A qualitative approach modified the original CAMI intervention designed for young women [14–16] for use with young men. Individual interviews with 27 men, aged 15–24 years, were conducted to assess their opinions on the content and delivery of a mobile app–assisted health intervention. There were three sets of interviews, each exploring a different component of the formative research: (1) coaching sessions and communication with the coach; (2) feedback format, timing and style from the app; and (3) role of the coach and impressions of a value card sort activity. [18] In addition, a Community Advisory Board of young men from the study population began meeting monthly in late 2016 and provided feedback which informed the design of many aspects of the study, including privacy and confidentiality, wording and length of surveys, recruitment strategies and materials, and participant engagement and retention. The adapted intervention will be pilot tested with nine participants to ensure functionality and feasibility, particularly of the mobile app and the coaching intervention.

Participants in the CAMI-TPP arm will receive four 30- to 45-minute sessions of one-on-one telephone MI coaching (at enrollment or within the first week of enrollment, at 2–4 weeks, at 8 weeks, and at 12 weeks). The sessions will be guided by an app-generated, personalized feedback aimed at increasing condom use, supporting female partners in contraceptive use, and obtaining reproductive health services and STI testing. Each participant will be offered, with permission, clear advice to avoid being involved in a pregnancy until he desires it and avoid getting STI’s by using condoms correctly and consistently, by limiting the number of sexual partners, and by supporting his female sexual partners in using MERCs or HERCs, and getting STI testing. Referral for a reproductive health clinic will be offered and encouraged.

Participants in the CAMI-Fitness arm will also receive four 30- to 45-minute sessions of one-on-one telephone MI coaching (at enrollment or within the first week of enrollment, at 2–4 weeks, at 8 weeks, and at 12 weeks.) The sessions will be guided by an app-generated, personalized feedback aimed at healthy diet, tobacco avoidance, and physical activity. Each participant will be offered, with permission, clear advice to avoid tobacco, maintain a healthy weight/body mass index, eat healthily, and participate in physical activity. With permission, the coach will offer or direct the participant to detailed information about tobacco avoidance, healthy diet, and physical activity according to the young man’s needs and requests. Referral for an athletic training visit will be offered and encouraged.

The two interventions are identical in length and timing but vary in the behavioral target (pregnancy prevention vs. fitness). All coaching sessions will be conducted by phone or video call. Formative research has reaffirmed that most young men in our study population will feel comfortable talking with their coaches by phone rather than in person and may find it more convenient than having to come in for an in-person meeting.

CAMI coaching sessions for topics in each intervention arm will embody the spirit of MI in every session (collaboration, acceptance, compassion, and evocation). Coaches will utilize MI processes (engaging, focusing, evoking, and planning) and core MI communication skills (open-ended questions, affirmations, reflections, and summaries).

Coaches will use MI style with each participant to collaboratively identify areas for behavior change and, with permission, develop a specific, tailored plan with each participant to consider and implement. The coach will use an agenda-mapping chart to determine what specific changes each participant wishes to make related to the intervention topic. Using a ruler scale, the coach will assess and explore each participant’s perceived importance of making specific intervention-related behavior changes as well as his confidence in making those changes. The coach will also use a ruler scale to assess and explore participants’ readiness to make changes. Finally, if the participant expresses readiness to create a behavior change plan, the coach will work with him to create a plan outlining specific steps toward change. The plan will identify each participant’s personal reasons for change(s), facilitating factors and barriers to the plan, and potential solutions to barriers.

A key component of any Ml-based intervention is for the coach to support each young man’s autonomy by explicitly stating that it is up to him, not the coach, to make decisions regarding behavior change. Coaches will lead the participant in a directed, empathic discussion that includes (1) reviewing intervention-specific behavioral data that the young man entered into his mobile device app (if he agrees to share his data) and (2) discussing future goals and values related to education, career, and intervention-specific goals such as relationship and parenting goals or fitness goals, depending on intervention group. The coach will support each young man’s decision to make (or consider making) change.

The CAMI protocol scripts for all four MI sessions in both intervention arms are available on request and these provide detailed examples of the types of behavioral changes participants could make and could discuss with the health coaches. CAMI conversations, when a participant is not in a current or stable relationship, might include open-ended questions and reflections about pregnancy and STI prevention goals with new or causal partners (and how this might be different from current or stable relationships), when and how to bring up the topic of pregnancy and STI prevention with new or casual sexual partners, potential barriers to keeping oneself safe from STIs and being involved in an unintended pregnancy with a new or casual sexual partner as well as potential solutions to these barriers. Coaches will also affirm participants’ willingness to think ahead about these potential situations and protect themselves as well as affirm their resourcefulness in generating solutions to potential barriers and challenges to consistent and effective condom use and other self-protective behaviors to avoid unintended pregnancy involvement and STI acquisition.

The app will provide each participant with feedback specific to his intervention arm and tailored to the behavioral data he entered for checkins and periodic assessments. CAMI-TPP will provide feedback to each participant on his personal risk for being involved in an unintended pregnancy and getting an STI. CAMI-Fitness will provide feedback to each participant on his tobacco use, body mass index, eating, and physical activity.

