-reported age, race/ethnicity, height and weight, and last grade completed. The CHW elicited from participants their top 3 health concerns and their top neighborhood concern. Participants were then asked whether they had ever been told by a health professional that they had a problem with high blood pressure, depression, heart disease, diabetes, cancer, or arthritis; a disease of the muscles or bones; asthma; or a kidney problem. CHWs then elicited participants’ smoking history and health insurance status. Next, they asked a series of questions about participants’ research participation history (beginning with “Have you ever been in a health research study?”) and their perceptions regarding research, after which they asked participants to indicate what they believed was “fair market” compensation for a 1.5-hour study that included a blood test. The final question elicited participants’ level of interest in taking part in a research study.
Data Management and Quality Control
Sites sent hard copies of their data to the Washington University team monthly. Data entry and quality control were completed at Washington University; feedback was given to each site on a regular basis to ensure protocol fidelity. General issues were discussed during biweekly conference calls with all team members.
Analyses
Data were analyzed by site and by participant race/ethnicity. At Washington University, 3.1% of individuals who were contacted decided not to participate. Although such records were not kept at other sites, observations indicated that the rate of nonparticipation seemed to be equivalent across the sites. We used means and standard deviations to calculate continuous variables and proportions and binomial confidence intervals to calculate categorical variables. SAS version 9.2 (SAS Institute, Cary, NC) was used in conducting the analyses.
RESULTS
The sample was racially/ethnically diverse (Table 1). The majority of Sentinel Network participants were African American (51.6%; n = 3058). Those who reported being Hispanic or Latino (regardless of other race) were counted as such and removed from any other category; they made up 16.9% (n = 1001) of the sample. In addition, 5.8% reported being Asian (n = 341) and 1.3% reported being biracial or multiracial (n = 82). American Indians, Alaska Natives, Pacific Islanders, and Middle Easterners took part in the study but made up less than 1% of the sample. White participants accounted for 22.3% (n = 1325) of the sample.
TABLE 1—
Demographic Characteristics of Sentinel Network Participants, by Site: 2010–2011
| Washington University in St. Louis (n = 2253) |
University of California, Davis (n = 1030) | University of Michigan (n = 1049) |
Albert Einstein College of Medicine (n = 714) |
University of Rochester (n = 933) |
Total (n = 5979) | ||||||
| Characteristic | No. (%; 95% CI) or Mean ±SD | Metropolitan %a | No. (%; 95% CI) or Mean ±SD | Metropolitan %a | No. (%; 95% CI) or Mean ±SD | Metropolitan %a | No. (%; 95% CI) or Mean ±SD | Metropolitan %a | No. (%; 95% CI) or Mean ±SD | Metropolitan %a | No. (%; 95% CI) or Mean ±SD |
| Hispanic/Latino | 24 (1.1; 0.6, 1.5) | 3.1 | 476 (46.4; 43.3, 49.5) | 19.5 | 59 (5.7; 4.3, 7.1) | 3.9 | 335 (48.0; 44.3, 51.7) | 51.1 | 107 (11.7; 9.6, 13.8) | 5.9 | 1001 (16.9; 15.9, 17.8) |
| Asian | 6 (0.2; 0.1 0.5) | 3.1 | 157 (15.3; 13.1, 17.5) | 14.9 | 145 (13.9; 11.8, 16.0) | 3.0 | 26 (3.7; 2.3, 5.1) | 3.6 | 7 (0.8; 0.2, 1.3) | 3.3 | 341 (5.8; 5.2, 6.3) |
| African American | 1890 (83.9; 82.4, 85.4) | 44.8 | 175 (17.1; 14.8, 19.4) | 8.6 | 301 (28.9; 26.1, 31.6) | 32.0 | 195 (27.9; 24.6, 31.3) | 30.5 | 497 (54.4; 51.2, 57.7) | 12.8 | 3058 (51.6; 50.3, 52.8) |
| White | 309 (13.7; 12.3, 15.1) | 47.1 | 161 (15.7; 13.5, 18.1) | 53.0 | 497 (47.7; 44.7, 50.7) | 59.1 | 89 (12.8; 10.4, 15.5) | 12.8 | 269 (29.5; 26.5, 32.4) | 76.6 | 1325 (22.3; 21.3, 23.4) |
| Biracial/multiracial | 15 (0.7; 0.3, 1.0) | 1.8c | 22 (2.1; 1.3, 3.0) | 4.0c | 15 (1.4; 0.7, 2.2) | 1.9c | 16 (2.3; 1.2, 3.4) | 1.9c | 14 (1.5; 0.7, 2.3) | 1.5c | 82 (1.3; 1.1, 1.7) |
| Otherb | 8 (0.4; 0.1, 0.6) | 35 (3.4; 2.3, 4.5) | 25 (2.4; 1.5, 3.3) | 37 (5.3; 3.6, 7.0) | 19 (2.1; 1.2, 3.0) | 124 (2.1; 1.7, 2.5) | |||||
| Female | 1204 (53.4; 51.4, 55.5) | 545 (53.2; 50.1, 56.3) | 706 (67.6; 64.6, 70.4) | 499 (71.3; 67.8, 74.6) | 570 (62.8; 59.6, 60.7) | 3524 (59.4; 58.2, 60.7) | |||||
| High school diploma | 1611 (72.1; 70.2, 74.0) | 718 (72.1; 69.2, 74.9) | 974 (93.6; 92.1, 95.1) | 493 (72.2; 68.7, 75.5) | 793 (87.4; 85.3, 89.6) | 4589 (78.3; 77.2, 79.4) | |||||
| Age, y | 39.2 ±13.7 | 42.3 ±14.6 | 40.9 ±19.0 | 46.1 ±15.1 | 46.5 ±15.3 | 42.0 ±15.6 | |||||
| Body mass index | 29.2 ±7.6 | 28.5 ±6.4 | 27.3 ±7.2 | 28.3 ±6.4 | 29.5 ±7.0 | 28.7 ±7.1 | |||||
Note. CI = confidence interval. Means and percentages were derived according to how many people answered a particular question.
