users in the youngest age group was small, as were the number of cycles they tracked compared to older teens and adult users. Because initially, the age at menarche was not a data point, the exact gynecologic age could not be determined, and data were analyzed by age. The mean cycle lengths for girls 9-13 years old ranged from 33-36, while the cycle lengths of girls 15 and older–32 daysdwere similar to those up to age 28. Data from Clue were compared with the historical data from Treloar 1967. The 95%-ile for cycle lengths for each age group tracked with Clue was longer than the cycle lengths previously reported. Conclusions: This is the first report to present menstrual cycle data and statistical analyses from a very large, geographically broad population in this age group. Similar to blood pressure, pulse, and respiration, the menstrual cycle has been described as a vital sign of health in adolescents. Disruptions in menstrual cyclicity can point to specific medical conditions in adolescents and adults, and thus knowledge of menstrual cycle data for adolescents is of importance. Data sets that are orders of magnitude larger than were previously available will add to our knowledge of menstruation in adolescents and young adults. 5. Retention of Adolescent Participants in Research Studies Frank M. Biro MD * , Anita R. Southwick MS, Christine Mosbaugh MS, Susan M. Pinney PhD Purpose: Longitudinal studies involving youth must identify and apply retention strategies that include both youth and their guardians. We examined which retention strategies utilized in our study were preferred by adolescent participants, as well as their adult guardians. Methods: Female participants, recruited at ages 6-7, were followed up to 11 years in an observational study. The participants were surveyed about retention strategies once between the ages of 10-15 years, and their guardians were surveyed one year later. An open-ended question to par- ticipants about continued participation, and Likert-type scales for aware- ness and preference of retention strategies to participants and to their guardians (the following year), were compared by age, race, and socio- economic (SES) categories. Responses to the open-ended question “You are in the study because.” were analyzed through thematic analysis and code development. Results: 230 participants and 212 guardians completed the surveys. The three most frequent themes cited by participants regarding moti- vation to remain in the study were advancing medicine, external motivation (such as encouragement from family members), and altruism. Of the 12 retention strategies used in the study, over 90% of participants reported receiving cash, receiving an annual gift, and fun at the visit as the most effective methods. Nearly all guardians endorsed advancing science as an important reason for continued participation, and greater than 90% endorsed fun at the visit; a study newsletter; and cash, gift card, and small thank you gift. Many of these preferences entailed minimal financial support. There were subtle but statistically significant differences in preferences by race and SES. For example, guardians with lower SES were more likely to endorse receiving a holiday card, participant birthday card, study T-shirt, and the girl’s so- cial gala (all p < 0.01). Conclusions: We examined awareness and preferences of retention strategies among adolescent participants and their guardians in a decade- long longitudinal study. Advancing medical science and altruism were two of the top three motivations to remain in this study; given the high costs of recruitment, investigators should examine cost-effective approaches to adolescent participant retention in their study population. Helping par- ticipants understand how meaningful their participation can be, and creating a culture of fun at study visits cost little, and can enhance study retention. 6. Adapting Evidence Based Pregnancy Prevention Programs to System- Involved Adolescents Ola Mscichowski MD 1 , Allison Muzzey BA 1 , Doug Cope-Barnes LCSW 2 , Abby Hunt MSW 2 , Mary A. Ott MD, MA *1 1 Indiana University School of Medicine Department of Pediatrics, Section of Adolescent Medicine 2 Health Care Education and Training, Inc. (HCET) Research Conducted: Indiana University School of Medicine, Depart- ment of Pediatrics, Section of Adolescent Medicine and Health Care Edu- cation and Training, Inc. (HCET), Indianapolis, IN. Purpose: Adolescents in juvenile corrections and foster care have high rates of pregnancy, and are thus important targets for evidence-based programs (EBP). Few EBPs exist for these system-involved adolescents. As part of a program evaluation and adaptation process for an EBP for system involved adolescents, we examine the influence of gender on adolescents’ perspectives on pregnancy prevention needs. Methods: Qualitative data were collected from 399 participants, age 12-20, from Indiana juvenile corrections and foster care programs that participated in a pregnancy prevention EBP, Wyman’s Teen Outreach ProgramÒ (TOPÒ). TOPÒ is a youth development EBP developed in schools, focusing on primary prevention of adolescent pregnancy. Our EBP was adapted to occur in weekly sessions over 3-4 months. We included information from 11 female-only cohorts (n¼136), 17 male- only cohorts (n¼228) and 3 mixed-gender cohorts (n¼35). At the end of each session, facilitators recorded participant responses on stan- dardized reflections forms about open-ended questions about their perspectives to the EBP, including suggestions for improvement on future sessions. The recorded responses were often summaries of discussions, but also included individual responses and direct quotes. Textual data were analyzed using qualitative content analysis. Figure 1. Clue cycle length for the age range 9-24 years. Mean and median cycle lengths are shown, as well as 5% and 95% range. Figure 2. Comparison of menstrual cycle length data between Treloar et al. and Clue database. The horizontal lines correspond to 5% -95% range, dots mark the median cycle length for each age. For Treloar et al, the gynecologic age is shown; for Clue data, the chronologic age. Oral Abstracts / J Pediatr Adolesc Gynecol 30 (2017) 268e274 270