Abstract
Introduction:
Recruiting racial/ethnic minorities in smoking cessation trials is a priority. This study described lessons learned from recruiting a diverse sample of African American, White, and Hispanic/Latinx smokers in a smoking cessation trial.
Methods:
We implemented a 42-month recruitment campaign utilizing reactive (e.g., word-of-mouth, newspaper, radio, online ads, flyers, community partnerships) and proactive (e.g., direct invitations) strategies. We included 821 participants in the analysis. We described our recruitment strategies’ implementation, their enrollment yield and rate (number enrolled/number screened) by race/ethnicity, and direct cost-per-participant (CPP: total cost/number of enrolled) for paid strategies.
Results:
Enrollment yields were higher using reactive strategies than proactive strategies (94.3% vs. 5.7%). The top source of enrollment was word-of-mouth among African Americans (36%) and Whites (44%), and flyers among Hispanics/Latinxs (34%). Proactive recruitment, word-of-mouth, and flyers were more successful among African Americans than other groups. Newspaper and online ads were more successful among Hispanics/Latinxs than other groups (P<.05). Word-of-mouth was cost-free and yielded 23.1% of enrollment. The most economic method among paid strategies was flyer distribution (CPP = $47.6; yield 17.5%), followed by newspaper ($194.7; 23.7%) and online advertisements ($264.6; 24.0%). Radio and television ads were the most expensive and produced the least participant yield ($4,755.6; 0.8%).
Conclusion:
Recruiting racially/ethnically diverse samples into smoking cessation clinical trials requires implementing multiple strategies and adjusting these strategies based on their enrollment yield and cost. Word-of-mouth, flyers, and newspaper and online ads were more successful among racial/ethnic minorities. Flyers and newspaper ads were the most economic methods for recruitment.
Keywords: Recruitment, Racial/ethnic minorities, Smoking Cessation, Clinical Trial, Cost Analysis
1. Introduction
Despite progress in reducing cigarette smoking in the United States (U.S.), racial/ethnic minority smokers still have less access to tobacco treatment than Whites (Trinidad et al., 2011; USDHHS et al., 2014), and bear a disproportionate burden of tobacco-related health consequences (Health & Services, 1998; Jemal et al., 2008). Smoking cessation trials are essential to develop effective treatments and improve cessation outcomes. However, racial/ethnic minority populations have been historically underrepresented in these trials and their recruitment has been a challenge among researchers (Cokkinides et al., 2008; Fagan et al., 2004; USDHHS et al., 2014). Adequate representation of minorities is needed in smoking-related studies to ensure that findings are applicable to diverse populations. The National Institutes of Health (NIH) mandated minorities’ inclusion in all randomized clinical trials (RCT) as part of the NIH Revitalization Act of 1993 (NIH, 1993). However, although minorities currently represent 23% of smokers in the US, their representation in smoking cessation trials is 16% (Dickerson et al., 2009). Therefore, identifying effective recruitment strategies to increase minorities’ participation in smoking cessation trials is a high priority.
Several barriers to minorities’ participation in smoking cessation research have been identified (King et al., 2011). These factors include the lack of cultural competency within the research setting (e.g., language-appropriate messaging) (Fujimoto, 1998); logistical limitations (e.g., transportation) (Yancey et al., 2006); and strict exclusion criteria (e.g., co-morbid medical, psychiatric conditions) (Hooper et al., 2019). Another important factor that might affect minorities’ participation in RCTs is how much targeting minorities was considered in the plannign of RCTs(Hussain‐Gambles et al., 2004). A systematic review evaluating recruitment strategies in smoking cessation studies indicated that intensive personal contact and the use of a combination of reactive (e.g., flyers) and proactive (e.g., direct invitations) strategies significantly increase recruitment (Belisario et al., 2012). However, there is a dearth of research examining strategies for enrolling minorities who smoke and results from that research were mixed. For example, two studies found that proactive recruitment was the most effective method for recruiting African American adults (Brodar et al., 2016; Webb et al., 2009). In contrast, others have found that reactive recruitment using mass mailings and media was effective among African American (Fouad et al., 2004; Yancey et al., 2006; Harris et al., 2003). Even fewer studies have examined strategies to recruit Hispanic/Latinx smokers. We found only two: the first study found that compared to other groups, Hispanics/Latinxs who smoke were more responsive to internet ads (e.g., Craigslist) and roadside signs, and less responsive to flyers and postcards (Brodar et al., 2016). The second study found that Facebook was the most effective method for recruiting this population (Medina-Ramirez et al., 2020). Furthermore, few studies have examined the costs of different recruitment strategies (Kredo et al., 2012), highlighitng the need for more studies to identify efficient strategies to recruit racial/ethnic minority smokers and reduce smoking-related health disparities.