Data on metropolitan populations > 18 years were obtained from the 2010 census. The metropolitan population for Washington University in St. Louis included St. Louis City County, MO. The metropolitan population for the University of California, Davis, included Sacramento and Yolo counties. The metropolitan population for the University of Michigan included Washtenaw, Wayne, and Genessee counties. The metropolitan metro population for the Albert Einstein College of Medicine included Bronx County, NY. The metropolitan metro population for the University of Rochester included Monroe County, NY.
Alaska Native/Eskimo, American Indian, Middle Eastern, and Pacific Islander totals were reported as ≤ 1%. In the 2010 census data, Other included Biracial/Multiracial populations.
Biracial/Multiracial and Other combined percentage.
Underrepresented populations were specifically targeted and thus were more highly represented than in the 2010 census demographic data for their community (Table 1). Washington University, the University of Michigan, the University of Rochester, and the University of California, Davis, oversampled African Americans, Hispanics/Latinos, or Asians relative to their area population. The Albert Einstein College of Medicine produced a sample roughly consistent with the local demographic profile of its area (≥ 80% of area residents belong to minority groups).
Other demographic characteristics are shown in Table 2. More than half of the respondents at each site were women (59.4% overall). The mean age of participants was 42.0 years, and 78.3% had graduated from high school. The mean body mass index (defined as weight in kilograms divided by the square of height in meters) was 28.7 across all sites, which is considered overweight.22 Mean ages stratified by race/ethnicity ranged from 35.8 years among Asians to 46.0 years among those classified as “Other.” The percentage of participants with a high school diploma ranged from 61.7% among Hispanics to 94.8% among Asians. Mean body mass index ranged from 24.0 among Asians to 29.6 among African Americans.
TABLE 2—
Demographic Characteristics of Sentinel Network Participants, by Race/Ethnicity: 2010–2011
| White (n = 1325) | African American (n = 3058) | Hispanic/Latino (n = 1001) | Asian (n = 341) | Biracial/Multiracial (n = 82) | Other (n = 124) | |
| Characteristic | No. (%; 95% CI) or Mean ±SD | No. (%; 95% CI) or Mean ±SD | No. (%; 95% CI) or Mean ±SD | No. (%; 95% CI) or Mean ±SD | No. (%; 95% CI) or Mean ±SD | No. (%; 95% CI) or Mean ±SD |
| High school diploma | 1177 (89.4; 87.7, 91.0) | 2315 (76.7; 75.2, 78.2) | 587 (61.7; 58.6, 64.8) | 311 (94.8; 91.8, 97.0) | 70 (86.4; 79.0, 93.0) | 93 (77.5; 70.0, 85.0) |
| Age, y | 44.3 ±16.8 | 41.8 ±14.9 | 41.2 ±14.4 | 35.8 ±17.0 | 38.3 ±14.9 | 46.0 ±17.3 |
| Body mass index | 27.7 ±6.9 | 29.6 ±7.6 | 28.9 ±6.2 | 24.0 ±4.1 | 28.7 ±6.4 | 26.8 ±5.7 |
Note. CI = confidence interval. Means and percentages were derived according to how many people answered a particular question.
Health and Neighborhood Concerns
The 5 most frequently mentioned health concerns across sites, among those reporting at least 1 concern (n = 5319; 89.0%), were hypertension, diabetes, cancer, weight, and heart problems (Table 3). Although not shown, the top 5 concerns did not vary according to respondents’ age or race/ethnicity.
TABLE 3—
Top Health and Neighborhood Concerns of Sentinel Network Participants, by Site: 2010–2011
| Washington University in St. Louis | University of California, Davis | University of Michigan | Albert Einstein College of Medicine | University of Rochester | Total | |
| Top health concernsa | ||||||
| Sample size | n = 1983 | n = 945 | n = 987 | n = 588 | n = 816 | n = 5319 |
| First | Hypertension | Diabetes | Cancer | Hypertension | Cancer | Hypertension |
| Second | Diabetes | Cancer | Weight | Diabetes | Diabetes | Diabetes |
| Third | Cancer | Hypertension | Diabetes | Weight | Hypertension | Cancer |
| Fourth | Heart problems | Heart problems | Heart problems | Cancer | Weight | Weight |
| Fifth | Weight | Weight | Hypertension | Heart problems | Heart problems | Heart problems |
| Top neighborhood concernsa | ||||||
| Sample size | n = 1598 | n = 759 | n = 854 | n = 346 | n = 661 | n = 4186 |
| First | Safety/crime | Safety/crime | Safety/crime | Health | Safety/crime | Safety/crime |
| Second | Drugs | Health | Health | Safety/crime | Health | Health |
| Third | Health | Drugs | Environment | Environment | Economy | Drugs |
| Fourth | Environment | Environment | Economy | Drugs | Drugs | Environment |
| Fifth | Economy/sexually transmitted diseases | Health care | Peace/respect | Youth programs | Health care | Economy |
Among those with a concern.
