More than 750,000 adolescents in the United States aged between 12 and 17 years have alcohol use disorder (AUD), according to current estimates [1]. Among adolescents who present for AUD treatment, >80% have a co-occurring mental health diagnosis [2,3]. Youth with AUD and co-occurring disorders (AUD+CODs) exhibit poor family functioning, low treatment motivation and engagement, poor treatment outcomes, and high rates of returning to substance use [4-6]. Although treatments for adolescents with AUD+CODs yield small, short-term benefits, caregiver-involved treatments are associated with the best outcomes [7,8]. This mirrors findings for caregiver-involved treatments for myriad adolescent mental health conditions (eg, depression and conduct disorder) [9]. Throughout this paper, we use the term caregiver to reflect that figures other than biological parents (eg, grandparents and stepparents) can serve as caregivers for youth in treatment. To date, limited work has examined caregiver behaviors during adolescent AUD+CODs treatment [10]. Building on theoretical and empirical research from clinical and developmental science emphasizing the importance of studying caregiving practices in the moment [11-13], this protocol describes the rationale and methods for a study seeking to address a central yet unanswered question: What specific caregiving behaviors during caregiver-involved treatment for AUD+CODs contribute to improved outcomes?

Caregiver involvement in adolescent services is positively associated with adolescent treatment engagement, motivation, retention, and outcomes [9,14-20]. Moreover, adolescent AUD treatments that involve caregivers have been found to reduce co-occurring externalizing and internalizing symptoms [21-24]. These findings align with evidence that caregiving practices mediate adolescent treatment for internalizing (eg, depression and anxiety) and externalizing (eg, attention-deficit/hyperactivity disorder and oppositional defiant disorder) conditions that often co-occur with adolescent AUD [25]. Consequently, current guidelines identify caregiver-involved adolescent AUD+CODs treatment as a best practice [8].

Research examining what caregivers are doing in their routine environments during the course of treatment to promote improved youth outcomes is at a nascent stage [26,27]. Moreover, although multiple evidence-based treatments (eg, cognitive behavioral therapy, family-based therapies, and multicomponent therapies) call for the involvement of caregivers in some capacity [26,28-31], typical community-based practices for treating adolescents with AUD+CODs do not involve caregivers [8,32-34]. Structural barriers to delivering caregiver-involved treatments in outpatient settings include the high costs of the training required for family-based treatments, session frequency and duration, and highly manualized protocols with limited flexibility [8,26,35]. These barriers may affect youth considered the most vulnerable whose families have limited knowledge, access, and resources to complete caregiver-involved treatments [26,36,37]. One way to advance treatment options for youth is to identify the specific caregiving practices (ie, change-promoting parenting behaviors) during AUD+CODs treatment that most strongly relate to positive changes in core treatment targets. By isolating these behaviors, we can develop scalable and targeted caregiving interventions to maximize those treatment ingredients that yield the highest impact while mitigating required resources [38].

Unfortunately, most of what we know about how caregiving behavior relates to adolescent AUD outcomes derives from observational studies of adolescents without AUDs and not in treatment [39]. To our knowledge, only 7 studies have assessed caregiving behaviors as putative mechanisms of adolescent substance use treatment outcomes [20,40-45], and only 1 of them assessed caregiving behavior during treatment [40]. In the study, growth in parental monitoring from baseline through 12-month follow-up predicted decreased growth in substance use from baseline to 12-month follow-up [40]. Although the study provides additional support for the importance of caregiver behavior to treatment outcomes, the study design did not prospectively examine how caregiving during treatment predicts outcomes [40]. Moreover, the study, like virtually all caregiving-AUD research [39,46], operationalized caregiving via assessments spaced months apart that aggregated caregiving behaviors across weeks or months [11,47]. Harnessing the impact of caregivers to boost adolescent AUD outcomes in the most impactful and efficient way requires methods that capture how caregiving behaviors, in real time in daily life, relate to core treatment targets during therapy.