The Quitville Study is a dualsite two-arm RCT aimed at enrolling equal samples of the three largest racial/ethnic groups in the U.S. (African Americans, Hispanics/Latinxs, Whites) to evaluate the efficacy of group cognitive behavioral therapy (CBT) versus General Health Education (GHE) intervention in eliminating racial/ethnic differences in smoking cessation (Hooper et al., 2018). With the aim of aiding reseachers recruiting racial/ethnic minority participants in smoking cessation trials, in this study we report: (1) recruitment strategies that we implemented, (2) their enrollment yield and effeciency by race/ethnicity groups, and (3) the direct cost-per-participant (CPP) of paid strategies. Lessons learned from this experience will provide insight into several recruitment strategies that may help increase minorities participation in smoking cessation research.
2. Methods
2.1. Overview of Clinical Trial Design
This study was approved by Institutional Review Boards of the University of Miami and the Moffitt Cancer Center (Clinicaltrials.gov NCT02511236). Details regarding design and methods of the trial appeared in (Hooper et al., 2018). Briefly, the study used stratified random assignment by race/ethnicity (African Americans, Hispanics/Latinxs, and Whites). Both interventions provided eight counseling sessions and eight weeks of nicotine patch therapy. Assessments occurred at end-of-therapy and at 3, 6, and 12-months. The primary abstinence outcome was biochemically confirmed 7-day point prevalence abstinence at 12-months. The two recruitment study sites were Miami and Tampa Bay metropolitan areas. The goal was to recruit most of African Americans and Hispanics from Miami, and most of Whites from Tampa. Therefore, Tampa focused on recruiting non-Hispanic Whites using Facebook because it was very effective in reaching Whites. The current study reports on recruitment strategies only in Miami.
2.2. Participants
Our target sample size was 180 participants (60 per racial/ethnic group). However, due to high attrition rate (40%), we increased our sample to 252 (84 per racial/ethnic group) to keep the study powered. A screening questionnaire over the telephone or in person was administered to potential participants (n=849). Inclusion criteria were being an adult (>18 years), self-identification as African American/Black, White, or Hispanic/Latinx (any race), smoking at least 5 cigarettes/day or having a carbon monoxide reading of 8 ppm or more, interested in quitting, and speaking/reading English or Spanish. Exclusion criteria were contraindications for nicotine patch therapy, cognitive, physical, or mental health conditions that inhibited group treatment, alcoholism or illicit drug use, current tobacco treatment, barriers to session attendance (e.g., transportation), and inappropriateness for the study (e.g., aggressive). We also assessed how smokers learned about the trial “How did you hear about the Quitville study?”. All those who were screened and self-reported how they learned about the study were included in the analysis.
2.3. Recruitment Strategies and their Cost
We implemented a 42-month recruitment campaign utilizing a combination of reactive and proactive recruitment strategies simultaneously to enroll participants from the community in Miami, Florida. Below we describe in detail the recruitment strategies and their cost.
Proactive recruitment. We directly contacted potential participants from previous trials to invite them to participate in the study.
- Reactive recruitment strategies were:
- Word-of-mouth. Participants heard about the trial from someone who may or may not have been in the trial. This strategy was cost-free.
- Newspaper advertisements. Color advertisements were displayed for 2–8 weeks, 1–2 times each year, in Miami New Times (cost $5,200.00) and Sun Sentinel (cost $1,608.00). In the last year, we purchased display advertisements targeting Spanish-only speakers in El-Clarin (cost $1,356.000) and Las Americas (cost $500.00).
- Flyers on public transportation. We ran color display advertisements for 4 weeks, 1–2 times a year in the interior of Metrorail cars and county buses (cost $3,000.00).
- Community partnership. We distributed and mailed general and culturally targeted flyers to community-based organizations such as churches, health care clinics, health fairs, and local businesses throughout the duration of the study. Our community partnership was active and bidirectional with churches as we provided other health educational services in return based on their request.
- Online advertisements. We placed paid and free advertisements on Craigslist in the jobs, volunteer, community classes, and event sections 2 to 3 times per week. In year 2, a website for the study was launched to provide information about the trial and how to contact the research staff for those who were interested in participation. Web banner advertisements were also featured simultaneously in online editions of the Miami paper, costing $1,600.00.
- Radio/television announcements. In year 1, we produced 15-second announcements about the study on CBS Radio for 11 consecutive days ($3,500.00). The Miami press generated four television spots (14 minutes total), four newspaper stories, and two 15-minute radio interviews (during a popular early morning FM show and on an AM channel owned by the same station), and public service announcements aired on those two Miami radio stations (cost $2,880.00).
2.4. Adjustment of recruitment strategies
Strategies that did not achieve enough yield on investment of personnel time or cost were discontinued and new strategies were implemented. For example, using the radio/television in year one was costly and resulted in enrolling only three participants. Therefore, we discontinued this strategy afterwards. In addition, it was difficult to achieve our target sample size of Hispanics/Latinxs during the first three years because only English-only speakers were eligible to participate in the study. Therefore, we expanded our inclusion criteria in year 4 to additionally include Hispanics who are Spanish-only speakers in the study. We also made the intervention available in Spanish and hired Spanish-speaking staff to deliver the intervention in Spanish and implemented a targeted recruitment campaign for Spanish-only speakers. This adjustment allowed us to reach our target sample size of Hispanics by enrolling 15 extra participants.
