“ Don't Close the Door on Them ” : Recruiting and retaining vulnerable Black adolescents in prevention research

There is a need to understand why some strategies work better than others for engaging vulnerable Black adolescents in prevention research, especially those who are affected by parental drug use. The current study sought to fill this gap. Thirty in ‐ depth interviews were conducted with parents who reported a history of illicit drug use ( N = 11), young adults who had a biological parent with a history of illicit drug use ( N = 14), and service providers who work with families affected by illicit drug use ( N = 5). Nearly all participants were Black and lived or worked in Baltimore, Maryland. Over half of the participants were female (62%). Interviews were recorded and ranged from 30 to 92 minutes in length. Data were analyzed using an inductive, content analysis approach. Four themes emerged (coined STAR): SAFE people and places minimize retraumatization; TEAMING UP with community partners increases acceptability; ADDRESSING a range of needs helps adolescents survive; and RELATABLE facilitators understand and listen to adolescents. Findings build onto and extend previous research that highlights strategies to life issues, and not only prevention behaviors, to increase the usability and relevance of interventions. Our findings the importance of previous trauma ‐ informed, prevention research that integrates an understanding of adverse childhood experiences, emphasizes the need for emotional safety, and creates opportunities for adolescents to rebuild a sense of control as needed (Earls, 2018; Oral et al., 2016). Among parents who use drugs, there are higher rates of concurrent mental health problems, trauma history, and PTSD, financial instability, interpersonal violence, incarceration, and stress (Stulac et al., 2019). These factors can increase the likelihood of trauma exposure among children affected by parental drug use. Participants highlighted the need to make both physical and emotional safety a priority in recruitment and retention efforts for this group because adolescents likely have a history of trauma and all measures should be taken to minimize retraumatization. This recommendation aligns with the Substance Abuse and Mental Health Services Administration's guidance for a trauma ‐ informed approach, whereby physical and psychological safety are recognized as one of the six key prin-ciples of a trauma ‐ informed approach (Substance Abuse and Mental Health Services Administration,