To promote and assess treatment integrity, all CAMI coaching sessions will be audio-recorded and reviewed weekly by the CAMI supervisor (M.A.G.), who will use a CAMI adherence checklist developed for the study to assess adherence to CAMI session content and process (quality MI microskills) and use the Motivational lnterviewing Treatment Integrity scale or MITI 3.1.1 to assess and give feedback to coaches about the quality of MI delivered at each session. Coaches will be laypeople hired from the community who have a minimum of a high school diploma and who are bilingual. They will have completed a written Helpful Responses Questionairre with six scenarios and separately will have completed a 10-minute audio recording that will be coded using the MITI 3.1.1. Laypeople will be hired and trained in MI for three reasons: (1) cost of staff; (2) better rapport between the participants and the coaches since coaches are not seen as “experts” but rather as peer helpers to participants in determining what types of behavior changes to choose and how to implement and build on those changes over time; and (3) the ease to train laypeople in MI than retrain professionals or paraprofessionals who may have prior, more didactic interactions with participants or clients in which they are seen and see themselves as “experts.” Before enrolling participants, coaches will undergo rigorous MI training and operate under significant oversight to ensure fidelity to the principles of MI. Each coach will receive individualized feedback and coaching by the supervisor (M.A.G.) who will review every audio-recorded practice session until the coaches are certified to competently conduct CAMI coaching based on the MITI 3.1.1 coding. After enrollment and MI coaching for participants has begun, coaches will receive personal feedback and coaching for their first 10 audio-recorded CAMI coaching sessions and overtime will be provided with ongoing written feedback and coaching for every other to every third audio-recorded session. To ensure fidelity, an independently contracted reviewer team will code at least 10% of all CAMI sessions using the MITI 4.2. Any coach who falls below acceptable levels of proficiency in MI will be identified and retrained by the supervisor until able to demonstrate competency and skills.

To promote engagement and minimize loss of data due to attrition, all participants in both arms will be asked to complete a biweekly check-in using the app platform to track a narrow set of sexual behavior and fitness behaviors (tobacco, marijuana, and alcohol use, diet, and physical activity) over the previous 7 days. Measures for the biweekly check-ins will use the same or very similar question phrasing as the assessment items, but using 7-day time periods, as summarized in Table 1. Check-ins are designed to be completed in fewer than 5 minutes.

Participants will complete assessments at baseline and weeks 12, 24, 36, and 64. Participants will use the app platform to complete an assessment to collect data on socio-demographic characteristics, relationship status and pregnancy intentions, reproductive and sexual health history and health care utilization, contraceptive knowledge and attitudes, sexual behavior, and physical fitness (eating, smoking, alcohol consumption, and physical activity), as summarized in Table 1. Assessments will use phrasing from items on existing national surveys that have been used specifically with young men aged 15–24 years [19–24]. Measures on tobacco use, eating, and physical activity will rely on survey items from the National Youth Physical Activity and Nutrition Survey [25]. Sexual behavior measures will include required performance measure items developed by the Office of Adolescent Health [26], to allow for coordinated outcome evaluations across funded interventions [26]. Validated scales will be used to measure key mediating variables including intention to seek medical care with the Action/Intention subscale of the Medical Help-Seeking scale [27–29] and therapeutic alliance (using the revised Helping Alliance Questionnaire [30]). Assessments will require 30–35 minutes to complete. Participants in both arms will complete the same assessments at each time period.

The primary outcome of interest is engaging in sexual intercourse without condom use at last sex during the last 3 months (“risky sex”), an outcome derived from two sexual behavior measures: having any sexual intercourse with a female partner in the last 3 months (yes/no) and condom use at last sexual intercourse. This primary outcome will include all participants, regardless of their frequency of sexual activity following enrollment [31]. A participant who reports no acts of sexual intercourse during the 3-month report period will be coded as not having the primary outcome. A participant who reports that a condom was not used, or that he is unsure if a condom was used at last sex, will be coded as having the primary outcome. A participant who reports that he used a condom at the last act of sexual intercourse will be coded as not having risky sex.

Secondary outcomes relating to behaviors that increase risk of being involved in a teen pregnancy (specific aim 2) include the use of a MERC or HERC, defined as a method with <10% failure rate over 12 months of typical use [32], by female sexual partners at last intercourse (asked separately for each sexual partner in last 3 months); number of acts of sexual intercourse in the assessment period that were unprotected by condoms; number of acts of sexual intercourse in the assessment period that were unprotected by partner use of MERC/ HERC; consistency of condom use; number of times each week participant had sexual intercourse without using a condom; and number of times each week participant had sexual intercourse without partner using MERC/HERC. Secondary outcomes relating to behaviors that increase risk of STIs and HIV (specific aim 3) include having a visit for sexual or reproductive health care since the last assessment period; receiving a test for gonorrhea, chlamydia, and/or HIV since the last assessment period. Physical fitness outcomes to be compared in both groups include physical activity (any, vigorous, or strength building); consumption of various food groups and beverages; cigarette smoking; marijuana use; alcohol use; and binge drinking.

Several mediating pathways will be assessed, including therapeutic alliance (a key theoretical component of the logic model); knowledge about condoms and MERC/HERC methods; communication with partners and health care providers about condoms and MERC/HERC; attitudes around male involvement in contraceptive decision-making; and intention to seek medical care.

Sociodemographic and sexual history variables that have been identified a priori as potential moderating factors that may influence the extent to which the intervention has an impact on the primary, secondary, and mediating outcomes include (as listed in Table 1) sociodemographic variables including age and race/ethnicity; access to health care; sexual partnerships; pregnancy intentions; past involvement in pregnancy; STI history; and perceived fertility.