2.5. Statistical Analysis
We used standard descriptive statistics based primarily on counts and proportions to describe the sample demographic characteristics (Table 1), and recruitment yield and efficiency by racial/ethnic groups and recruitment strategies (Table 2). Racial/ethnic differences in screening, eligibility (ratio of number eligible to number screened), and enrollment (ratio of number enrolled to number screened) by recruitment strategies when utilizing non-targeted recruitment were calculated (Table 3). The Chi-squared goodness of fit test was used for comparing the enrollment rate across three racial/ethnic groups by recruitment sources. Overall type-I error (alpha) was set at 5% for statistical hypothesis testing procedures. Participants who were recruited using Spanish targeted strategies (n = 15) were excluded from this analysis. We also estimated the direct cost of implementing each paid recruitment strategy, the number and rate of participants who were enrolled by each strategy, and the corresponding direct CPP (direct cost divided by number of enrolled participants) for each strategy (n=254) (Table 4). We only included the direct expenses of services by independent providers. Proactive recruitment was not included in the cost analysis because we did not track cost data for staff time (to develop, manage, and implement the strategies). The cost of word-of-mouth was also not included because it was impossible to estimate. Statistical analysis was completed using SAS v9.4 statistical software (SAS Institute Inc. Cary, NC, USA).
Table 1.
Participant Characteristics (n=821)
| Characteristics | N (%) |
|---|---|
| Screened | 821 (100) |
| Eligible | 487 (59.3) |
| Enrolled | 211 (25.7) |
| Age (mean ± SD) | 48.4 ±11.6 |
| Sex | |
| Female | 387 (47.1) |
| Male | 434 (52.9) |
| Race/Ethnicity | |
| African American | 396 (48.2) |
| Hispanic/Latinx | 329 (41.1) |
| White | 96 (11.7) |
Table 2.
| Recruitment Sources | Screened N (%) |
Eligible N (%) |
Enrolled N (%) |
||
|---|---|---|---|---|---|
| Yield | Efficiency | Yield | Efficiency | ||
| All | 821 (100) | 487 (100) | 487 (59.3) | 211 (100) | 211 (25.7) |
| Proactive Recruitment | 36 (4.4.) | 16 (3.3) | 16 (44.4) | 12 (5.7) | 12 (33.3) |
| Reactive Recruitment | 786 (95.6) | 471 (96.7) | 471 (60.0) | 197 (94.3) | 197 (25.4) |
| Word-of-Mouth | 276 (33.6) | 153 (31.4) | 153 (55.6) | 68 (32.5) | 68 (24.7) |
| Newspaper | 107 (13.0) | 78 (6.0) | 78 (72.2) | 25 (12.0) | 25 (24.1) |
| Flyer | 237 (28.8) | 125 (25.7) | 125 (52.7) | 61 (29.2) | 61 (26.2) |
| Community Partner | 47 (5.7) | 26 (5.3) | 26 (56.5) | 16 (7.7) | 16 (34.8) |
| Online | 112 (13.6) | 83 (17.0) | 83 (74.1) | 25 (12.0) | 25 (22.3) |
| Radio & Television | 7 (0.9) | 6 (1.2) | 6 (85.7) | 2 (1.0) | 2 (28.6) |
| African American | |||||
| All | 397 (100) | 203(100) | 203(51.3) | 98 (100) | 99 (25.0) |
| Proactive Recruitment | 23 (5.8) | 12 (5.9) | 12 (52.2) | 9 (9.2) | 9 (39.1) |
| Reactive Recruitment | 374 (94.2) | 191 (94.1) | 191 (51.2) | 90 (90.9) | 90 (24.1) |
| Word-of-Mouth | 194 (48.9) | 102 (50.2) | 102 (52.6) | 48 (49.0) | 48 (24.7) |
| Newspaper | 23 (5.8) | 10 (4.9) | 10 (43.5) | 1 (4.3) | 1 (4.3) |
| Flyer | 100 (25.2) | 43 (21.2) | 43 (43.0) | 26 (26.5) | 26 (27.0) |
| Community Partner | 25 (6.3) | 14 (6.9) | 14 (58.3) | 8 (8.2) | 8 (33.3) |
| Online | 32 (8.1) | 22 (10.8) | 22 (68.8) | 6 (6.1) | 6 (18.8) |
| Radio & Television | - | - | - | - | |
| Hispanic/Latino | |||||
| All | 329 (100) | 217(100) | 217(66.0) | 89 (100) | 89 (27.1) |