| INTRODUCTION
Nearly nine million children in the USA have a parent suffering from a substance use disorder (American Academy of Pediatrics, 2018). Having a parent who uses illicit drugs (e.g., heroin and cocaine) increases risk for early substance use (alcohol or drug use before the age of 13) among adolescents ages 10 to 19 (Adger & Belcher, 2008;Biederman et al., 2000;Godleski & Leonard, 2019;Howell et al., 2019;O'Loughlin et al., 2019). Preventing early substance use among adolescents affected by parental drug use, is a top public health priority, given that early substance use is associated with low educational attainment, incarceration, and substance use disorder later in life (Adger & Belcher, 2008;Biederman et al., 2000;Godleski & Leonard, 2019;Howell et al., 2019;O'Loughlin et al., 2019). Ensuring that adolescents affected by parental drug use are engaged in substance use prevention programs is one way to prevent early substance use.
Recruiting and retaining adolescents affected by parental drug use into substance use prevention programs can be challenging for at least five reasons. First, logistical issues such as time requirements, misaligned incentives, transportation issues, and obtaining parental consent may be barriers that vulnerable adolescents are unwilling or unable to overcome to participate in an intervention (Seibold-Simpson & Morrison-Beedy, 2010). Second, a few substance use prevention programs acknowledge the lived experiences of children who have parents with substance use problems (Botvin & Griffin, 2014;Cuijpers, 2002;Foxcroft & Tsertsvadze, 2012;Wynn et al., 2000). This oversight may limit the extent to which this population finds existing substance use prevention programs useful and relevant. Third, adolescents affected by parental drug use may be reluctant and feel unsafe discussing problems with others (Adger & Leff, 2002;Fenster, 2011). Fourth, housing instability, including multiple moves and experiences of disconnected phones, is common among adolescents affected by parental drug use and may limit regular communication (Callaghan et al., 2011;Morrison-Beedy et al., 2016;Powell et al., 2018). Finally, nonparticipation or premature withdrawal among vulnerable adolescents may reflect a necessary coping strategy that allows them to prevent reminders of their trauma or retraumatization (Campbell et al., 2014;Vogel et al., 2020).
This suggests there is a need for trauma-informed strategies to engage this vulnerable population in interventions that both understand and meet their needs.
Successfully recruiting and retaining participants requires a set of tailored and consistently employed strategies (Abshire et al., 2017;Hanna et al., 2014). Accordingly, researchers have proposed a number of strategies to enhance the recruitment and retention of vulnerable adolescents in interventions. For example, partnering with trusted community partners and collecting multiple pieces of contact information enables researchers to have more frequent and consistent contact with vulnerable Black adolescents (Grape et al., 2018;Mendelson et al., 2020;Morrison-Beedy et al., 2016).
Having this information allows the research team to contact participants frequently to both inform participants of the study and inquire about their well-being. Social media information has also become a popular method to recruit a wider range of participants and also assist with retention since social media profiles are often more stable than cell phone numbers and home addresses (Grové, 2019;Gu et al., 2016;Jones et al., 2012;Van Doesum et al., 2016;Whitaker et al., 2017). Finally, when researchers allocate funds to hire and train staff to develop and consistently use personable recruitment and retention strategies, they are often more successful at keeping vulnerable adolescents engaged POWELL ET AL. | 995 (Catherine et al., 2020;Seibold-Simpson & Morrison-Beedy, 2010). Research studies often benefit from using a combination of the methods above to engage vulnerable adolescents.
Despite best efforts, few have investigated why some recruitment and retention strategies work better than others, especially among vulnerable Black adolescents. The negative consequences of parental drug use are exacerbated among Black adolescents as evident by their disproportionate representation in the criminal justice system and overrepresentation in the foster care system (Welty et al., 2016;Zapolski et al., 2014). Although concern has increased for engaging Black adolescents affected by parental drug use in prevention research (Yaster et al., 2020), relatively little is known about the perspectives that adolescents and their families may have about participating. This is a missed opportunity given that adolescents affected by parental drug use may benefit most from prevention research; however, they remain least engaged (Stulac et al., 2019). In the current study, we used qualitative methods to describe why some strategies may be more effective in overcoming barriers to recruitment and retention efforts with Black adolescents affected by parental drug use in prevention research.

| METHODS
Data analyzed for this study were drawn from the Better Together Project. Better Together is a multiphase project that takes a community approach to prevent early substance use and injury among Black adolescents in Baltimore, Maryland, a city that has had a longstanding problem with illicit drugs use. Baltimore had the highest overdose fatality rate of any city in the United States in 2017, with nearly 700 opioid-related deaths (Baltimore City Department of Health, 2018). In Baltimore, more than one in five teens reported experiences of parental drug use; 70% of these teens identified as Black (Centers for Disease Control and Prevention, 2020). Thirty-eight percent of Black adolescents affected by parental drug use reported early substance use compared to only 28% of White teens (Centers for Disease Control and Prevention, 2020).