Dynamic system theory posits that momentary interactions between a caregiver and their child create a reciprocal interaction system (ie, relationship pattern) that impacts long-term adjustment [11,48,49]. In a treatment context, this suggests that examining how caregivers act toward their children during AUD+CODs treatment is critical to understanding long-term treatment outcomes. A growing body of work from developmental science highlights the importance of studying caregiving behaviors and their associations with the indicators of adolescent adjustment at a more refined timescale [50-53]. Consistent with dynamic systems theory, these studies have found caregiving practices to vary from moment to moment depending on context, affect, and adolescent behavior [11-13]. Importantly, emerging research indicates that associations between caregiving practices and adolescent adjustment indicators do not generalize across different timescales [54]. In other words, capturing caregiver practices and adolescent adjustment on broader time courses (eg, monthly) may not representatively encode what occurs moment to moment in daily life.

Core adolescent AUD+CODs treatment targets, such as treatment motivation [55,56], alcohol craving [57-59], substance use [59,60], and internalizing and externalizing symptoms [61-63], vary considerably within and across days. Understanding improvements prompted by caregiver-involved treatments for adolescent AUD+CODs requires momentary assessments that can capture and correlate the dynamic nature of caregiving and adolescent treatment targets. Ecological momentary assessment (EMA) provides a methodology to study caregiving behaviors and evaluate their impact on adolescent functioning in real time [64,65]. EMA has been suggested to minimize recall bias, memory heuristics, and demand characteristics that can occur in caregiving research [64,66-68]. It has the unique ability to go beyond looking at how caregiving predicts behaviors across families (between-family associations) and provides rich data on how changes in a caregiver’s behavior relate to changes in their adolescent’s behavior in real time in the family’s daily life (ie, within-family associations [64]). Despite the developmental salience of these dynamic associations, no research has examined these momentary associations between caregiving practices and adolescent behavior during AUD+CODs treatment.

The proposed protocol aims to examine caregiving practices, in the moment, during adolescent community outpatient AUD+CODs treatment. To advance our understanding of caregiving practices that promote improvements in core treatment targets, the proposed study has 3 overarching research aims and corresponding hypotheses.

Aim 1 is to examine momentary associations between caregiving practices and core treatment targets. Hypothesis 1 is that momentary caregiver reports of caregiver support, caregiver monitoring, substance use communication quality, and alcohol-specific caregiving practices will be positively associated with adolescent reports of motivation to reduce or stop drinking and positive affect and negatively associated with alcohol craving, alcohol use, and internalizing and externalizing symptoms. Momentary caregiver reports of caregiver-adolescent conflict will be negatively associated with adolescent reports of motivation to reduce or stop drinking and positive affect and positively associated with alcohol craving, alcohol use, and internalizing and externalizing symptoms.

Aim 2 is to examine changes in the strength of the associations between caregiving practices and core treatment targets over the course of treatment, consistent with treatment perspectives arguing that skill use should be more effective across time [69]. Hypothesis 2 is that aim 1 associations for caregiver support, caregiver monitoring, substance use communication quality, and alcohol-specific caregiving practices will become stronger over the course of AUD+CODs treatment as caregivers and adolescents learn and practice skills over the course of treatment. Aim 1 associations for caregiver-adolescent conflict will become weaker as caregivers and adolescents learn and practice skills to better manage conflict in the moment.

Aim 3 is to examine whether the use of family-focused therapy techniques are associated with changes in the use of caregiving practices. Hypothesis 3 is that caregiver reports of learning or practicing family-focused therapy techniques in therapy will be prospectively positively associated with increases in caregiver support, caregiver monitoring, substance use communication quality, and alcohol-specific caregiving practices and with decreases in caregiver-adolescent conflict.

The adolescents eligible for Project Momentary Assessment of Parenting Practices (MAPP) will be those who (1) are aged between 13 and 18 years; (2) meet the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, diagnostic criteria for AUD; (3) are enrolled in outpatient treatment at the time of recruitment; and (4) have a caregiver willing to participate in the study. Considering that >80% of adolescents with an AUD presenting to treatment have a co-occurring mental health condition [2,3]—and consistent with current dimensional conceptualizations of psychopathology [70]—a formal diagnosis of a co-occurring condition will not be required for eligibility. Exclusion criteria will include (1) adolescents exhibiting acute psychosis and (2) adolescents or caregivers who are unable to read or comprehend the study forms in English or Spanish (eg, consent procedures).