| Proactive Recruitment | 6 (1.8) | 2 (0.9) | 2 (33.3) | 2 (2.3) | 2 (33.3) |
| Reactive Recruitment | 323 (98.2) | 215 (99.1) | 215 (66.6) | 86 (97.7) | 87 (26.9) |
| Word-of-Mouth | 63 (19.1) | 38 (17.5) | 38 (59.7) | 13 (14.8) | 13 (21.0) |
| Newspaper | 56 (17.0) | 47 (21.7) | 47 (84.2) | 19 (21.6) | 19 (35.1) |
| Flyer | 118 (35.9) | 69 (31.8) | 69 (58.5) | 31 (35.2) | 31 (26.3) |
| Community Partner | 17 (5.2) | 8 (3.7) | 8 (47.1) | 6 (6.8) | 6 (35.3) |
| Online | 64 (19.5) | 49 (22.6) | 49 (76.6) | 15 (17.0) | 15 (23.4) |
| Radio & Television | 5 (1.5) | 4 (1.8) | 4 (80.0) | 2 (2.3) | 2 (40.0) |
| White | |||||
| All | 96 (100) | 67 (100) | 67 (69.8) | 23 (100) | 23 (24.0) |
| Proactive Recruitment | 7 (7.3) | 2 (3.0) | 2 (28.6) | 1 (4.3) | 1 (14.3) |
| Reactive Recruitment | 89 (92.7) | 65 (97.0) | 65 (73.0) | 22 (95.7) | 22 (24.7) |
| Word-of-Mouth | 19 (19.8) | 14 (20.9) | 14 (73.7) | 7 (30.4) | 7 (36.8) |
| Newspaper | 28 (29.2) | 20 (29.9) | 20 (71.4) | 5 (21.7) | 5 (17.9) |
| Flyer | 19 (19.8) | 13 (19.4) | 13 (68.4) | 4 (17.4) | 4 (21.1) |
| Community Partner | 5 (5.2) | 4 (6.0) | 4 (80.0) | 2 (8.7) | 2 (40.0) |
| Online | 16 (16.7) | 12 (17.9) | 12 (75.0) | 4 (17.4) | 4 (25.0) |
| Radio & Television | 2 (2.1) | 2 (3.0) | 2 (100.0) | - | - |
Yield: number of participants reached;
Efficiency: the effect of the strategy in meeting eligibility criteria and be enrolled in the study; Highlighted numbers are the top in their category.
Table 3.
Differences in screening, eligibility, and enrollment ratesa by recruitment strategies and racial/ethnic groupsb.
| Recruitment Strategies | African American N (%) |
Hispanic/Latino N (%) |
White N (%) |
P-value |
|---|---|---|---|---|
| Screening | ||||
| All | N=397 | N=329 | N=96 | |
| Proactive Recruitment | 23 (63.9) | 6 (16.7) | 7 (19.4) | 0.0005 |
| Reactive Recruitment | ||||
| Word-of-Mouth | 194 (70.3) | 63 (22.8) | 19 (6.9) | <0.0001 |
| Newspaper | 23 (21.5) | 56 (52.3) | 28 (26.2) | 0.0001 |
| Flyer | 100 (42.2) | 118 (49.8) | 19 (8.0) | <0.0001 |
| Community Partner | 25 (53.2) | 17 (36.2) | 5 (10.6) | 0.0016 |
| Online | 32 (28.6) | 64 (57.1) | 16 (14.3) | <0.0001 |
| Radio & Television | --- | 5 (71.4) | 2 (28.6) | --- |
| Eligibility | ||||
| All | N=203 | N=217 | N=67 | |
| Proactive Recruitment | 12 (75.0) | 2 (12.5) | 2 (12.5) | 0.0019 |
| Reactive Recruitment | ||||
| Word-of-Mouth | 102 (66.2) | 38 (24.7) | 14 (9.1) | <0.0001 |
| Newspaper | 10 (13.0) | 47 (61.0) | 20 (26.0) | <0.0001 |
| Flyer | 43 (34.4) | 69 (55.2) | 13 (10.4) | <0.0001 |
| Community Partner | 14 (53.8) | 8 (30.8) | 4 (15.4) | 0.0538 |
| Online | 22 (26.5) | 49 (59.0) | 12 (14.5) | <0.0001 |
| Radio & Television | --- | 4 (66.7) | 2 (33.3) | --- |
| Enrollment | ||||
| All | N=98 | N=73 | N=23 | |
| Proactive Recruitment | 9 (75.0) | 2 (16.7) | 1 (8.3) | .0086 |
| Reactive Recruitment | ||||
| Word-of-Mouth | 48 (73.9) | 10 (15.4) | 7 (10.8) | <0.0001 |
| Newspaper | 1 (4.8) | 15 (71.4) | 5 (23.8) | 0.0006 |
| Flyer | 26 (49.1) | 23 (43.4) | 4 (7.6) | 0.0003 |
| Community Partner | 8 (50.0) | 6 (37.5) | 2 (12.5) | 0.1737 |
| Online | 6 (24.0) | 15 (60.0) | 4 (16.0) | 0.0162 |
| Radio & Television | --- | 2 (100.0) | --- | --- |
Pairwise difference between racial/ethnic groups
We included in the analysis only those who were reached out using nontargeted strategies.
Participants who were enrolled by the Hispanic/Latino targeted recruitment strategies were excluded from this analysis (n=15).
Bolded values indicate statistically significant findings at p<.05.
Table 4.