| Recruitment & participants
The study team partnered with local organizations to recruit participants. Four recruitment strategies were used: (1) existing networks of the study team, (2) referrals from partner organizations, (3) participant referrals, and (4) flyers and information sheets posted in partner organizations. All data were collected between November 2018 and July 2019. Research protocols were approved by the Institutional Review Board at Johns Hopkins Bloomberg School of Public Health.
A maximum variation purposive sampling strategy was used to provide a diverse range of cases relevant to recruiting and retaining Black adolescents affected by parental drug use in prevention research (Patton, 1990).
Specifically, three groups were included in this study (N = 30): young adults affected by parental drug use (n = 14), parents with a history of illicit drug use (n = 11), and providers who work with families affected by parental illicit drug use (n = 5); we refer to these samples as "young adults", "parents" and "providers", respectively. The final number of study participants was determined by the richness and saturation of the data (Malterud et al., 2016). For the purposes of this study, a parent with a history of illicit drug use was defined as a biological parent who abused an illicit drug for longer than 6 months; parents who were currently in treatment or recovery were included.
Parents and young adults were not related.
A screening tool was used to determine eligibility. Inclusion criteria for the young adult sample were as follows: be between 18 and 24 years old, identify as Black/African American, and have a biological parent with a history of illicit drug use. There are considerable ethical concerns related to interviewing children and adolescents exposed to trauma experiences (Carter, 2009 (Pinto et al., 2014), young adults, rather than adolescents, were included in this study.
Furthermore, adolescents who are still living these experiences may have been less well equipped to provide information about successful strategies for long-term engagement in prevention research. Inclusion criteria for the parent sample were (1) at least 25 years of age, (2) be a parent with a history of any illicit drug use, and (3) speak and understand English. Providers also had to be at least 25, speak and understand English and work with families affected by parental drug use. Ninety-six percent of participants were African American. Over half of the participants (60%) were women. Over half of participants completed high school or received a GED (60%). Table 1 lists detailed demographic information of participants.

| Demographic questionnaire
Participants completed a brief demographic questionnaire. The demographic questionnaire asked young adult participants' age, sex, race, Hispanic ethnicity, and highest grade in school. The demographic questionnaire asked parent and provider participants' age, race, sex, highest education level, and their social interactions with adolescents (from every day/almost every day to never). The brief demographic questionnaire was completed on a tablet provided by the project.

| Interview guide
Three semistructured interview guides were used: one for young adults, one for parents, and one for providers. The interview guides had 16 primary questions that were categorized into four sections (i.e., personal history, family relationships, health, and program recommendations). Sample items include, "What topics should be included in an intervention designed to prevent substance use among adolescents affected by parental drug use?", "How would we find families and keep them engaged?" and "Are there other things we need to consider as we develop this intervention?" All questions asked participants to focus on Black adolescents (ages 10-19) as they answered questions about engagement in prevention research. Follow-up probes were used to ensure the completeness of participant responses.

| Procedures
In-depth interviews were scheduled at a time, date, and location convenient for participants. Round-trip transportation was offered to each participant. Each participant was provided written, informed consent before participation. After a review of the consent form, the interviewer answered participant questions about the project. All interviews were conducted by a trained member of the study team. Participants were assigned a unique study identification number to protect their privacy.
All interviews were digitally recorded and ranged from 32 to 92 min in duration. Upon completion of the demographic survey and interview, participants were thanked for their participation and given information about the next steps of the broader research project. They were also given $25 for their time.

| Data analysis
The digital recordings of the interviews were transcribed verbatim using an online transcription service. Study team members compared each of the digital recordings to the transcripts to verify the accuracy of the transcriptions. Transcripts were edited as needed by a study team member and then imported into the qualitative software program, Atlas.ti (version 8.0), to assist with data management and analysis. Transcripts were analyzed inductively using a content analytic approach (Hsieh & Shannon, 2005). Using this approach, the study team created a coding manual to identify the basic themes in the data. Initially, nine in-depth interview transcripts (i.e., four young adults, three parents, and two providers) were independently read and coded by the study team. The codes were then discussed and refined as a group. There were 58 codes, categorized into seven groups, in the final codebook. This codebook was applied to the full data set.
Each transcript was coded by two study team members. To ensure consistent coding, team members met after coding each interview to discuss the coded material and address any discrepancies. Discrepancies were discussed and resolved by consensus. Once all interviews were coded, comparisons were made across interviews to allow themes to emerge.
Ultimately, the study team developed themes based on patterns and topics that persisted throughout the interviews.
Three techniques were used to enhance the trustworthiness of findings: peer debriefing, triangulation, and negative case analysis. For example, the study team met weekly to discuss major and minor points both during data collection and analysis. The themes presented had to be supported by at least two of the three stakeholder groups. Once themes emerged, study team members also searched for and discussed text that did not support the salient patterns. Compelling contradictory evidence was used to rethink and remove themes. Codes describing research recommendations were used in the current analyses. Demographic data were analyzed using STATA (version 16).

