Rigorous evaluation of a substance use and teen pregnancy prevention program for American Indian girls and their female caregivers: a study protocol for a randomized controlled trial

Extensive research documents how both early sexual debut and substance use intiation adversely impact the health of an individual. Specifically, early sexual debut increases the likelihood of poor sexual and reproductive health outcomes, such as sexually transmitted infections (STI) and unintended teen pregnancy [1‐4]. It is also associated with a greater number of sexual partners and increased risk of being a victim of intimate partner violence [4‐6]. In terms of substance use, early initiation of substance use adversely impacts the cognitive and emotional growth and development of adolecents [7‐9].

Early sexual debut and substance use are also costly for the individual, society and tax payer. A recent analysis conducted by Rotz et al. found large economic savings of preventing teen sexual activity and substance use. The analysis estimates a net benefit of up to $52,109 for females and $27,861 for males from delaying voluntary sexual activity to age 15 or older [10]. For substance use, the estimated cost savings for preventing underage drinking is estimated to be up to $12,313 for individuals [10].

Not surprisingly, substance use initiation is one of the most significant risk factors for sexual initiation and unprotected sex among adolescents. In fact, these behaviors tend to co-occur, particularly among more vulnerable subgroups of adolescents [11, 12]. These behaviors also share similar risk factors including lack of family cohesion [13‐17] and poor adolescent functioning (having high levels of internalizing and externalizing behaviors). They also share many protective factors including connection to culture [18‐21], parent-child relationship, self-efficacy and future aspirations [21‐26]. Programs targeting these shared risk and protective factors may have a dual impact: reducing both sexual risk taking and substance use among adolescents.

National data show Native American (Native) adolescents are more likely to initiate sex before age 13 than all other U.S. racial/ethnic groups, except African American adolescents [27]. Native high school youth are also more likely to have multiple sex partners and to ever have had sex compared to youth of all races [27, 28]. Native youth also experience higher rates of STIs and teen pregnancy than their non-Native counterparts; in 2018, U.S. national Chlamydia and Gonohrrea rates among Native Americans were 3.7 and 4.6 times higher than that among Whites [29]. Further, in 2017, the national teen birth rate among 15–19 year old Native Americans was the highest of all races and ethnicities [30, 31]. It follows that Native adolescents have higher rate of substance use and abuse. Compared to all U.S. adolescents, in 2018, Native adolescents ages 12–17 are more likely to engage in past-month binge drinking (5.7% vs. 4.9%), marijuana use (11.1% vs. 7.4%), and other illicit drug use (12.8% vs. 8.7%) [32].

Given the adverse outcomes of early sexual activity and substance use and the high rates of sexual risk taking and substance use among Native adolescents, prevention efforts to delay these behaviors in Native communities are warranted [33]. These efforts may be more advantageous if delivered during early adolescence (ages 10–14), before a child initiates sex or substance use. Early adolescence is a key developmental period in which children continue to solidify their values and relationships, and learn foundational life skills [34‐37]. Hindered growth in these areas during this time period has been linked to poorer health outcomes including early sexual initiation, substance use and abuse in later adolescence and adulthood [33]. Thus, programs that promote values, focus on building healthy relationships and teach life skills such as problem-solving and goal-setting in early adolescence are needed.

As opposed to many Western societies, most Native cultures do not have a self-centered orientation to health, but instead have family- and community-oriented health frameworks [38]. Thus, intergenerational programming may be more culturally appropriate [39] than individual-level programs. In addition to cultural congruency, engaging parents alongside adolescents in programming may be advantageous as they can support and reinforce behaviors [40, 41]. While robust efforts have begun to address behavioral health disparities among Native adolescents, currently there are no intergenerational, efficacious primary prevention programs targeting the dual threat of sexual risk taking and substance use for young Native adolesents [42‐44]. This study seeks to fill this gap.

The objective of the is to assess the efficacy of a program called Asdzáán Be’eena’ (Female Pathways, AB) for increasing protective factors associated with delayed substance use and sexual risk taking among Native girls in early adolescence (ages 10–14) through a randomized controlled trial (RCT). This paper describes the study protocol.

The study will test the efficacy of the AB program on risk and protective factors associated with delaying adolescent substance use and sexual initiation, as well as actual sexual initiation and substance use. By assessing protective and risk factors as well as behaviors among the adolescent and the caregiver, the study team will be able to assess how each of these are impacted by the intervention, and how these interact with one another to moderate program impact.

