Homeless youth were recruited via referrals from case managers at a youth homeless shelter located in a large urban city. Individuals who participated in the focus groups met the following inclusion criteria: (a) English-speaking, (b) homeless, (c) aged 18 – 22 years old, (d) currently sheltered, and (e) willing and able to comply with requirements of the study protocol. We had no eligibility requirement regarding access or use of technology. Homelessness was defined as lacking “a fixed, regular, and adequate nighttime residence”, and thus, currently sleeping in the shelter at least 50% of the nights in the past week (U.S. Department of Education, 2001).

The main goal of the focus groups was to explore the difficulties homeless youth face with receiving mental health care to guide potential solutions to these problems, especially solutions that make use of technology-based resources. Borrowing methodologies from the participatory design model, these focus groups aimed to generate a creative and collaborative experience, to build understanding between researchers and the homeless youth, and to engage youth prior to any formal designs to help understand needs and resources (Muller, 2003). We focused on the youth stakeholder group, rather than staff, as our goal was to design youth-facing tools that could be used regardless of living situation (i.e., sheltered or unsheltered) or staff involvement. Focus groups lasted 90-minutes and took place at the shelter during a time that was convenient for all shelter residents (6:00-7:30pm, as not to interfere with school or work obligations). Participants were compensated with a $15 gift card and dinner was provided during the focus group sessions. Five focus groups were conducted consisting of 4–6 participants per group along with 2–3 facilitators from our research team, resulting in 24 homeless youth participating in this study. The study was approved by the Institutional Review Board at Rush University Medical Center.

Informed consent was obtained at the beginning of each focus group. Participants were instructed to keep all information discussed during the group confidential. In order to respect privacy, only the first names of all group members were used. In addition, participants were reminded that their participation was voluntary and that they were not required to disclose anything that would make them uncomfortable. Following this disclaimer, a brief demographics questionnaire which collected information about age, gender, sexual orientation, race, ethnicity, education, and homeless experiences was administered. A semi-structured interview was used to guide a collaborative discussion of the youths’ needs and expectations for a smartphone-based mental health intervention (Table 1). The semi-structured interview explored four main topics: current use of technology, experiences with the mental health system, target areas of mental health concerns, and privacy and security related to the use of technology-based interventions. Interviews were audio recorded to facilitate transcription and analysis of focus group data.

The interviews were audio recorded, transcribed, and coded using ATLAS.ti (Muhr, 1997). Analysis was guided by the grounded theory approach, which provides a structure and techniques to interpret and organize the data. Grounded theory is a qualitative method that focuses on the evaluation of data in order to generate new theoretical explanations, rather than testing a previous theory (Strauss & Corbin, 1990). As such, themes are allowed to emerge and change over time as new insights are uncovered and new data is gathered. Coding began after the first focus group was transcribed. The lead author, who did not participate in the focus groups, led the coding. Coding was conducted by creating labels to describe and reduce the data while maintaining the substance of the text (Charmaz & Belgrave, 2002). As the data were coded, the lead author created codes for each new idea and made memos of emerging interpretations and alternative conceptualizations. As similar codes emerged, these codes were grouped into themes and reviewed with the senior author. Themes were modified throughout the analysis as they were compared to one another, to new codes, and to new data. Once all of the data were coded and all themes identified, the lead author again reviewed themes with the senior author (who led all focus groups), adjusting and consolidating themes where necessary. Grounded theory continues until theoretical saturation is obtained, which means that the themes maintain consistency and little information is gained by additional data. Saturation was reached after the fifth focus group. Because codes and themes evolve over the analysis, consistency is reached through constant comparison among the data and consultation among the research team members.

The sample consisted of 24 homeless youth with an average age of 18.8 years (SD = 0.8). The sample was 62.5% female and 66.7% African American. About 47.8% were in school and less than half were employed. Participants reported an average of 3.3 (SD = 2.6) homelessness episodes in their life, with the longest episode lasting an average of 10.7 months (SD = 13.4). Complete demographics are displayed in Table 2.

These findings need to be interpreted with the study’s limitations in mind. Participants were recruited from one homeless youth shelter in a large urban city and were selected by the shelter’s case managers as being a good fit for the study. Thus, it is unclear whether these findings would generalize to unsheltered youth or if other characteristics of our participants made them more likely to be identified by our case management. In addition, the questions asked during the focus groups were adapted as more sessions were completed. Therefore, themes that were present in early focus groups may have guided questioning in subsequent groups. This practice, however, is common in qualitative work, in which the goal is to allow themes to emerge and to gain more information about those areas (Strauss & Corbin, 1990). Also, individual participants were not identified during the transcription process. Therefore, we could not analyze if certain characteristics of the participants (e.g., history of homelessness, mental health characteristics) impacted their feelings to any of the topics discussed during the focus groups. Furthermore, we were not able to identify how many youth actually endorsed each topic or statement discussed. As such, a highly vocal youth might be more likely to impact the discussion and thus the prevalence of a particular theme. We appreciate that this is often a concern when conducting the focus groups, and group moderators used several methods, such as soliciting feedback from each user through writing ideas on sticky notes, looking for overlap and themes within each group, and having focus group members vote or indicate their preference with a thumbs up or thumbs down. Finally, based on participatory design principles, we focused this investigation on youth stakeholders. This overlooks input from the shelter staff or other organizational leaders. We focused on youth because this study was meant to inform the development of a youth-facing intervention that would not require the involvement of the shelter staff and could be used whether or not a youth was currently sheltered. However, future research should take into consideration the views and experiences of other stakeholder groups as their input, needs, and workflow might identify additional requirements and might contribute to scalability and sustainability of developed resources.

Access to mental health services is limited among homeless youth, and thus, new intervention models must be considered in order to provide the necessary care. Based on our findings, future interventions should address the barriers to mental health access while taking into consideration the importance of trust and rapport between providers and homeless youth, and the desire for general support and help with mental health problems. While it is sometimes assumed that homeless youth do not have access to technology, our findings validate other recent work demonstrating that homeless youth access and use mobile phones regularly, especially for communication, music and social media. Technology-supported interventions provide a potential avenue for reaching homeless youth and addressing their mental health needs. Researchers utilizing technology-supported interventions should be transparent regarding the risks and benefits of the mobile phone platform, especially if data sharing occurs. Allowing homeless youth to have a say in what information is tracked and shared is critical for building their trust. Further research is needed to validate the findings of this study in other homeless youth populations and support the development of technology-based platforms for delivering mental health services to homeless youth.

Overall though we have highlighted several potential features that could be present in future resources to meet the needs of this population. Apps and programs should focus on providing frequent, actionable information possibly through e-mails, text messages, or push notifications. Furthermore, monitoring and feedback systems that help youth stay connected with providers and pass useful information (such as symptom change and side effects of medications) might help guide evidence-based practices when frequent office visits are not possible. Future research projects and efforts should make use of the findings from this study to develop apps and programs to address the mental health needs of homeless youth while being mindful of convenient and feasible ways to receive mental health services. We believe novel types of services such as technology-supported programs have the potential to transform and expand mental health care for this population and address the overwhelming need in a context where traditional resources are insufficient.