Table 3 also shows the top neighborhood concerns by site, again among those reporting at least 1 such concern (n = 4186; 70.0%). Respondents at all of the sites, with the exception of the Albert Einstein College of Medicine, mentioned safety and crime as the most salient concern. Another concern in the top 5 among all sites was health; drugs and environment were mentioned as priorities by respondents from 4 of the 5 sites. When stratified by age, sexually transmitted diseases replaced economy as a top neighborhood concern among respondents younger than 30 years (data not shown). Hispanic respondents reported health care as one of their top 5 neighborhood concerns.
Data on actual health conditions and topics were elicited with yes–no questions; concerns are shown in Table 4. Overall, high blood pressure was most commonly cited, followed by depression and arthritis. With the exception of respondents at the Washington University site, high blood pressure was the most prevalent health condition reported. The least common conditions reported overall were heart disease and cancer, each at 4.8%.
TABLE 4—
Health Conditions, Insurance Status, and Smoking Behavior Reported by Sentinel Network Participants, by Site and Race/Ethnicity: 2010–2011
| Site |
Race/Ethnicity |
|||||||||||
| Reported Health Condition | Washington University in St. Louis (n = 2253), No. (%; 95% CI) | University of California, Davis (n = 1030), No. (%; 95% CI) | University of Michigan (n = 1049), No. (%; 95% CI) | Albert Einstein College of Medicine (n = 714), No. (%; 95% CI) | University of Rochester (n = 933), No. (%; 95% CI) | Total (n = 5979), No. (%; 95% CI) | White (n = 1325) | African American (n = 3058) | Hispanic/Latino (n = 1001) | Asian (n = 341) | Biracial/Multiracial (n = 82) | Other (n = 124) |
| Arthritis | 427 (19.8; 18.1, 21.5) | 190 (20.0; 17.5, 22.7) | 213 (20.3; 17.9, 22.9) | 155 (22.0; 19.0, 25.2) | 195 (25.0; 22.0, 28.2) | 1180 (20.9; 19.9, 22.0) | 350 (27.3; 24.9, 29.8) | 597 (21.0; 19.5, 22.5) | 159 (17.0; 14.6, 19.4) | 31 (9.4; 6.3, 12.5) | 20 (24.4; 15.1, 33.7) | 16 (13.7; 7.5, 19.9) |
| Asthma | 408 (18.0; 16.5, 19.7) | 169 (17.9; 15.4, 20.4) | 158 (15.1; 13.0, 17.4) | 110 (15.6; 13.0, 18.5) | 169 (21.6; 18.8, 24.7) | 1014 (17.7; 16.7, 18.7) | 231 (17.9; 15.8, 20.0) | 553 (18.8; 17.4, 20.2) | 158 (16.9; 14.5, 19.3) | 32 (9.7; 6.5, 12.9) | 16 (19.5; 10.9, 28.1) | 16 (13.7; 7.5, 19.9) |
| Cancer | 74 (3.3; 2.6, 4.0) | 39 (4.1; 2.9, 5.6) | 79 (7.6; 6.1, 9.3) | 32 (4.5; 3.1, 6.3) | 51 (6.6; 5.0, 8.6) | 275 (4.8; 4.3, 5.4) | 100 (7.8; 6.3, 9.2) | 120 (4.1; 3.4, 4.8) | 32 (3.4; 2.3, 4.6) | 8 (2.4; 0.8, 4.1) | 5 (6.1; 0.9, 11.3) | 7 (5.9; 1.7, 10.2) |
| Depression | 582 (25.8; 24.0, 27.6) | 293 (28.8; 26.0, 31.6) | 245 (23.4; 20.9, 26.1) | 124 (17.5; 14.8, 20.5) | 225 (24.9; 22.1, 27.8) | 1469 (24.8; 23.7, 25.9) | 478 (36.2; 33.6, 38.8) | 659 (21.7; 20.3, 23.2) | 235 (23.6; 21.0, 26.3) | 25 (7.4; 4.6, 10.1) | 30 (36.6; 26.2, 47.0) | 33 (27.3; 19.3, 35.2) |
| Diabetes | 194 (8.7; 7.5, 9.9) | 161 (15.8; 13.6, 18.2) | 132 (12.6; 10.7, 14.8) | 105 (14.9; 12.3, 17.7) | 148 (16.3; 14.0, 18.9) | 740 (12.5; 11.7, 13.4) | 134 (10.2; 8.6, 11.9) | 367 (12.2; 11.0, 13.3) | 166 (16.7; 14.4, 19.1) | 28 (8.3; 5.3, 11.2) | 13 (15.9; 8.0, 23.8) | 24 (19.8; 12.7, 26.9) |
| Diseases of the muscles or bones | 840 (37.3; 35.3, 39.3) | 96 (10.1; 8.3, 12.2) | 76 (7.3; 5.8, 9.0) | 53 (7.5; 5.7, 9.7) | 69 (8.9; 7.0, 11.2) | 1134 (19.8; 18.8, 20.8) | 238 (18.5; 16.4, 20.6) | 751 (25.6; 24.0, 27.2) | 102 (10.9; 8.9, 13.0) | 20 (6.1; 3.5, 8.6) | 16 (19.5; 10.9, 28.1) | 5 (4.3; 0.6, 7.9) |