Comparison of recruitment strategies on direct recruitment costs, number of participants recruited, and cost-per-participant (CPP)a (n=254)b
| Recruitment Sources Details | Total Cost | Number of Enrolled & Enrollment rate (%) | CPP |
|---|---|---|---|
| Reactive Recruitment | |||
| Newspaper | $16,551.00 | 85 (23.7) | $194.7 |
| Flyer | $3,000.00 | 63 (17.5) | $47.62 |
| Community Partner | $0 | 17 (4.7) | $0 |
| Online advertisement | $22,759.00 | 86 (24.0) | $264.6 |
| Radio & Television | $14,267.00 | 3 (0.8) | $4,755.6 |
CPP: total direct cost divided by number of enrolled participants
Participants who were recruited using Spanish targeted strategies (n=15) were excluded from this analysis.
3. Results
We screened 849 potential participants; from those, 821 (47.1% women; Mean age [SD]= 48.4 [11.6]) were included in the analysis (Table 1). Overall, 59.3% were eligible and 25.7% were enrolled in the study. Reasons for exclusion were being <18 years old (n = 7), self-identification as “other” race/ethnicity (n = 24), and not meeting other study eligibility criteria (n = 5). Most participants were African American (48.2%), followed by Hispanic/Latinx (41.1%), and White (11.7%).
3.1. Enrollment yield and efficiency by recruitment strategy and race/ethnicity groups.
Overall, compared to proactive recruitment, reactive recruitment yielded the highest number of participants screened (95.6% vs 4.4%), were eligible (96.7% vs 3.3%), and enrolled in the study (94.3% vs 5.7%) (Table 2). Word-of-mouth was the top source for screening (33.6%), eligibility (31.6%), and enrollment (32.5%). The most efficient strategy for screening was word-of-mouth among African Americans (48.9%), flyers among Hispanics/Latinxs (35.9%), and newspaper among Whites (29.2%). The most efficient strategy for eligibility was online advertisement among African Americans (68.8%) and newspaper among Hispanics/Latinxs (84.2%). For enrollment, proactive recruitment among African Americans (39.1%) and radio/television among Hispanics/Latinxs (40.0%) were the most efficient strategies.
3.2. Racial/ethnic differences using non-targeted recruitment.
Compared to Whites and Hispanics/Latinxs, African Americans were significantly more likely to be screened/reached by word-of-mouth (70.3%) and community partners (53.2%), to be eligible if they were recruited by proactive recruitment (75.0%) or word-of-mouth (66.2%), and to be enrolled if they were recruited by proactive recruitment (75.0%; p = .0086), word-of-mouth (73.9%; p <.0001), and flyers (49.1%; p = .0003) (Table 3). Hispanics/Latinxs were significantly more likely than other groups to be reach by online advertisement (57.1%), newspaper (52.3%), and flyers (49.8%) (p < .001 for all), to be eligible if they were recruited by newspaper (61.0%), online advertisement (59.0%), or flyers (55.2%), and to be enrolled via newspaper advertisements (71.4%) and online ads (60.0%) (p < .05 for all).
3.3. Enrollment and CPP.
Ads on public transportation accounted for 17.5% of enrollment and was the least expensive method (total direct cost = $3,000.00; CPP = $47.6) (Table 4). The cost for newspaper advertisements (total direct cost = $16,551.00; CPP = $194.7) and online advertisement (total direct cost = $21,159.00; CPP = $264.6) were higher but yielded a higher enrollment rate (23.4%, 24.0%; respectively). Radio and television ads were the most expensive strategies (total direct cost = $14,267.00; CPP = $4,755.6) and yielded only 0.8% of enrollment.
4. Discussion
The current study reports on lessons learned from recruiting a racially/ethnically diverse sample of adults into a smoking cessation trial. Our recruitment strategies were effective in obtaining our target sample size, in particular among minorities. It was important to implement and constantly evaluate several proactive and reactive recruitment strategies and be flexible in adjusting those strategies to successfully reach our target populations. In addition, targeted and culturally tailored recruitment was necessary to successfully enroll certain groups (e.g., Hispanics/Latinxs). Reactive recruitment strategies yielded higher enrollment compared to proactive strategies. In terms of differences by race/ethnicity, the top source for enrollment was word-of-mouth among African Americans and Whites, and flyers among Hispanics/Latinxs. Our non-targeted recruitment approaches indicated that word-of-mouth, flyers, and newspaper and online ads are more successful among minorities compared to Whites. Word-of-mouth accounted for enrolling a quarter of the total sample and was free. Flyer distribution as the least expensive method among the paid recruitment strategies. These results provide insight on methods for reaching racial/ethnic minority adults who smoke and highlight the value of being flexible and utilizing multiple recruitment strategies to obtain the required numbers of the target populations. Findings also have implications for targeted recruitment to achieve accrual and project goals.