| RESULTS
Participants acknowledged the challenges associated with recruiting and retaining adolescents affected by parental drug use into prevention research. These challenges included skepticism of the research purpose and person/ organization leading the intervention, the stigma associated with participation among adolescents affected by parental drug use, inability to obtain consent from parents, and fear of participating alone. Given these challenges, participants explained why some strategies may be more effective than others. Four themes emerged, which we coined the STAR features of recruiting and retaining vulnerable Black adolescents in prevention research: SAFE people and places minimize retraumatization; TEAMING UP with community partners increases acceptability; ADDRESSING a range of needs helps adolescents survive, and RELATABLE facilitators understand and listen to adolescents. The themes are discussed below in detail and summarized in Table 2.

| SAFE people and places minimize retraumatization
Having a parent who uses drugs was one of many adverse childhood experiences described by participants. Other trauma exposures included community violence, excessive death of significant others (i.e., experiencing two or more deaths a year for successive years), parental incarceration, neglect, and abuse (physical and sexual). Adolescents learned early that they should not talk about their families and their parents' struggles with drug use. In times when adolescents did talk with others, the results were often not favorable (e.g., anger from parents, involvement of social services). These experiences led to participants being and feeling unsafe. Safety, described as the chance of harm, injury, or loss, was discussed both as an important physical and emotional concept. Participants noted that adolescents affected by parental drug use often have little control over how they are treated, and few places to go or trusted adults to consult when trouble arose. The lack of physical and emotional safety left these adolescents less willing to believe that others could and would help when needed. As a result, adolescents were reluctant to participate in prevention research for fear of what may happen if they shared details about their lives.
As noted by two participants,

T A B L E 2 Summary of STAR themes
Theme

Key points
Safe people and places minimize retraumatization -Youth affected by parental drug abuse may have experienced multiple and complex traumas -Providing physical and emotional safety allows youth to relax and fosters supportive interactions -Long-term interventions may be best suited to offer safety and minimize retraumatization Teaming up with community partners increases acceptability -Endorsement from community partners implies that research reputable -Key stakeholder in communities include adult partners in local organizations, formal institutions, and other adolescents Addressing a range of needs helps adolescents survive -Substance use prevention may be a low priority for youth who have more pressing issues such as financial instability and navigating negative community contexts -Incentives may include food, money, and other essential items -Connecting intervention content with community impact demonstrates holistic investment in youth Relatable facilitators understand and listen to adolescents -Relatable facilitators have genuine intentions and are willing to be vulnerable -Facilitators of the same race and with lived experiences are preferred and seen as relatable -Relatable facilitators are as important as intervention content in retaining youth affected by parental drug use in prevention research POWELL ET AL.