| Heart disease | 38 (1.7; 1.2, 2.2) | 79 (7.8; 6.2, 9.6) | 76 (7.3; 5.8, 9.0) | 28 (4.0; 2.6, 5.7) | 60 (6.7; 5.1, 8.5) | 281 (4.8; 4.2, 5.3) | 79 (6.0; 4.7, 7.3) | 115 (3.8; 3.1, 4.5) | 49 (5.0; 3.6, 6.3) | 17 (5.0; 2.7, 7.3) | 11 (13.4; 6.0, 20.8) | 8 (6.6; 2.2, 11.0) |
| High blood pressure | 600 (26.6; 24.8, 28.5) | 355 (34.8; 32.0, 37.9) | 292 (27.9; 25.2, 30.7) | 240 (34.1; 30.6, 37.7) | 351 (38.1; 34.9, 41.3) | 1838 (30.9; 29.8, 32.1) | 372 (28.2; 25.8, 30.6) | 1005 (33.0; 31.4, 34.7) | 312 (31.5; 28.6, 34.4) | 59 (17.3; 13.3, 21.3) | 28 (34.2; 23.9, 44.4) | 46 (37.7; 29.1, 46.3) |
| Kidney problem | 245 (10.9; 9.6, 12.2) | 83 (8.8; 7.0, 10.7) | 45 (4.3; 3.2, 5.7) | 29 (4.1; 2.8, 5.9) | 49 (6.4; 4.7, 8.3) | 451 (7.9; 7.2, 8.6) | 106 (8.3; 6.8, 9.8) | 245 (8.4; 7.4, 9.4) | 74 (7.9; 6.2, 9.7) | 12 (3.6; 1.6, 5.7) | 9 (11.1; 4.3, 18.0) | 3 (2.6; 0.0, 5.4) |
| Other health topics | ||||||||||||
| Has any type of health insurance | 1095 (48.8; 46.7, 50.8) | 507 (49.8; 46.6, 52.9) | 873 (83.5; 81.2, 85.7) | 594 (83.7; 80.7, 86.3) | 776 (84.4; 81.8, 86.6) | 3845 (64.7; 63.5, 66.0) | 977 (74.1; 71.8, 76.5) | 1857 (61.2; 59.4-62.9) | 563 (56.6; 53.5, 59.7) | 249 (73.0; 68.3, 77.7) | 65 (79.3; 70.5, 88.0) | 96 (78.1; 70.7, 85.4) |
| Has smoked cigarettes in the past 30 d | 1217 (54.5; 52.4, 56.5) | 283 (27.8; 25.0, 30.6) | 172 (16.5; 14.3, 18.9) | 170 (23.9; 20.9, 27.3) | 208 (22.6; 20.0, 25.5) | 2050 (34.6; 33.4, 35.8) | 431 (32.7; 30.2, 35.3) | 1295 (42.8; 41.0-44.5) | 235 (23.7; 21.0, 26.3) | 23 (6.7; 4.1, 9.4) | 33 (40.2; 29.6, 50.9) | 27 (22.0; 14.6, 29.3) |
Note. CI = confidence interval. Percentages were derived according to how many people answered a particular question.
Health conditions were examined according to race/ethnicity as well (Table 4). In the case of all conditions, Asians reported the lowest prevalence of any racial/ethnic group. Whereas high blood pressure was the most prevalent condition reported among African Americans (33.0%), Hispanics (31.5%), and Asians (17.3%), the most prevalent condition reported among Whites was depression (36.2%). Enrollment in health insurance ranged from 56.6% among Hispanics/Latinos to 79.3% among biracial or multiracial participants. Current smoking rates ranged from 6.7% among Asians to 42.8% among African Americans.
Research Experiences and Perceptions
Overall, 87.3% of the respondents (ranging from 71.5% at the University of Rochester site to 99.7% at the Washington University site) were interested in participating in research (Table 5). With the exception of Washington University, respondents at all sites reported that they were least willing to participate in a study in which they would have to take medication (mean = 49.3%; overall range: 27.8%–68.1%). Participants stated that they would be most willing to take part in a study wherein they were asked about only their health (mean = 85.2%; range: 76.1%–93%). At the Washington University site, participants overwhelmingly reported being willing to provide a blood sample (95.2%), relinquish their medical records (92.4%), and provide a sample for genetic purposes (91.8%). With respect to fair compensation for a study that involved a 1.5-hour interview and a blood draw, responses (means) ranged from $63.41 among participants at the University of Rochester site to $79.72 among participants at the Albert Einstein College of Medicine site (overall mean = $73.54).