We experienced substantial challenges in achieving our goal of recruiting racial/ethnic groups. We therefore implemented several recruitment strategies and demonstrated flexibility by discontinuing strategies that did not achieve enough return on investment of time and cost and applying new strategies. The adjustment in the strategies was based primarily on attracting specific groups to respond to the various recruitment methods. For example, using generic non-targeted recruitment strategies during the first three years failed to attain our recruitment goal for Hispanics/Latinxs mainly because early in the study only English-only speakers were eligible to participate. To overcome this barrier, we expanded our inclusion criteria to include Hispanics who are Spanish-only speakers and hired Spanish-speaking staff to deliver the intervention in Spanish. We also implemented a targeted recruitment campaign for Spanish-only speakers. This change was critical to achieve our target sample size for this group. Given that 16 million Hispanics/Latinxs in the US reported speaking English less than “very well” or “not at all,” and another 11 million are bilingual but prefer to use Spanish (Krogstad, Stepler, & Lopez, 2015), future smoking cessation trials targeting Hispanics/Latinxs need to consider their language barriers to improve their participation (Lopez-Class, Cubbins, & Loving, 2016).
Consistent with several previous studies, our results indicated that reactive strategies yield greater number of participants relative to proactive strategies, as measured by the proportions who were screened, met eligibility criteria, and enrolled in the study (Brodar et al., 2016; Harris et al., 2003; Webb, Seigers, & Wood, 2009). Reactive strategies have the potential to reach a wider pool of potential participants at the population level. In particular, reactive recruitment usually reaches those who are more motivated to make a serious quit attempt, as their participation requires that they initiate the contact with the research staff (Velicer et al., 1995). When comparing the efficiency of each strategy across racial/ethnic groups, online recruitment among African America and newspaper among Hispanics/Latinxs were resulted in more eligible participants. This might be related to participants socioeconomic status (e.g., higher education, greater income) and characteristics (e.g., younger age). Further investigation of participants characteristics who were reach by these strategies is warranty.
Using non-targeted recruitment strategies, on the other hand, allowed us to examine differences in the efficiency of each strategy by racial/ethnic groups at three levels: screening, meeting eligibility criteria, and enrollment in the study. Consistent with two previous studies (Brodar et al., 2016; Webb et al., 2009), our results indicated that African Americans were more likely than other racial/ethnic groups to be screened, met eligibility criteria, and enrolled via proactive contact and word-of-mouth. Hispanics/Latinxs were more likely than other groups to be screened, met eligibility criteria, and enrolled by newspaper, flyers, and online advertisements (Fitzgibbon et al., 1998). Brodar and colleagues also found that Hispanic/Latinx adults who smoked in California and North Carolina were more likely to be recruited by online recruitment (e.g., Craigslist) than non-Hispanics (Brodar et al., 2016). Another study indicated that Facebook was the most effective methods to recruit Hispanics/Latinxs in smoking cessation trial (Medina-Ramirez et al., 2020). However, evidence suggests that there is no one-size-fits-all approach to recruitment across different racial/ethnic groups, as what works for one group may not work for another (Nicholson et al., 2011).
Among the paid reactive strategies, radio ads were very expensive and yielded very few participants at a very high CPP. A similar result was documented, indicating that using local and national media was very expensive and not effective in recruiting participants (Lopez et al., 2008). One possibility is that mainstream TV or radio advertisements may not be sufficient to reach racial/ethnic minority audiences. In contrast, newspaper and online advertisements yielded a steady rate of accrued participants, at a moderate estimated cost ranging from $264 to $194 per participant, respectively. Several studies have also shown that online advertisements (e.g., Craigslist, Facebook) are effective and relatively inexpensive compared to other recruitment strategies (Brodar et al., 2016; Buller et al., 2012; Ramo et al., 2010). However, advertising on public transportation was a relatively low-cost method that yielded a high enrollment number with low CPP ($47.62). A similar pattern to our findings was reported in which flyers were lower cost than other strategies (Rait et al., 2015). More importantly, word-of-mouth was no cost and accounted for 24% of the total enrolled participants. One potential explanation for the effectiveness of word-of-mouth communication in recruitment is the social aspect of smoking behavior, given that adults who smoke usually know and socialize with others who smoke (Kotz & West, 2009). In addition, one of the most important benefits of word-of-mouth is its credibility because hearing about the study from a friend, family member or co-worker usually carries more weight than other public media advertisements. Future trials in this field could benefit from concerted efforts to enhance word-of-mouth recruitment, as it is a highly economic source of participant recruitment.
The main strength of the current study is its focus on recruiting an equal sample of smokers from three racial/ethnic groups. Other strengths include the assessment of specific recruitment sources and our large sample size, which allowed us to make comparisons by recruitment source and across racial/ethnic groups. There are also a few challenges and potential limitations with this study. First, it is very challenging to assess the real source of enrollment when using the word-of-mouth approach because it may result from a series of other proactive and reactive recruitment efforts. The proactive recruitment approach is also confounded with the fact that many of these individuals had participated in previous smoking cessation trials. Therefore, it may simply be that responsiveness to this approach was influenced by past experiences with study participation (either positive or negative). Second, our screening tool was designed to be brief and was limited to the main eligibility criteria, and thus we were unable to examine participants demographic characteristics that may affect response to recruitment strategies. The staff in-person recruitment cost data was missing. We treat the computed CPP ratios in Table 4 as only a rough index of the cost of each recruitment strategy, not as fully refined indicators of their cost. However, time spent on recruitment was equal across all strategies. In addition, we were unable to compare recruitment methods across Hispanic/Latinx subgroups. We believe that our results may be most applicable to smokers of Cuban and Puerto Rican origin because they comprise the majority of Hispanics/Latinxs in the areas where we targeted our recruitment efforts. Similarly, overall findings from this analysis may not generalize to other geographic areas within the United States. Finally, some of our exclusion criteria such as having serious mental illness and the need for transportation increased minority’s barriers to participate in the trial (Hooper et al., 2019). These criteria were based on the nature of the intervention, a standard (i.e., non-adapted) group CBT which was developed with people without diagnoses of serious mental illness. Moreover, our interventionists were not equipped to manage such conditions, especially in a group intervention context. The intervention was also delivered in-person. Thus, the ability to obtain transportation was required. However, we provided reimbursement to offset the cost of transportation to each session.