| 999
He said his uncle came into the house drunk. He's traumatized in a sense because his uncle is yelling.
Then everybody else starts yelling. Then he's trying to find a safe haven to not have to deal with that, and because he can't, he gets upset and he has nowhere to go. -Provider (man, age 47) You can't sit there and go to somebody and be like, "My mother did drugs, and I don't eat," and then CPS shows up at the house on the same day. You the only child in the house, and now your mother knew you said something, and once they leave, you're being beat on. Young Adult (woman, age 19) Recognizing this reality, participants emphasized the need to ensure that people and places were both physically AND emotionally safe for adolescents. Specifically, providing a physically safe place for adolescents to convene and that also fosters supportive, authentic interactions were essential components for recruiting and retaining these young people in prevention research. Safe people and places allowed adolescents to relax and minimized retraumatization. Below, three participants describe features of physically and emotionally safe spaces.
Have a group with the same age group that might be going through trauma and have them sit down and talk about their similarities because at the end of the day you're building a place for them to bond and feel like they're safe. Recreation centers and churches were consistently described as safe places for interventions to be delivered because they were familiar, and the people were inviting. However, recent closures of recreation centers and limited hours of local churches made these locations less accessible to adolescents. Thus, there is a need to identify additional locations within communities to deliver interventions that meet both the physical and emotional safety needs of this population.
To ensure physical and emotional safety among adolescents, all participants championed long-term interventions. They believed that short-term interventions (i.e., less than one month) displayed a temporary interest in adolescents and focused narrowly on individual behaviors instead of promoting lifelong success. Participants suggested interventions range no fewer than five weeks and could run for over a year, with additional contacts after the intervention is complete. Most also suggested having sessions for 90 min once or twice a week. Young adults described participating in short-term programs where they were beginning to build relationships but ended abruptly at the close of the intervention. For example, Not short term, that's definitely not…You're going to need a long time for where they can come in, and it should be if they leave, whatever, don't close the door on them. Always let them know, "You can come back, it's not over," because they might not be ready. "We here for you," Young adult (woman, age 20) You got to be consistent with kids, man…. You have people start up these programs in the community, and the kids feel comfortable talking to them, and then they're gone. It's gone. If you want to start something, follow through…. And when you're doing it [leaving], you're doing nothing but reinforcing the neglect that they already have. -Provider (woman, age 34) Participants thought longer interventions would give youth a better chance to see how their actions, large and small, affected their communities. They also believed that the longer duration would provide consistent care and concern that adolescents affected by parental drug use may lack at home.

| Teaming up with the community increases acceptability
Participants highlighted the need for researchers to build and cultivate relationships with a broad network of partners to support prevention efforts for adolescents affected by parental drug use. This relational, networkbased approach was described as a way to increase the acceptability of the intervention from adolescents, families, and other important adults. Increased acceptability would yield more engagement among adolescents. Furthermore, acceptability across stakeholders, and not only adolescents, was deemed to be an effective strategy for recruitment and retention because it implied the intervention was reputable. The "community" was described in three ways by participants: community partners, formal institutions, and adolescents.
First, participants suggested collaborating with community partners such as recreation centers, as well as groups who regularly used the parks and basketball courts. Partnerships with these groups were most promising because adolescents trusted and respected the adults who led these organizations, and they often had low barriers to entry and allowed adolescents to have fun.
I would say go through trusted organizations who have built strong relationships with kids because they already trust that person versus you just making a random post or just showing up at a school saying, "Hey. We want to do this." Provider (woman, age 39) Send somebody out to local parks and courts. Send people to basically kids' after-school programs.

Young adult (man, age 19)
Where do the kids go when they leave school? What's popular in that town? Are they hanging out at the playground? Are they hanging out at rec Center? In the summer, they're at the pool. We know that, the playground. We got to [get in] the community. -Parent (woman, age 52) Second, participants suggested identifying and working with trusted individuals within formal institutions such as schools and social service agencies. These partnerships were thought to be reliable and stable given the high level of interaction that they have with adolescents. Unlike community organizations, some, but not all of the adults in these places were trustworthy. As noted by one participant, The issue is finding that contact in the school of that person that all the bad kids go to. They love, trust and respect that person, like, that's the person you want to get everything filtered through.
That's how you want to get that vulnerable population. Provider (woman, age 39) Finally, allowing adolescents affected by parental drug use to be spokespeople of the intervention was suggested. When given proper guidance, adolescents who participated and had a positive experience could tell their peers about the intervention both in person and via their social media sites. This snowball recruitment and retention method was thought to reduce some of the anxiety among adolescents associated with participating in an intervention alone. With groups of friends, so even if they're still learning to trust me, they trust somebody on the team.
That's the first, because they're peers, that's who they're going to trust firsttheir peers.