Seventy-five dyads will be recruited from 5 community-based mental health clinics treating adolescents with co-occurring conditions in Massachusetts. On average, these clinics have 9 (SD 7.36, range 3-20) therapists working with adolescents with co-occurring conditions and have the capacity to see 159 (SD 123.74, range 50-350) adolescents per year. Across these mental health clinics, several treatments are offered, including the adolescent community reinforcement approach (A-CRA; all 5 clinics), trauma-focused cognitive behavioral therapy (1 clinic), eye movement desensitization and reprocessing (1 clinic), and home-based services (1 clinic). A-CRA is the preferred treatment model endorsed by the Office of Youth and Young Adult Services in the Bureau of Substance Addiction Services, Massachusetts. Caregiver involvement varies across settings and treatments, with examples of caregiver involvement being weekly family therapy sessions for home-based services and 2 formal caregiver sessions in A-CRA.

Before recruiting families, the study team will provide presentations to each partnering clinic about the study so that clinicians can inform their clients about the study and answer any questions they may have. At each clinic, recruitment will occur using flyers that will be distributed to all new adolescent clients and their caregivers. In line with recent recommendations for facilitating compliance and building rapport in EMA research with adolescents [71], the flyers will also briefly introduce the Project MAPP research team and include pictures of team members. Interested families will be able to notify study staff if they are interested in participating and would like to complete an eligibility questionnaire through a QR code on the flyer. This method of recruitment was selected so that no clinic needs to provide the study team with any protected health information.

Figure 1 provides an overview of the study components of Project MAPP. Project MAPP consists of two components: (1) an eligibility screening and baseline assessment and (2) an EMA burst design (Figure 2). All components of Project MAPP will be conducted remotely via Health Insurance Portability and Accountability Act (HIPAA)–compliant video communication platforms.

Figure 3 provides an overview of the procedures for the baseline study visit. Eligible dyads interested in participating in Project MAPP will be scheduled for their baseline assessment session during the second week of their treatment. If dyads are unable to schedule the study visits during the second week of treatment, they will be scheduled as close to the second week of treatment as possible. The baseline session will be offered in 2 formats: In format 1, dyads complete consent, assent, and parental permission forms; complete a formal assessment for AUD; complete baseline assessment measures, including a timeline followback (TLFB) interview for the adolescent; and are trained in EMA study procedures. In format 2, adolescents and caregivers complete all components of the baseline other than the baseline assessments, which they can complete at their own convenience within 72 hours of the baseline session.

These 2 options are provided to accommodate caregiver and adolescent preferences and time constraints.

For both formats, after completing the consent (for caregivers aged ≥18 y and adolescents aged 18 y), parental permission (for caregivers of adolescents aged <18 y), and assent (for adolescents aged <18 y) procedures, adolescents and caregivers will each complete the Schedule for Affective Disorders and Schizophrenia for School-Aged Children in separate web-based breakout rooms to formally assess for adolescent AUD [73]. Dyads will be notified that they are not eligible for Project MAPP if the adolescent does not meet the diagnostic criteria for AUD, and they will be compensated for their time. Eligible adolescents will complete a 30-day TLFB assessing their alcohol and other substance use (eg, cannabis and nicotine) in the past 30 days. The TLFB is a gold standard assessment for capturing alcohol and other substance use behaviors [74,75].

In format 1 of the baseline visit, caregivers and adolescents will next complete an assessment battery measuring demographic information (eg, age, gender, and family composition) as well as validated self-report measures of the EMA parenting domains and the treatment outcomes of interest (Table 1, column 3). A strength of the parenting measures selected for this study is that they contain parent- and adolescent-report versions, attenuating concerns regarding demand characteristics in parenting measures [76]. Internalizing and externalizing mental health symptoms will be measured continuously using the Child Behavior Checklist and the Youth Self-Report, which are measures from the Achenbach System of Empirically Based Assessment (Table 1) [77,78]. In format 2, adolescents and caregivers will complete these measures on their own time within 72 hours of completing the baseline visit.

