Journal Article

Yields and costs of recruitment methods with participant phenotypic characteristics for a diabetes prevention research study in an underrepresented pediatric population

Keller, Colleen S., Konopken, Yolando P., Pimentel, Janiel, Shaibi, Gabriel Q., Olson, Micah L., Peña, Armando, Fernández, Arlene D. R., Williams, Allison N., Patrick, Donald L., Soltero, Erica G., Hu, Houchun H., Ayers, Stephanie L., Vander Wyst, Kiley B., Castro, Felipe G., Hooker, Elva, Knowler, William C.

2020

Overview

Background/aims Prediabetes and diabetes disproportionately impact Latino youth, yet few diabetes prevention programs have prioritized inclusion of this underrepresented population. This report describes the recruitment process, yields, associated costs, and phenotypic characteristics of Latino youth with obesity and prediabetes enrolled in a randomized controlled diabetes prevention study in the USA. Methods Recruitment efforts included referrals from clinics, community outlets, local media, and word of mouth with the goal of enrolling 120 Latino adolescents aged 12–16 with obesity (BMI >  95th percentile) and prediabetes. Prediabetes eligibility was determined by any of the following: HbA1c between 5.7 and 6.5%, fasting glucose between 100 and 125 mg/dL, or a 2-h glucose between 120 and 199 mg/dL following a 75-g oral glucose tolerance test (OGTT), but not meeting any of the diagnostic criteria for diabetes. Eligible participants were randomized 2:1 to either a 6-month community-based lifestyle intervention that included group nutrition and health education classes (1 day/week) and group exercise classes (2 days/week) or usual care control arm. Recruitment yields were determined by review of referral source in the study screening database. Recruitment costs were determined by an after-the-fact financial review of actual and in-kind costs. Participant phenotypic characteristics (i.e., demographics, anthropometrics, and biochemical data) were compared by recruitment strategy using a one-way ANOVA. Results Recruitment efforts covered 160 mile 2 (414 km 2 ) across 26 ZIP codes (postcode) in the Phoenix Metropolitan Area and yielded 655 referrals from clinics ( n  = 344), community ( n  = 143), media ( n  = 137), and word-of-mouth ( n  = 31). From this pool, 26% ( n  = 167) did not meet general, pre-screening eligibility criteria; 29% ( n  = 187) declined participation; and 10% ( n  = 64) were unable to be contacted. A total of 237 youth were invited to the clinical research unit to determine final eligibility. Following the OGTT, 52% ( n  = 122) met prediabetes criteria and 117 were subsequently randomized. Clinical recruitment yielded the highest number of referrals (53%; n  = 344) while word-of-mouth yielded the highest proportion (35%; n  = 11) of randomized participants per referred youth. There were no significant differences in anthropometric or biochemical measures among youth by recruitment strategy. Based upon final enrollment numbers, community recruitment was the costliest approach ($486/randomized participant) followed by clinical ($248/randomized participant) and media ($236/randomized participant). Conclusions The ability to meet enrollment goals for a clinical trial of an underrepresented population required multiple recruitment strategies. Although strategies vary in yields and costs, it appears they produce similar phenotypical risk profiles of eligible youth. Trial registration ClinicalTrials.gov NCT02615353 . Registered on 26 November 2015

• Publisher: BioMed Central,BioMed Central Ltd,Springer Nature B.V,BMC
• Subject: Biomedicine / Care and treatment / Children & youth / Clinical trials / Community / Community health services / Demographic aspects / Diabetes / Diabetes mellitus / Dietitians / Enrollments / Glucose / Health aspects / Health education / Health Sciences / Identification and classification / Intervention / Laboratories / Lifestyles / Management / Medical referrals / Medicine / Medicine & Public Health / Methodology / Pediatric research / Pediatrics / Phenotypes / Recruiting / Social networks / Statistics for Life Sciences