| Addressing a range of needs help adolescents survive
Participants noted that researchers often do not fully understand the lives of adolescents affected by parental drug use, and thus fail to integrate important contextual factors into the design and content of interventions. Although substance use prevention is important, it was often viewed as a lower priority by adolescents affected by parental drug use. Survival was the highest priority; intervention practices and content that enhanced chances of survival would be successful. For example, participants acknowledged that many families affected by drug use experience financial instability. Adolescents in these families are may be negatively affected by their parents' financial challenges. Participants emphasized that incentives may mitigate the negative financial effects of parental drug use.
Thus, participants stated that food and money should always be available in the intervention or as incentives because some adolescents may be experiencing food insecurity. Other basic items such as shoes, socks, and clothes were also listed as appropriate incentives given that some families may be experiencing financial strain and unable to purchase new items regularly.
That [money] could help. They could want to go to school and get something from the cafeteria and if they can't afford it, they mother didn't give them or can't give it to them, or just doing something over the weekend with their friends or whatever. Some kids do need financial support system and they don't have that. Young adult (woman, age 19) Although monetary and essential items were believed to be the highest priority, they were not the only suggestions. Other incentive options such as academic support, electronics, accessories, gift cards, and general giveaways were also suggested as they allowed adolescents to receive products they wanted and not only ones they needed.
Participants also noted that adolescents affected by parental drug use needed strategies to navigate negative community contexts. One way to do this was to have intervention components that connected individual substance use behaviors to their impact on the community. This approach would allow adolescents to both recognize the leader in themselves and identify ways to contribute to their communities. It would also give adolescents the opportunity to discuss topics without feeling hopeless. As noted by two participants, Participants thought that connecting the content to the community would more easily allow adolescents to believe that they can abstain from substance use despite their family history and neighborhood surroundings. They also argued that connecting adolescents' behaviors to community impact within an intervention would demonstrate concern for not only the young person but also the context in which they lived. Furthermore, participants stated that connecting the content to the community would permit facilitators and supporters of the intervention to reinforce strategies for why and how adolescents can abstain from substance use.

| Relatable facilitators understand and listen to adolescents
Participants noted that the facilitators are critical to participant retention. Participants described the ideal facilitators as having genuine intentions and a willingness to be vulnerable with adolescents. Participants believed that adolescents would not listen to someone who knew little about the complexity of their lives and feelings about their parents who use drugs. Therefore, facilitators who were the same race as adolescents and those with lived experience (e.g., had a parent or other family who used drugs or had a history of drug use) were preferred.
However, at the minimum, having worked with families affected by parental drug use seemed to be acceptable.
Leaders who only have extensive knowledge in the intervention, but not in the lives of adolescents would not be able to inspire hope, nor would they be well received. Two participants noted: I would definitely not advise somebody who doesn't have a real story to share to lead it. The whole project is going to be stripped of its credibility if you're asking me to share something that you can't even relate to. The young people are going to sniff it and they're going to smell it and they are going to be completely out of it. -Provider (woman, age 39) …before they can cope with it, they have to be able to understand it. Before they're open to understand it, they have to feel like somebody can relate. -Provider (woman, age 34) In addition to relatability, the ability to patiently listen and respond empathetically to adolescents was also an essential characteristic of facilitators. Participants shared that adolescents affected by parental drug use had few opportunities to talk and have someone listen, rather than immediately solve their problems or give them advice.
Listening to what they had to say about a range of topics, not only their families, was described as helping.
Not just listen, listen and understand what they're talking about. If they can listen, that would help a lot. That actually means a lot, you just listening to 'em. Because a lot of kids, they… think adults don't really care as much. But you got to listen. -Young adult (man, age 21) POWELL ET AL.

| 1003
Someone who has a lot of empathy, someone who is very skilled at listening to everything that they're not saying and being able to pull it out in a way that doesn't feel judgmental. -Provider (woman, age 39) Sometimes, we just have a rap session where kids just talk about wherever they talk about and we listen. -Provider (man, age 47) Participants believed relatable and responsive facilitators were as important as the intervention content when it came to retaining young people in interventions. Having personal experiences with families affected by parental drug use was important, but being able to listen to adolescents was equally important. Thus, facilitators who balanced their knowledge with an openness to learning more about adolescents and their experiences were assumed to be most successful in delivering an intervention for adolescents affected by parental drug use.