5. Conclusions
Our results indicated that multiple recruitment strategies are needed to recruit an adequate sample of racially and ethnically diverse adults who seek help quitting smoking. Flexibility and adjustment in recruitment strategies were essential when planned strategies failed to recruit the desired number of participants. Targeted and culturally adapted recruitment was necessary to reach low acculturated Hispanics/Latinxs. Word-of-mouth, flyers, and newspaper and online ads were more successful among racial/ethnic minorities compared to Whites. Advertising on public transportation was the most economic strategy.
HIGHLIGHTS.
Improving minorities’ recruitment in smoking cessation trials is a high priority.
Multiple recruitment strategies were needed to recruit minorities in the trial.
Proactive recruitment and flyers were successful among African Americans.
Culturally tailored recruitment strategies were crucial to reach Hispanics/Latinxs.
Flyer distribution and newspaper ads were the most economic recruitment methods.
ABBREVIATIONS
- U.S.
United States
- NIH
National Institutes of Health
- CBT
Cognitive Behavioral Therapy
- GHE
General Health Education
- CPP
Cost-per-participant
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 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.
Refences
- Belisario JSM, Bruggeling MN, Gunn LH, Brusamento S, & Car J (2012). Interventions for recruiting smokers into cessation programmes. Cochrane Database of Systematic Reviews(12). [DOI] [PMC free article] [PubMed] [Google Scholar]
- Brodar KE, Hall MG, Butler EN, Parada H, Stein-Seroussi A, Hanley S, & Brewer NT (2016). Recruiting diverse smokers: enrollment yields and cost. International journal of environmental research and public health, 13(12), 1251. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Buller DB, Meenan R, Severson H, Halperin A, Edwards E, & Magnusson B (2012). Comparison of 4 recruiting strategies in a smoking cessation trial. American journal of health behavior, 36(5), 577–588. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cokkinides VE, Halpern MT, Barbeau EM, Ward E, & Thun MJ (2008). Racial and ethnic disparities in smoking-cessation interventions: analysis of the 2005 National Health Interview Survey. American journal of preventive medicine, 34(5), 404–412. [DOI] [PubMed] [Google Scholar]
- Control, C. f. D., & Prevention. (2005). State-specific prevalence of cigarette smoking and quitting among adults--United States, 2004. MMWR. Morbidity and mortality weekly report, 54(44), 1124. [PubMed] [Google Scholar]
- Dickerson DL, Leeman RF, Mazure CM, & O’Malley SS (2009). The inclusion of women and minorities in smoking cessation clinical trials: A systematic review. American Journal on Addictions, 18(1), 21–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fagan P, King G, Lawrence D, Petrucci SA, Robinson RG, Banks D, … Grana R (2004). Eliminating tobacco-related health disparities: directions for future research. American Journal of Public Health, 94(2), 211–217. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fitzgibbon ML, Prewitt TE, Blackman LR, Simon P, Luke A, Keys LC, … Singh V (1998). Quantitative assessment of recruitment efforts for prevention trials in two diverse black populations. Preventive medicine, 27(6), 838–845. [DOI] [PubMed] [Google Scholar]
- Fouad MN, Corbie-Smith G, Curb D, Howard BV, Mouton C, Simon M, … White C (2004). Special populations recruitment for the Women’s Health Initiative: successes and limitations. Controlled Clinical Trials, 25(4), 335–352. [DOI] [PubMed] [Google Scholar]
- Fujimoto WY (1998). Community involvement and minority participation in clinical research. Diabetes Spectrum, 11(3), 161–166. [Google Scholar]
- Haiman CA, Stram DO, Wilkens LR, Pike MC, Kolonel LN, Henderson BE, & Le Marchand L (2006). Ethnic and racial differences in the smoking-related risk of lung cancer. New England Journal of Medicine, 354(4), 333–342. [DOI] [PubMed] [Google Scholar]
- Harris KJ, Ahluwalia JS, Catley D, Okuyemi KS, Mayo MS, & Resnicow K (2003). Successful recruitment of minorities into clinical trials: The Kick It at Swope project. Nicotine & Tobacco Research, 5(4), 575–584. [DOI] [PubMed] [Google Scholar]
- Health, U. D. o., & Services, H. (1998). Tobacco use among US racial/ethnic minority groups. A report of the Surgeon General. [Google Scholar]