For both formats 1 and 2, the last component of the baseline study visit consists of EMA training. Participants will be instructed to download the EMA software, created using a MetricWire platform (MetricWire Inc), and register for Project MAPP. Research assistants will use an EMA training manual developed for adolescent alcohol and other drug EMA research [102,103]. First, adolescents and caregivers are walked through a MetricWire demonstration survey. Research assistants walk participants through completing question types that they will encounter (eg, check boxes, typed responses, and Likert-scale items) and answer any questions that participants may have. Second, participants are led through completing a problem survey report. These surveys are intended for participants to use when they experience any difficulties with the MetricWire app. In addition, research assistants demonstrate the messaging feature of MetricWire where participants can contact any study staff if they have questions or issues with the EMA software. Third, to facilitate the reporting of standard drinks of alcohol, research assistants walk adolescents through a standard drink training on MetricWire. This survey, which is available throughout the study, includes detailed information on what constitutes a standard drink of alcohol.

As the last step of the baseline visit, research assistants will walk participants through a series of procedures intended to maximize the ease of completing the EMA protocol and EMA compliance. These procedures consist of study staff reviewing times where participants are unable to complete EMA (eg, during work or extracurricular activities) and informing participants that they do not have to complete EMA surveys during these periods and that these surveys will not count against their compliance. Furthermore, participants are instructed not to complete EMA surveys when it would be unsafe or uncomfortable (eg, when driving or when a caregiver or peer is looking over their shoulder). Research assistants also inform participants that they will receive push notification reminders, SMS text messages, and telephone calls to facilitate their EMA compliance (greater detail on these procedures is provided in the subsections that follow). Finally, research assistants will inform participants that they will be in frequent contact with them to help facilitate EMA compliance and address any questions that may arise [104]. EMA compliance will be calculated in two ways: (1) total accountable EMA compliance (total number of EMA surveys completed/total number of EMA accountable surveys—accountable surveys refer to reports where an adolescent or caregiver could have completed the survey but did not; this will be the primary means of examining EMA compliance) and (2) total EMA compliance (total number of EMA surveys completed/total number of EMA surveys).

High levels of EMA compliance are essential to obtaining high-quality data that can rigorously answer the questions outlined in this protocol. The proposed study will implement the following procedures to maximize EMA compliance: (1) As noted in the Baseline Study Visit subsection, the dyads will be trained during the baseline visit by research assistants on using MetricWire and review the times when participants are unable to answer prompts and should not answer prompts (eg, when driving). (2) Research assistants will monitor the EMA compliance of all participants daily. (3) Participants will be contacted frequently through the MetricWire messaging feature or SMS text message, based on their preference, to receive EMA compliance feedback. (4) On the morning of the fourth day of each 7-day EMA week period, participants will be sent a message noting whether they are on pace to earn a US $10 bonus for maintaining a weekly compliance rate of ≥75%. For participants who are below the 75% threshold, they will be informed of the number of surveys they need to complete during the remainder of the week to earn the US $10 compliance bonus. (5) At the end of each study week, research assistants will reach out to participants by text message or telephone call, depending on participant preference, to review the participants’ EMA compliance from the last week, discuss missed surveys and whether there were any reasons for missed surveys, and inquire whether there were any issues that occurred with the EMA software. (6) Information from procedures 3 to 5 will be shared with all Project MAPP staff on an internal messaging service (eg, Google Chat or Slack). This will help ensure that the entire study team is aware of any emerging issues with EMA compliance and that plans can be discussed early on to address any compliance issues. In addition, this information will be reviewed at weekly Project MAPP study meetings.