TABLE 5—
Research Experience and Perceptions Reported by Sentinel Network Participants, by Site and Race/Ethnicity: 2010–2011
| Site |
Race/Ethnicity |
|||||||||||
| Experience or Perception | Washington University, St. Louis (n = 2253), No. (%; 95% CI) | University of California, Davis (n = 1030), No. (%; 95% CI) | University of Michigan (n = 1049), No. (%; 95% CI) | Albert Einstein College of Medicine (n = 714), No. (%; 95% CI) | University of Rochester (n = 933), No. (%; 95% CI) | Total (n = 5979), No. (%; 95% CI) | White (n = 1325) | African American (n = 3058) | Hispanic/Latino (n = 1001) | Asian (n = 341) | Biracial/Multiracial (n = 82) | Other (n = 124) |
| Has ever been in a health research study | 394 (17.8; 16.2, 19.4) | 102 (10.1; 8.3, 12.0) | 271 (25.9; 23.3, 28.7) | 80 (11.3; 9.1, 13.9) | 158 (17.3; 14.9, 20.0) | 1005 (17.1; 16.1, 18.0) | 279 (21.2; 19.0, 23.4) | 521 (17.4; 16.0, 18.7) | 113 (11.4; 9.4, 13.4) | 49 (4.4; 10.7, 8.1) | 17 (21.0; 12.1, 29.9) | 19 (15.5; 9.1, 21.8) |
| Interested in participating in research studies | 1857 (99.7; 99.3, 99.9) | 785 (83.8; 81.3, 86.1) | 901 (86.1; 83.8, 88.1) | 558 (78.5; 75.3, 81.5) | 553 (71.5; 68.2, 74.7) | 4654 (87.3; 86.4, 88.2) | 1067 (85.5; 83.5, 87.5) | 2357 (91.0; 89.9, 92.1) | 782 (84.5; 82.2, 86.9) | 263 (79.7; 75.4, 84.0) | 69 (86.3; 78.7, 93.8) | 82 (70.1; 61.8, 78.4) |
| Would participate in a study | ||||||||||||
| If asked only about health | 2090 (93.0; 92.0, 94.1) | 847 (83.5; 81.0, 85.7) | 871 (83.2; 80.8, 85.4) | 541 (76.1; 72.8, 79.2) | 701 (77.3; 74.4, 80.0) | 5050 (85.2; 84.3, 86.1) | 1143 (87.3; 85.5, 89.1) | 2655 (87.5; 86.3, 88.7) | 816 (82.2; 79.8, 84.6) | 252 (73.9; 69.2, 78.6) | 66 (80.5; 71.9, 89.1) | 83 (68.0; 59.8, 76.3) |
| If researchers wanted to see your medical records | 2075 (92.4; 91.3, 93.5) | 726 (71.7; 68.8, 74.4) | 628 (60.0; 56.9, 63.0) | 439 (61.8; 58.1, 65.4) | 423 (46.6; 43.4, 50.0) | 4291 (72.5; 71.3, 73.6) | 946 (72.1; 69.6, 74.5) | 2339 (77.2; 75.7, 78.7) | 689 (69.5; 66.7, 72.4) | 173 (50.9; 45.6, 56.2) | 51 (63.0; 52.5, 73.5) | 70 (56.9; 48.2, 65.7) |
| If you had to give a blood sample | 2140 (95.2; 94.4, 96.1) | 770 (75.9; 73.2, 78.5) | 690 (65.9; 62.9, 68.8) | 464 (65.4; 61.7, 68.9) | 565 (62.2; 59.0, 65.4) | 4629 (78.1; 77.1, 79.2) | 1021 (77.7; 75.5, 80.0) | 2505 (82.6; 81.3, 84.0) | 747 (75.5; 72.8, 78.1) | 195 (57.2; 51.9, 62.4) | 61 (74.4; 64.9, 83.8) | 73 (59.4; 50.7, 68.0) |
| If asked to give a genetic sample | 2062 (91.8; 90.7, 92.9) | 718 (70.9; 68.0, 73.7) | 600 (57.3; 54.3, 60.3) | 412 (58.0; 54.3, 61.7) | 462 (51.3; 48.0, 54.7) | 4254 (71.9; 70.8, 73.1) | 940 (71.8; 69.4-74.3) | 2328 (76.9; 75.4, 78.4) | 657 (66.4; 63.4, 69.3) | 176 (51.6; 46.3, 56.9) | 61 (74.4; 64.9, 83.8) | 67 (54.9; 46.1, 63.8) |
| If you had to take medicine | 1527 (68.1; 66.2, 70.1) | 520 (51.7; 48.6, 54.8) | 319 (30.5; 27.7, 33.3) | 292 (41.3; 37.6, 45.0) | 250 (27.8; 24.8, 30.8) | 2908 (49.3; 48.0, 50.6) | 638 (48.7; 46.0, 51.5) | 1599 (53.0; 51.2, 54.8) | 478 (48.4; 45.3, 51.6) | 94 (27.7; 22.9, 32.4) | 40 (48.8; 38.0, 59.6) | 47 (38.8; 30.2, 47.5) |
| If asked to stay overnight in a hospital or clinic | 1896 (84.6; 83.1, 86.1) | 584 (57.7; 54.6, 60.8) | 515 (49.2; 46.1, 52.3) | 329 (46.4; 42.7, 50.2) | 399 (43.8; 40.5, 47.0) | 3723 (62.9; 61.7, 64.1) | 796 (60.6; 57.9, 63.2) | 2133 (70.5; 68.9, 72.1) | 558 (56.4; 53.3, 59.5) | 112 (32.8; 27.9, 37.8) | 48 (58.5; 47.9, 69.2) | 55 (45.1; 36.3, 53.9) |
| If you might have to use medical equipment | 2023 (90.2; 89.0, 91.5) | 697 (68.9; 65.9, 71.7) | 582 (55.6; 52.5, 58.6) | 390 (54.9; 51.2, 58.6) | 426 (47.2; 43.9, 50.6) | 4118 (69.6; 68.5, 70.8) | 897 (68.4; 65.9, 70.9) | 2278 (75.4; 73.8, 76.9) | 631 (63.7; 60.7, 66.7) | 169 (49.6; 44.3, 54.9) | 56 (68.3; 58.2, 78.4) | 67 (54.9; 46.1, 63.8) |
| If you didn’t get paid | 1198 (64.3; 62.2, 66.5) | 639 (68.9; 65.8, 71.8) | 553 (52.9; 49.8, 55.9) | 400 (56.7; 53.0, 60.4) | 419 (52.1; 48.6, 55.6) | 3209 (60.0; 58.7, 61.4) | 751 (60.1; 57.4, 62.8) | 1572 (60.1; 58.2, 62.0) | 604 (65.9; 62.8, 68.9) | 158 (48.5; 43.0, 53.9) | 49 (60.5; 49.9, 71.1) | 58 (50.4; 41.3, 59.6) |
| Fair compensation for a study with a 1.5-h interview and a blood test, $ | 78.25 | 67.52 | 71.57 | 79.72 | 63.41 | 73.54 | 58.34 | 81.60 | 75.15 | 68.71 | 61.68 | 65.52 |
Note. CI = confidence interval. Percentages were derived according to how many people answered a particular question.