| DISCUSSION
The goal of this study was to describe why some strategies may be more effective than others in overcoming barriers to recruitment and retention efforts with Black adolescents affected by parental drug use in prevention research. Common barriers to recruitment and retention of vulnerable youth in prevention research include concerns about safety, fear of retraumatization, logistical issues, housing instability as well as limited utility and relevance of content. Strategies such as working with community partners, using social media, and training staff to be personable have been proposed as effective methods to better recruit and retain vulnerable youth in prevention research. Our findings extend previous research with qualitative data to explain how these strategies (and others) overcome the barriers to recruitment and retention. Specifically, participants noted that safe people and places are inviting and can minimize re-traumatization. Relatable facilitators also help to overcome barriers related to safety and trauma because youth are more easily able to build connections and trusting relationships with people who have shared their experiences. Thus, adolescents may be more willing to enroll and stay in an intervention when they feel heard, understood, and free from harm. Relying on trusted community networks may reduce barriers related to logistics and housing instability because partners become both champions of the program and connections to participants. Having a network of community ambassadors for the intervention may assist in finding safe places to host interventions and reduce the chances of attrition because someone will always know where youth are and how to encourage them to continue with an intervention. Finally, participants suggested researchers connect the content to community and life issues, and not only prevention behaviors, to increase the usability and relevance of interventions.
Our findings echo the importance of previous trauma-informed, prevention research that integrates an understanding of adverse childhood experiences, emphasizes the need for emotional safety, and creates opportunities for adolescents to rebuild a sense of control as needed (Earls, 2018;Oral et al., 2016). Among parents who use drugs, there are higher rates of concurrent mental health problems, trauma history, and PTSD, financial instability, interpersonal violence, incarceration, and stress (Stulac et al., 2019). Substance use within the household is a well-documented adverse childhood experience (Cronholm et al., 2015;Dube et al., 2003;Oral et al., 2016). Parental drug use has been associated with violent and erratic behavior that places adolescents at higher risk for neglect, physical and sexual abuse (Manly et al., 2013). Consistently, participants believed that adolescents affected by parental drug use were also at risk of suffering physical or emotional harm as a result of their parent's problematic substance use. Unfortunately, adolescents were limited in knowledge of and access to safe people and places. Future researchers might consider using an ecological approach to prevention research with this group, whereby an individual substance prevention curriculum is paired with a cross-sector strategy to connect local organizations that could provide ongoing and long-term support for these adolescents. In addition to churches and recreation centers, nontraditional spaces such as libraries, fast food restaurants, and firehouses may be options to consider. Safe Place, a national adolescent outreach and prevention program, has successfully employed a similar strategy to support adolescents in need of immediate help and safety (National Safe Place Network, 2020).
Participants endorsed a community-partnered approach to successfully recruit and retain adolescents affected by parental drug use in prevention research (Grape et al., 2018;Graves & Sheldon, 2018). The relational, networkbased recruitment and retention strategies described by participants represent a partnership between researchers and trusted adults that enable vulnerable adolescents to receive support that they may not otherwise receive.
Once adolescents are invested, they are able to become advocates and invite others via their social media pages.
Recent research has shown online social networking sites to be a promising method to connect and retain hard-toreach adolescents in interventions (Ezell et al., 2013;Thomas et al., 2020). Given the ability to create private groups and send private messages, social media platforms may circumvent stigma and conversational barriers that arise during the in-person implementation of prevention research. Ultimately, these relational approaches may yield greater and more sustained involvement among adolescents affected by parental drug use in prevention research because they start with an investment in the community.