Considering that aim 3 seeks to examine whether the use of family-focused therapy techniques are associated with changes in caregiving practices, Project MAPP requires a means of collecting information on what is occurring in weekly therapy sessions. The mental health update report will collect information, weekly, on the skills that adolescents and caregivers are learning in outpatient therapy. These reports will capture skills learned or practiced in therapy, caregiver involvement in the session, therapeutic alliance, and the degree to which caregivers support their adolescent’s mental health and treatment goals (refer to Multimedia Appendix 1 for a complete list of items). The skills that caregivers and adolescents learn or practice in session will be assessed using the Therapy Process Observation Coding Scale–Self-Reported Therapist Intervention Fidelity for Youth [111], modified for caregiver and adolescent self-reports. This measure captures adolescent-oriented skills (eg, emotion regulation, managing negative thinking, and behavioral activation) and caregiver and family-oriented skills (eg, communication skills, consistent rules, and consequences). When adolescents or caregivers report that an adolescent missed their weekly therapy session, they will complete the mental health update report, which will include specific items assessing why they missed their therapy visit along with most of the constructs assessed in the mental health update report for attended sessions (Multimedia Appendix 1).

Unfortunately, only approximately 4 in 10 individuals of any age who present to substance use services complete treatment [112]. This number may be lower for adolescents [113]; for example, whereas the average outpatient treatment protocol is 16.5 sessions, youth only attend 3.9 sessions on average [114]. Acknowledging the high dropout rates in adolescent outpatient treatment, this study will retain dyads where the adolescent drops out of treatment. If an adolescent or caregiver reports that the adolescent dropped out of treatment, they will complete a version of the mental health update report assessing the reasons for discontinuing treatment, alternative recovery supports that the adolescent is receiving, caregiver support of the adolescent’s mental health and substance use, the use of skills learned from therapy or other recovery supports, and experiences of minority stress (refer to Multimedia Appendix 1 for a complete list of items). Dyads where the adolescent successfully completes treatment will also complete the same version of the mental health update report but not be administered questions regarding the reasons for discontinuing treatment (Multimedia Appendix 1).

The mental health update report will be available every Saturday starting at 4 AM. Reminders for the mental health update will follow the same schedule as those for the morning reports. If the participant does not complete the mental health update on Saturday, the mental health update will be available to complete on Sunday.

As part of a National Institute on Alcohol Abuse and Alcoholism K99 award (K99AA030030), a pilot trial was conducted to assess the acceptability and feasibility of conducting a 7-week EMA study with caregiver-adolescent dyads, in which the adolescent is receiving AUD+CODs treatment in an intensive outpatient treatment program. The acceptability and feasibility benchmarks for the K99 phase are presented in Table 2. Whereas some core benchmarks were met during the K99 pilot study (ie, retention rate ≥80% for caregivers), most of the benchmarks were not met. These included a targeted enrollment rate of ≥75% versus the actual dyad enrollment rate of 62% (16/26), retaining 75% (12/16) of the adolescents versus an 80% target retention rate, EMA random prompt completion rate of ≥80% versus adolescents completing 47.6% (772/1622) and caregivers completing 70.76% (1331/1881) of the random prompts, and a morning report compliance rate of ≥80% versus adolescents completing 55.5% (361/651) and caregivers completing 91.3% (601/658) of the morning reports. The proposed study, which would be funded through the R00 portion of this National Institute on Alcohol Abuse and Alcoholism Pathways to Independence R00 award, contains modifications based on the information learned from the K99 phase, which are discussed in detail in the Methods section (Table 2; Multimedia Appendix 2).

No data have been collected for the proposed study. Data collection is anticipated to begin in December of 2024. Project MAPP will take place over the course of 3 years. Study startup tasks (eg, training study staff) will occur during months 1 to 3, and data collection will occur during months 4 to 30. On the basis of this recruitment time frame, the study will need to enroll approximately 3 dyads per month to meet the recruitment goal of 75 dyads. Data checks to ensure the quality of the data will be performed weekly throughout data collection as will data cleaning procedures. Beginning in month 31, by which time most of the data collection will have occurred, data preparation and analysis will be performed for the primary aims of Project MAPP.