We also assessed research interest according to race/ethnicity. Rates of past study participation were low in all groups, ranging from 11.4% among Hispanics/Latinos to 21.2% among Whites (Table 5). Interest in study participation was higher among respondents in the “other” racial/ethnic category (70.1%) and among African American respondents (91.0%).
More African Americans than respondents in any other racial/ethnic group were interested in participating in a study (91%), providing a blood sample (82.6%), providing a sample for a genetic study (76.9%), providing access to medical records (77.2%), staying overnight in a hospital (70.5%), and using medical equipment (Table 5). Asian respondents were least willing to report interest along each of these categories.
Regarding fair compensation for a study 1.5 hours in duration that involved a blood draw, African Americans were highest in their expectations (mean = $81.60), followed by Hispanics (mean = $75.15). The lowest end of the range was $61.68, the average reported by Whites.
Sharing Findings With Communities
Efforts were made to share Sentinel Network findings with the communities in which the data were collected as well as among other researchers. A community brochure was produced in which summarized study findings were presented. To ensure bidirectional sharing of information, we shared the data with our community advisory boards and asked for their input into the development of study materials.
DISCUSSION
The goals of the Resource Development Workgroup—to collaborate across CTSA sites, to build the capacity of CHWs, to increase the rigor of health evaluation metrics in the field, and to establish a network for ongoing, real-time assessment of health and neighborhood needs and concerns—are being met. With the data obtained, CTSA sites are working with their local communities to increase the representativeness and relevance of research by facilitating community participation.
The first phase of the Sentinel Network study focused on assessing community needs and bidirectional communication through a new network of 5 sites and 2 national organizations. Findings about health concerns were consistent according to site, race/ethnicity, and age, with hypertension, diabetes, and cancer among the top concerns. These self-reported health concerns are among the top 25 priority research topics listed in a recent Institute of Medicine report.23 Similarly, the conditions that were most likely to be reported by Sentinel Network respondents, hypertension (30.9%) and depression (24.8%), are consistent with national prevalence rates of these conditions.24,25
In addition, neighborhood concerns were consistent across site, race/ethnicity, and age groups, with safety and crime ranked as the top concern (an issue also raised by participants in a series of Federal Interagency Working Group sessions held around the country). Such concerns call attention to the need to understand and evaluate the social determinants of health: “the conditions in which people live, learn, work, play, and pray.”26(pS15)
A majority of Sentinel Network respondents reported interest in taking part in health research and had clear opinions about the kinds of research in which they would be willing to participate; they also identified fair levels of compensation. Across all sites, participants were most likely to report willingness to take part in a study that is noninvasive and does not require a hospital stay, medical equipment, or taking medication. Their reported rate of previous participation in health studies was well above the national average of less than 1%.1
One of our salient findings is that African Americans reported more willingness to participate in research than respondents from other racial/ethnic groups. Interestingly, this included more willingness to be involved in studies in which blood and genetic samples were drawn from participants, which may suggest that the reluctance of African Americans to participate in medical research as a result of the US Public Health Service study of syphilis in Tuskegee, Alabama,27–29 is waning. This finding confirms other recent study results suggesting that African Americans may not be less willing than members of other racial/ethnic groups to participate in clinical studies30,31 and may even be overrepresented in certain areas of research, such as phase I trials.32 One genetic study showed that when greater efforts were made to contact African Americans, their rate of participation was higher than that of Whites.33
Thus, our results contradict previous studies indicating that African Americans are less willing to participate in medical research.12,34–36 The willingness of our respondents to participate in clinical/health research bodes well for the new NIH National Center for Advancing Translational Sciences, which aims to facilitate the implementation of diagnostics, therapeutics, and devices pertinent to improving health37 and speed the translation of basic discoveries into therapies.38
Limitations and Strengths
Although the Sentinel Network study has many strengths, its limitations include the use of a convenience sample, the brevity of the assessment, and the cross-sectional nature of the data. These limitations are overshadowed by the network’s strengths, which include its ability to scale up efforts to collect real-time data from underrepresented populations commonly missed by most clinical research trials and community health endeavors. This maximizes the generalizability of the findings. Given that non-White racial/ethnic groups now constitute more than one third of the population in the United States,39 their inclusion in clinical trials is increasingly important.3
A better understanding of racial/ethnic variations in clinical trial participation and health research outcomes is critical in helping meet the health care needs of an increasingly diverse US population. Thus, the CTSA mission to enhance the health of local communities requires innovative methods that reach out to underrepresented populations. With the Sentinel Network model, this aim was achieved through ongoing multidirectional feedback loops that included community members, agency staff and leaders, researchers, and NIH.