Findings suggest that substance use interventions for adolescents affected by parental drug use may be an opportunity to address more than substance use. For example, they may be a space for adolescents to expand their network of supportive adults. Social network members have powerful influences on adolescent substance use behaviors (Ennett et al., 2006;Haas et al., 2010;Henry & Kobus, 2007;Loke & Mak, 2013). Unfortunately, many prevention interventions have narrowly focused on family or peers (alone or jointly) without considering the range of other influencers young people may have (Bröning et al., 2012;Kumpfer & Summerhays, 2006), including intervention facilitators. In particular, natural mentors, or nonparent adults who are confidants and advocates for adolescents, can be role models, provide reliable support, and promote positive adolescent development (Deutsch et al., 2020;Spencer et al., 2016;Van Dam et al., 2018;Zimmerman et al., 2002). Future recruitment and retention efforts may benefit by identifying and including community partners (e.g., faith leaders, coaches, and librarians) as facilitators to ensure that adolescents can remain connected with them after the research.
Prevention research projects may meet the basic needs of adolescents who are involved. Food insecurity (i.e., limited or uncertain access to adequate food) is a widespread problem among adolescents in low-income neighborhoods (Mmari et al., 2019;Popkin et al., 2016;Waxman et al., 2015). Offering food and small sums of money to participants may serve as both a token of appreciation and a much-needed resource. Thus, adequate, adolescentcentric incentives for participants to remain engaged should be included in the budget.
Finally, engagement in prevention research could empower adolescents to share their voices about substance use in the community. This finding aligns with research on critical youth engagement which suggests that when adolescents are engaged in activities that focus outside of themselves, they are well suited to connect research to action. Sustained engagement by adolescents and researchers could spark hope for a better future for participants and their communities, which makes prevention research not only helpful but also relevant.
There are three limitations to note. First, although consistent with a qualitative approach (Creswell, 1998;Patton, 1990), the sample size of this study is small and specific. Our participants were mostly African American, lower-income individuals who reside in a city with high overdose rates. This does not represent the range of families affected by parental drug use. Thus, our ability to generalize findings beyond this sample is limited.
However, our findings may be essential to those seeking to conduct interventions with urban, Black families affected by parental drug use. Second, our participants were willing to talk about their experiences with familial drug use. Individuals who are less comfortable discussing these issues with people outside of their close social network may describe different experiences. Finally, young adults, not adolescents, were included in this study.
Their responses may be subject to recall bias. Future researchers desiring to conduct in-depth interviews with vulnerable adolescents might consider partnering with mental health providers and embedding such research within the context of long-term treatment. This collaborative effort best ensures that the benefits outweigh the harm of vulnerable youth sharing their stories.
Despite these limitations, this study makes several important contributions to the literature. First, it is one of the first studies to qualitatively explore perspectives of effective recruitment and retention strategies for engaging Black adolescents affected by parental drug use in prevention research. Each theme expanded our understanding of how to better meet the needs of families affected by parental drug use. Second, this study incorporated input from three stakeholder groups: young adults affected by parental drug use, parents with a history of drug use, and providers who work with families affected by parental drug use. These diverse perspectives reinforced the need for tailored, trauma-informed substance use prevention programs and provided directions for future research.

ACKNOWLEDGMENTS
The authors would like to thank the families who participated in this study.

CONFLICT OF INTERESTS
The authors declare that there are no conflict of interests.

FUNDING STATEMENT
This study was funded by a grant from the National Institute on Drug Abuse (1K01DA042134; PI: Powell). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute on Drug Abuse or the National Institutes of Health. The sponsor had no role in the study design, collection, analysis, or interpretation of the data, writing the manuscript, or the decision to submit the paper for publication.

ETHICS APPROVAL
All procedures performed in studies involving human participants were in accordance with the ethical standards of the Institutional Review Board at Johns Hopkins Bloomberg School of Public Health and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

CONSENT TO PARTICIPATE
Informed consent was obtained from all participants before participation using an IRB-approved consent procedure.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.