Project MAPP seeks to advance our knowledge of how caregiver involvement in AUD+CODs treatment improves outcomes by examining (1) momentary associations between caregiving practices and core treatment targets (eg, motivation to reduce or stop drinking and alcohol craving), (2) changes in the strength of the associations in aim 1 to understand whether caregiving practices become more effective over the course of therapy, and (3) whether learning and practicing family-focused therapy techniques in session are associated with changes in the use of caregiving practices in real time in daily life outside of therapy. Information from this study will hopefully aid clinical practice by informing which caregiving practices positively change adolescent core treatment targets on a more refined and clinically aligned timescale. In contrast to existing caregiving research that indicates that greater monitoring before treatment is associated with improved outcomes [125], the proposed study will be able to determine whether caregiver use of monitoring strategies in the moment protects their children from using substances. Data from this study will be on a timescale most relevant to caregivers and clinicians who want to know what caregivers can do during a specific moment in time to promote well-being and mitigate risk for their teens when in treatment [126,127].

In addition to testing the primary aims of Project MAPP, we intend to use the data from this study to advance knowledge on associations across multiple timescales (eg, the association between learning or practicing skills and then using them outside of therapy). Theoretical and empirical papers in psychological science have raised awareness of the importance of timescale [54,128]. Aims 1 to 3 of Project MAPP focus on momentary (aims 1 and 2) and proximal (ie, weekly; aim 3) associations. It remains an open empirical question as to whether these are the most appropriate timescales to better understand caregiving processes during treatment; for example, aim 3 will examine whether caregiver skills learned in therapy are associated with the use of caregiving practices in the week after therapy. Alternative timescales (eg, lags of half a week, 2 weeks, 1 month, and the entire study period) may better capture the relationship between learning or practicing skills in therapy and implementing them outside of therapy.

Aims 1 to 3 emphasize the unidirectional role of caregiving behaviors on adolescent core treatment outcomes. A large body of research demonstrates that caregiving and adolescent behavior influence one another reciprocally across time [129-131]. Changes in adolescent core treatment outcomes during AUD+CODs treatment may be a function of improvements in adolescent behavior, which result in improved caregiving behaviors. In turn, these caregiving improvements facilitate positive changes in adolescent treatment outcomes. Furthermore, caregiver involvement in treatment may facilitate treatment outcomes by enhancing the effectiveness of skills that adolescents learn during therapy. Learning or practicing specific skills in therapy, such as emotion regulation strategies, may be more strongly associated with reduced mental health symptoms among adolescents whose caregivers monitor their behavior or are responsive to them. Project MAPP is designed to test alternative mechanisms of positive outcomes, such as these examples, in addition to the primary aims of the study.

Although Project MAPP represents an important extension to caregiving research during AUD+CODs treatment, the proposed study will still contain several limitations. First, the study will recruit only a single caregiver per adolescent. Caregiving research has demonstrated unique effects of separate caregivers (eg, unique effects of mothers vs fathers [132]). By not including all caregivers responsible for the adolescent, Project MAPP will miss important information regarding what some caregivers are doing during AUD+CODs treatment. Second, caregivers and adolescents will be repeatedly asked questions about caregiving practices, skill use, treatment motivation, and substance use. There is the possibility for reactivity—that repeated assessments systematically alter caregiver and adolescent behaviors or attitudes [133,134]. Without the inclusion of a control condition (ie, a group of dyads in which the adolescents are receiving treatment who do not receive EMA), Project MAPP will not be able to determine the extent of reactivity to the EMA burst design protocol. Third, given that adolescents and their caregivers will be participating in outpatient treatment, which can last more than a year, Project MAPP will not be able to sample all of an adolescent’s time in AUD+CODs treatment. Finally, there are many caregiving practices associated with adolescent substance use [39]. Capturing all these practices via EMA would not be feasible. The selection of caregiving practices was based on the core elements of caregiver-involved interventions for adolescents [8,32] and EMA caregiving research from community samples [53,135].