Another strength is the large, geographically and ethnically diverse sample of 5979 community members. The diversity of this sample would not have been possible without CHWs, who, as a result of their experience and training, are respectful and culturally sensitive. These qualities are more important than homophily for successful community engagement.40
Conclusions
The reliance on CHWs in the Sentinel Network assessment model was found to be successful and feasible, and it led to a further expansion of the model. Specifically, a second phase of the Sentinel Network is currently being implemented. In this phase, along with a sixth site (University of Florida) being added, participants are actively being recruited into a study that links them to medical and social services as well as opportunities to participate in research. CHWs then track the success and satisfaction of participants at 1-month intervals; this pilot is intended as a first step toward meeting the HealthStreet program’s aims and planning future collaborative comparative effectiveness research across these and other CTSA sites.
The results of this study show that CHWs are crucial to the success of community-engaged research. With its focus on expanding the role of CHWs to include research and an emphasis on reducing health disparities by facilitating broader research participation, the Sentinel Network is an innovative approach that challenges the status quo and works toward person-centered research to promote better health outcomes for all.
Acknowledgments
This project was supported by federal funds from the National Center for Research Resources and the National Center for Advancing Translational Sciences (NCATS) through the Clinical and Translational Science Awards Program (grants NCRR UL1 RR024992-03S2, NCRR 2 UL1 RR024146-06, NCRR UL1RR024986, NCRR UL1RR025750, P60MD00514, NCRR UL1RR024992-03S3, NCRR UL1RR024992-05S1, and NCRR UL1RR029890).
We acknowledge our team of community health workers, Linda Alexander, Jazmin Amparo, Anita Johnson, Khai Nguyen, Ashley Neuhaus, Collette Noel, Marie Watkins, and Lisa Wines, and our Sentinel Network project coordinator, Dan Martin.
Human Participant Protection
The Sentinel Network protocol was approved by each site’s institutional review board (IRB), with Washington University serving as the coordinating center. Community health workers completed standard IRB training at their respective institutions. Interested participants provided verbal informed consent at some sites if identifiable data were not collected (University of Rochester, University of Michigan, University of California, Davis) and signed consent forms at the other sites (Washington University, Albert Einstein College of Medicine).
References
- 1.Mozes A. Report claims clinical trials miss many populations. Available at: http://health.usnews.com/usnews/health/healthday/080401/report-claims-clinical-trials-miss-many-populations.htm. Accessed December 2, 2012
- 2.Murthy VH, Krumholz HM, Gross CP. Participation in cancer clinical trials: race-, sex-, and age-based disparities. JAMA. 2004;291(22):2720–2726 [DOI] [PubMed] [Google Scholar]
- 3.Rochon PA, Mashari A, Cohen Aet al. The inclusion of minority groups in clinical trials: problems of under representation and under reporting of data. Account Res. 2004;11(3–4):215–223 [DOI] [PubMed] [Google Scholar]
- 4.Osler M, Schroll M. Differences between participants and non-participants in a population study on nutrition and health in the elderly. Eur J Clin Nutr. 1992;46(4):289–295 [PubMed] [Google Scholar]
- 5.Patel MX, Doku V, Tennakoon L. Challenges in recruitment of research participants. Adv Psychiatr Treat. 2003;9:229–238 [Google Scholar]
- 6.Cottler LB, Callahan C, Striley CL. HealthStreet: A community-based approach to include mental health in public health research. In: Cottler LB, ed. Mental Health in Public Health: The Next 100 Years. New York, NY: Oxford University Press; 2011:247–273 [Google Scholar]
- 7.Balcazar H, Rosenthal EL, Brownstein JN, Rush CH, Matos S, Hernandez L. Community health workers can be a public health force for change in the United States: three actions for a new paradigm. Am J Public Health. 2011;101(12):2199–2203 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Terpstra J, Coleman KJ, Simon G, Nebeker C. The role of community health workers (CHWs) in health promotion research: ethical challenges and practical solutions. Health Promot Pract. 2011;12(1):86–93 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Landers SJ, Stover GN. Community health workers—practice and promise. Am J Public Health. 2011;101(12):2198. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Cottler LB, Nagarajan R. Real-time assessment of community health needs and concerns. Sci Transl Med. 2012;4(119):119mr2. [DOI] [PubMed] [Google Scholar]
- 11.Yancey AK, Ortega AN, Kumanyika SK. Effective recruitment and retention of minority research participants. Annu Rev Public Health. 2006;27:1–28 [DOI] [PubMed] [Google Scholar]
- 12.Corbie-Smith G, Thomas SB, St. George DM. Distrust, race, and research. Arch Intern Med. 2002;162(21):2458–2463 [DOI] [PubMed] [Google Scholar]
- 13.Striley CL, Callahan C, Cottler LB. Enrolling, retaining, and benefiting out-of-treatment drug users in intervention research. J Empir Res Hum Res Ethics. 2008;3(3):19–25 [DOI] [PubMed] [Google Scholar]