- Hooper MW, Asfar T, Unrod M, Dorsey A, Correa JB, Brandon KO, … Lee DJ (2019). Reasons for exclusion from a smoking cessation trial: An analysis by race/ethnicity. Ethnicity & disease, 29(1), 23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hooper MW, Lee DJ, Simmons VN, Brandon KO, Antoni MH, Unrod M, … Brandon TH (2018). Reducing racial/ethnic tobacco cessation disparities via cognitive behavioral therapy: Design of a dualsite randomized controlled trial. Contemporary clinical trials, 68, 127–132. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hussain‐Gambles M, Atkin K, & Leese B (2004). Why ethnic minority groups are under‐represented in clinical trials: a review of the literature. Health & social care in the community, 12(5), 382–388. [DOI] [PubMed] [Google Scholar]
- Jemal A, Thun MJ, Ries LA, Howe HL, Weir HK, Center MM, … Anderson R (2008). Annual report to the nation on the status of cancer, 1975–2005, featuring trends in lung cancer, tobacco use, and tobacco control. JNCI: Journal of the National Cancer Institute, 100(23), 1672–1694. [DOI] [PMC free article] [PubMed] [Google Scholar]
- King AC, Cao D, Southard CC, & Matthews A (2011). Racial differences in eligibility and enrollment in a smoking cessation clinical trial. Health Psychology, 30(1), 40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kotz D, & West R (2009). Explaining the social gradient in smoking cessation: it’s not in the trying, but in the succeeding. Tob Control, 18(1), 43–46. [DOI] [PubMed] [Google Scholar]
- Kredo T, Bateganya M, Pienaar E, & Adeniyi F (2012). Cochrane database of systematic reviews. In: John Wiley & Sons, Ltd. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Krogstad JM, Stepler R, & Lopez MH (2015). English proficiency on the rise among Latinos.
- Lopez-Class M, Cubbins L, & Loving AM (2016). Considerations of methodological approaches in the recruitment and retention of immigrant participants. Journal of racial and ethnic health disparities, 3(2), 267–280. [DOI] [PubMed] [Google Scholar]
- Lopez EN, Simmons VN, Quinn GP, Meade CD, Chirikos TN, & Brandon TH (2008). Clinical trials and tribulations: lessons learned from recruiting pregnant ex-smokers for relapse prevention. Nicotine & Tobacco Research, 10(1), 87–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Medina-Ramirez P, Calixte-Civil P, Meltzer LR, Brandon KO, Martinez U, Sutton SK, … Simmons VN (2020). Comparing methods of recruiting Spanish-preferring smokers in the United States: Findings from a randomized controlled trial. Journal of medical Internet research, 22(8), e19389. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nicholson LM, Schwirian PM, Klein EG, Skybo T, Murray-Johnson L, Eneli I, … Groner JA (2011). Recruitment and retention strategies in longitudinal clinical studies with low-income populations. Contemporary clinical trials, 32(3), 353–362. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rait MA, Prochaska JJ, & Rubinstein ML (2015). Recruitment of adolescents for a smoking study: use of traditional strategies and social media. Translational behavioral medicine, 5(3), 254–259. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ramo DE, Hall SM, & Prochaska JJ (2010). Reaching young adult smokers through the internet: comparison of three recruitment mechanisms. Nicotine & Tobacco Research, 12(7), 768–775. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sherer R (2002). Surgeon General’s report highlights mental health problems among minorities. Psychiatric Times, 19(3). [Google Scholar]
- The National Institutes of Health (NIH). (1993). National Institutes of Health Revitalization Act of 1993. Retrieved from https://orwh.od.nih.gov/sites/orwh/files/docs/NIHRevitalization-Act-1993.pdf
- Trinidad DR, Pérez-Stable EJ, White MM, Emery SL, & Messer K (2011). A nationwide analysis of US racial/ethnic disparities in smoking behaviors, smoking cessation, and cessation-related factors. American Journal of Public Health, 101(4), 699–706. [DOI] [PMC free article] [PubMed] [Google Scholar]
- USDHHS, CDC, NCCDPHP, & Office on Smoking and Health. (2014). The Health Consequences of Smoking—50 Years of Progress: A Report of the Surgeon General. Retrieved from
- Velicer WF, Fava JL, Prochaska JO, Abrams DB, Emmons KM, & Pierce JP (1995). Distribution of smokers by stage in three representative samples. Prev Med, 24(4), 401–411. doi:S0091–7435(85)71065–1 [pii] 10.1006/pmed.1995.1065 [DOI] [PubMed] [Google Scholar]
- Webb MS, Seigers D, & Wood EA (2009). Recruiting African American smokers into intervention research: relationships between recruitment strategies and participant characteristics. Research in nursing & health, 32(1), 86–95. [DOI] [PubMed] [Google Scholar]
- Yancey AK, Ortega AN, & Kumanyika SK (2006). Effective recruitment and retention of minority research participants. Annu. Rev. Public Health, 27, 1–28. [DOI] [PubMed] [Google Scholar]