Therese S. Richmond PhD, CRNP, FAAN 1 Rose Cheney PhD 2 Liana Soyfer BA 3 Rebecca Kimmel BA 1
description
Transcript of Therese S. Richmond PhD, CRNP, FAAN 1 Rose Cheney PhD 2 Liana Soyfer BA 3 Rebecca Kimmel BA 1
Therese S. Richmond PhD, CRNP, FAAN1
Rose Cheney PhD2
Liana Soyfer BA3
Rebecca Kimmel BA1
Adrian Raine D.Phil2,3
EMPIRICALLY-BASED RECOMMENDATIONS FOR
RECRUITMENT OF COMMUNITY-RESIDING YOUTH INTO
RANDOMIZED CLINICAL TRIALS
1School of Nursing – Biobehavioral Research Center; 2Perelman School of Medicine; 3School of Arts & Sciences/ Criminology
ACKNOWLEDGMENTS
Funder: PA DOH 2008 Health Research Nonformula Grant Award*
*This project is funded, in part, under a grant with the Pennsylvania Department of Health. The Department specifically disclaims responsibility for any analyses, interpretations or conclusions
BACKGROUND• Violence is a priority public health problem1
• Studies to reduce violence require enrollment of community-residing individuals• Challenging
• Data-driven recommendations to enhance recruitment are sparse
• Inadequate sample sizes• Generalizability
• Timely infusion of findings into practice
1 Krug EG et al. (2002). World report on violence and health. Geneva: WHO.
PURPOSE• To analyze factors & strategies that affect
successful recruitment of community-residing youth into RCTs requiring follow-up
• Focus – 3 factors that may affect recruitment: • access to study information - recruitment sources,
referrals, gatekeepers
• minority status - race
• personal costs to participate - time, transportation
HEALTHY BRAINS & BEHAVIOR
Public Health Translation:
Cheney, Fein, Richmond
Public Health Translation:
Cheney, Fein, Richmond
Intervention:
Richmond, Cheney
Biosocial Prediction:Raine, Gur
Follow-up
Recruitment
4. Develop a new biosocial approach for treatment of aggression
1. Assess risk factors in order to predict early adolescent violence
2. Identify factors that protect children socially at risk
5. Test differential prediction and treatment of 2 variants of aggression
Genetic Mouse Model
Brodkin
Genetic Mouse Model
Brodkin
3. Assess biosocial risk factors and efficacy of nutritional intervention in a mouse model
Subjects (mice)
APPROACH• Goal - 500 youth from Phila & contiguous
zipcodes• 240 square miles, population 1.9 million with 70,000
11-12 year olds
• Sampling Frame • 11-12 year old boys & girls
• Behavior – ranging from developmentally appropriate to highly aggressive
• Plan – enroll from the Phila School District
• Reality – Award budget cut by 20%; School District refused access
RECRUITMENT STRATEGIES• Targeted mailings
• Study flyers• Cooperating organizations
• In study newsletter
• Posted & sent home
• News/internet/public transportation ads
• Referral incentive for current subjects
• Incentives, expense reimbursement, bonuses
DATA SOURCES
• Recruitment log - ads (date, time, cost)
• 2010 census data by zipcode
• Travel - Google maps (time/distance)• Centroid of each zip code
• Incentives
• Recruitment log - contacts & enrollment
ANALYSIS• Mail saturation rates (# of mailings/estimated
population of 11-12 year olds in zip code)
• Contact # (# of initial contacts to study)
• Conversion rate (consented/contacts)
• Personal costs1 (travel & opportunity)
• Recruitment rate (#consents/11-12y in zip)
• Recruitment efficiency (success & cost/source)
• Modeled using linear regressions
1Stringer M et al. (2005). The cost of prenatal care attendance and pregnancy outcomes in low –income working women. Journal of Gynecologic & Neonatal Nursing, 34, 551-560.
RESULTS
• 1038 contacts yielded 354 (37%) consented subjects • 185 boys/169 girls; 84% Black
• Recruitment rate• 1.5/week in first 6 months
• Increased to 3.6/week, 9/week, 13.5/week over time as strategies synergistically overlapped & incentives increased
RECRUITMENT COSTSSource Enrolled n (%)* Est. Cost (staff effort) Cost/Enrolled
Targeted mailings 73 (33%) $4,000 (low-med) $55
Community flyer 49 (22%) $360 (high) $7
Metro news ads 26 (12%) $2,200 (low) $85
Personal referral** 25 (11%) $625 (low) $25
Craigslist 18 (8%) $0 (low) $0
Healthcare providers 11 (5%) $0 (medium) $0
School targeted mailings 7 (3%) $340 (high) $49
Mass transit ads 5 (2%) $830 (low) $166
Other news ads 3 (1%) $410 (low) $137
Other 17 (8%) - -
*n = 234 (of 354 participants) who provided information on referral source**Participants received $25 for personal referrals into the study
TRAVEL TIME & RECRUITMENT RATE Region Est. Car Travel
Time (min)Est. Public Transit Travel Time (min)
# Youth Recruited
Recruitment Rate per 1,000
West Phila 8.6 21.4 143 14.5
Center City 10.1 17.0 0 0
South Phila 12.6 37.2 29 6.12
Lower North Phila 13.9 36.8 43 7.05
Upper North Phila 18.2 45.5 41 2.78
Northwest Phila 19.7 47.2 15 2.51
Northeast Phila 26.2 64.3 14 1.09
Delaware Cty* 16.0 37.4 43 6.83
Montgomery Cty* 26.3 60.0 10 1.58
Bucks Cty* 32.2 83.2 1 0.35
*county zip codes contiguous to Phila County
CONTRIBUTORS TO RECRUITMENT RATE
Variable B Beta SigTargeted mailing 0.015 0.419 0.001Travel time by public transit -0.087 -0.299 0.002% Black residing in zip code 3.516 0.200 0.048
N = 69 zip codes; R2 = 0.562Non-significant to model - average household income & travel time by car
CONCLUSION• Recruitment of community-based youth into
complex studies is challenging but can be successful with • careful planning of recruitment strategies
• consideration of the personal costs of study participation
• alignment of study incentives with personal costs
• Targeted mailings proved to be the most efficient strategy in successful recruitment