- 14.McCloskey DJ, McDonald MA, Cook Jet al. Community engagement: definitions and organizing concepts from the literature. In: Silberberg M, ed. Principles of Community Engagement. 2nd ed. Bethesda, MD: National Institutes of Health; 2011:1–41 [Google Scholar]
- 15.Aungst J. Exploring Challenges, Progress, and New Models for Engaging the Public in the Clinical Research Enterprise: Clinical Research Roundtable Workshop Summary. Washington, DC: National Academies Press; 2003 [Google Scholar]
- 16.Ford JG, Howerton MW, Bolen Set al. Knowledge and access to information on recruitment of underrepresented populations to cancer clinical trials. Evid Rep Technol Assess (Summ). 2005;122:1–11 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Seifer SD, Michaels M, Collins S. Applying community-based participatory research principles and approaches in clinical trials: forging a new model for cancer clinical research. Prog Community Health Partnersh. 2010;4(1):37–46 [DOI] [PubMed] [Google Scholar]
- 18.Giuliano AR, Mokuau N, Hughes Cet al. Participation of minorities in cancer research: the influence of structural, cultural, and linguistic factors. Ann Epidemiol. 2000;10(suppl 8):S22–S34 [DOI] [PubMed] [Google Scholar]
- 19.Comis R, Aldige CR, Stovall EL, Krebs LU, Risher PJ, Taylor HJ. A Quantitative Survey of Public Attitudes Towards Cancer Clinical Trials. Philadelphia, PA: Coalition of National Cancer Cooperative Groups, Cancer Research Foundation of America, and Cancer Leadership Council and Oncology Nursing Society; 2000 [Google Scholar]
- 20.Fouad MN, Partridge E, Green BLet al. Minority recruitment in clinical trials: a conference at Tuskegee, researchers and the community. Ann Epidemiol. 2000;10(suppl 8):S35–S40 [DOI] [PubMed] [Google Scholar]
- 21.Horowitz CR, Robinson M, Seifer S. Community-based participatory research from the margin to the mainstream: are researchers prepared? Circulation. 2009;119(19):2633–2642 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults. WMJ. 1998;97(9):20–21, 24–25, 27–37 [PubMed] [Google Scholar]
- 23.100 Initial Priority Topics for Comparative Effectiveness Research. Washington, DC: Institute of Medicine; 2009 [Google Scholar]
- 24.Roger VL, Go AS, Lloyd-Jones DMet al. Heart disease and stroke statistics—2012 update: a report from the American Heart Association. Circulation. 2012;125(1):e2–e220 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Kessler RC, Berglund P, Demler Oet al. The epidemiology of major depressive disorder: results from the National Comorbidity Survey Replication (NCS-R). JAMA. 2003;289(23):3095–3105 [DOI] [PubMed] [Google Scholar]
- 26.Gracia JN, Koh HK. Promoting environmental justice. Am J Public Health. 2011;101(suppl 1):S14–S16 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Protection of human subjects; Belmont Report: notice of report for public comment. Fed Regist. 1979;44(76):23191–23197 [PubMed] [Google Scholar]
- 28.How Tuskegee Changed Research Practice. Atlanta, GA: Centers for Disease Control and Prevention; 2009 [Google Scholar]
- 29.Guidelines for the Conduct of Research Involving Human Subjects at the National Institutes of Health. Washington, DC: US Dept of Health and Human Services; 2004 [Google Scholar]
- 30.Brown DR, Topcu M. Willingness to participate in clinical treatment research among older African Americans and whites. Gerontologist. 2003;43(1):62–72 [DOI] [PubMed] [Google Scholar]
- 31.Wendler D, Kington R, Madans Jet al. Are racial and ethnic minorities less willing to participate in health research? PLoS Med. 2006;3(2):e19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Fisher JA, Kalbaugh CA. Challenging assumptions about minority participation in US clinical research. Am J Public Health. 2011;101(12):2217–2222 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Hartz SM, Johnson EO, Saccone NL, Hatsukami D, Breslau N, Bierut LJ. Inclusion of African Americans in genetic studies: what is the barrier? Am J Epidemiol. 2011;174(3):336–344 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Thompson HS, Valdimarsdottir HB, Jandorf L, Redd W. Perceived disadvantages and concerns about abuses of genetic testing for cancer risk: differences across African American, Latina and Caucasian women. Patient Educ Couns. 2003;51(3):217–227 [DOI] [PubMed] [Google Scholar]
- 35.Brandon DT, Isaac LA, LaVeist TA. The legacy of Tuskegee and trust in medical care: is Tuskegee responsible for race differences in mistrust of medical care? J Natl Med Assoc. 2005;97(7):951–956 [PMC free article] [PubMed] [Google Scholar]
- 36.Williams MM, Scharff DP, Mathews KJet al. Barriers and facilitators of African American participation in Alzheimer disease biomarker research. Alzheimer Dis Assoc Disord. 2010;24(suppl):S24–S29 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Collins FS. Reengineering translational science: the time is right. Sci Transl Med. 2011;3(90):90cm17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Kaiser J. Acting director Thomas Insel explains new NIH Translational Center’s aims and structure. Available at: http://news.sciencemag.org/scienceinsider/2012/01/acting-director-thomas-insel-exp.html. Accessed December 2, 2012
- 39.Humes KR, Jones NA, Ramirez RR. Overview of Race and Hispanic Origin: 2010. Washington, DC: US Census Bureau; 2011 [Google Scholar]
- 40.Striley CW, Margavio C, Cottler LB. Gender and race matching preferences for HIV post-test counselling in an African-American sample. AIDS Care. 2006;18(1):49–53 [DOI] [PubMed] [Google Scholar]