Primary research questions include: 1) Is the intervention effective in increasing adolescents’ intention to abstain from sex (intention to abstain was chosen as opposed to sexual initiation as we do not anticipate high rates of sexual initiation for the duration of the follow-up time period); 2) Is the intervention effective in improving caregiver-adolescent relationships. Secondary research questions include: 1) Is the intervention effective in improving adolescent risk and protective factors associated with delayed sexual initiation and/or substance use initiation; 2) Is the intervention effective in improving caregiver risk and protective factors associated with delayed adolescent sexual initiation and/or adolescent substance use initiation; 3) Is the intervention effective in delaying sexual initiation and substance use initiation among adolescents; 4) Is the intervention effective in reducing substance use among enrolled caregivers.

The study team hypothesizes a higher proportion of adolescents in the intervention group will state they intend to abstain from sex in the next year and until they graduate high school compared to the control group at 6- and 12-months post intervention. Additionally, we hypothesize that there will be 1) an increase in caregiver-adolescent communication at 6- and 12-months post intervention as measured by a communication scale adapted from a trial with Ojibwe adolescents, 2) an increase in caregiver-adolescent quality time together at 6- and 12-months post intervention as measured by a quality time with parent scale adapted from a trial with Ojibwe adolescents, 3) an increase in maternal warmth at 6- and 12-months post intervention as measured by the Authoritative Parenting Index, and 4) an increase in caregiver monitoring at 6- and 12-months post intervention as measured by the parental monitoring scale.

The study is a two-armed randomized controlled trial examining the primary outcomes of adolescent-caregiver relationship, adolescent intention to have sex and adolescent adaptive/maladaptive functioning at 12 months post intervention. The trial is being conducted at two study sites on the Navajo Nation with Navajo girls ages 10–14 and their female caregiver. Randomization will be performed with a stratified block randomization sequence to ensure a 1:1 allocation of study conditions within each of the two study sites Fig. 1.

The study will be conducted by the Navajo Nation in partnership with Johns Hopkins Center for American Indian Health (Center). The study design was approved by the appropriate Navajo governing bodies including Chapters, Agency Councils and Health Boards as well as the JHU research review board and the Navajo Nation Human Subjects Research Review Board (NNHRRB). This manuscript was approved by the NNHRRB. All deviations from the protocol will be approved by IRBs before being implemented. Once approved, they will be updated in clinicaltrials.​gov by the PI. As needed, participants will be reconsenting utilizing consenting documents that include changes to the protocol and have been approved by the appropriate IRBs.

The study is community based intervention and conducted on the Navajo Nation in the United States of America.

Participants are Native female adolescents and their female caregiver (referred to as dyads). Inclusion criteria for the adolescent includes: Native American ethnicity, 10–14 years of age, living within 50 miles of one of the two study sites, and written parental permission and adolescent assent. Participant inclusion criteria for the caregiver includes: Native American ethnicity, ≥ 18 years of age, caregiver to a female child age 10–14 years who will enroll in the study with them, living within 50 miles of a study site, and written informed consent. If the caregiver is not the legal guardian of the adolescent, the adolescent’s legal guardian must agree for the caregiver to participate. To honor cultural views around caregiving, female caregivers who live outside of the home but play a pivotal role in child rearing are welcome to participate. To enroll, both members of the dyad (adolescent and caregiver) must be eligible and enroll. Adolescents in foster care are ineligible to participate.

Participants will be recruited through a combination of non-probability and snowball sampling. First, recruitment fliers will be posted in community gathering spots (i.e., supermarket, community centers, health and human services offices, fitness centers, etc.), on social media (e.g. Facebook) and disseminated during public gatherings through informational booths. Second, caregivers enrolled in the project will be asked to refer other caregivers who may be eligible and interested in the program. Third, IEs will provide information about the study to local agencies that provide services to eligible participants (e.g. Office of Self Reliance, local schools, counseling services etc.) and provide them with referral forms. Independent Evaluators will collect referral forms from these agencies on a regular basis (e.g. weekly). All recruitment materials have contact information for the local project coordinator. To learn about the project and enroll, interested individuals can call the local project coordinator. A study staff member will also reach out to referred caregivers or parents of a referred child. When contact is made with a potential participant, the IE will review a brief script describing the study and assess for eligibility. For those individuals meeting eligibility criteria who express interest in study participation, the study staff member will work with them to set up a time to complete informed consent/parental permission/assent and work with the participant to contact the caregiver who will enroll with them (adolescent and adolescent’s female caregiver). The informed consent process will include a review of how personal information will be collected, shared and maintained to protect confidentiality of the participant’s study data. After both the caregiver and adolescent participants have completed informed consent and adolescent assent respectively, they will complete the baseline assessment and schedule their first lesson.

Participants will receive gift cards for participation in assessments ($20 per assessment time point per participant). Additionally, FHCs and IEs will do all they can to keep participants enrolled and engaged including providing transportation to study visits and snacks or meals at these visits. Participants will only be dropped from the program if they request. Outcome data will be collected from all participants regardless of their intervention completion status.

Assessments will be administered at baseline, immediately following program completion and 6- and 12-months following program completion (see Table 2). Data will be collected in participants’ homes or another private location via a tablet using Research Electronic Data Capture (REDCap). Caregivers will complete all assessments via self-report. Adolescents will complete portions of the outcome assessment asking about sexual and substance use behaviors via self-report. All other portions will be be completed via interview. Participants will be given $20 after completion of the baseline, immediate, six and twelve month post intervention assessments. All measure items were previously piloted with adolescent and caregivers from the community to assess comprehension, language, and cultural relevance. Edits were made accordingly and pilot participants were not eligible for the RCT.

Please see Table 2 for a description of the outcome measures, including method and frequency of administration.

All data will be directly entered into REDCap by an intervieweer (IE) or the participant. If assessments are completed via paper, they will be entered by the IE. A data manager will review all data entered into REDCap to ensure completeness and conduct data quality assurance (including checking ranges for variables as applicable). All data will be stored via REDCap, a secure, HIPPA compliant, data collection and storage database. Data will be analyzed part way through the trial by the evaluator. The study management team (including the evaluator, the PI and the program manager) will review data and assess the need to continue the trial.

This study will use a 1:1 randomized controlled trial design; the unit of randomization will be the individual dyad. Unique participant identification numbers will be used to randomize individual dyads to one of two study groups: intervention or control. Stratified block randomization, using random block size will be used to ensure a 1:1 allocation of study conditions across the two study sites. The sequence will be generated by the program evaluation via STATA and will not be shared with study staff who enroll participants.. Independent Evaluators will enroll participants and reveal study group after the participants complete the baseline assessment. The study assignments are sequentially numbers in sealed envelopes that are only opened at the time of randomization. No blinding will occur.

Sample size and statistical power estimates are predicated on 2 primary hypotheses: 1) that those randomized to the AB intervention will have improved parent-adolescent communication compared to the control group, and 2), that those randomized to the AB intervention group will have a higher proportion of participants who report they intend to abstain from sex.

The between study group equivalence of participant characteristics and outcome measures at baseline will be compared using Chi-squared tests (dichotomous/categorical data) and t-tests (continuous data). Study hypotheses related to the confirmatory questions will be initially tested under an “intent-to-treat” model in which data are analyzed according to treatment assignment at randomization. We will then conduct “completer analyses” on those subjects completing at least two-thirds of intervention sessions. Intervention impact will be evaluated by comparing primary study outcomes between intervention and control groups across the three time points: baseline, 6 months and 12 months post-intervention. Initially, summary scores of outcomes will be stratified by participation in the intervention versus control groups and by time point, and will be compared using chi-square tests (binary or categorical outcomes), t-tests and analysis of variance (continuous outcomes). Generalized linear mixed effects models (GLMM) accounting for within-group correlation across longitudinal data will then be applied. Outcomes will be modeled as a function of group assignment and time since intervention. The association between outcomes and time will be explored visually using scatterplots and lowess smoothers. Restricted cubic splines will be added to reflect deviations from linearity.

If siblings enter the study, they will be randomized to the same intervention group to limit contamination. In this case, analyses will be adjusted to take into account intra-sibling correlation as well. Additionally, sensitivity analyses will be conducted including and excluding siblings to examine the impact on study results. Missing data will be handled as follows: 1) identify erroneous data, 2) document reason(s) for missing and erroneous data to inform model development; 3) compare demographic characteristics of those who dropped out and those who did not to compare if data is missing at random; and 3) conduct sensitivity analysis to compare inferences that are based on different plausible reasons for missingness. Navajo and Hopkins investigators will collaborate with community partners in data interpretation to assure accuracy, cultural acceptance and relevance.