A Randomized Controlled Trial on Brief Behavioral Parent Training

1. INTRODUCTION AND RATIONALE Behavioral difficulties, including oppositional, defiant, aggressive, hyperactive, impulsive, and inattentive behavior, are one of the most common reasons for referral to mental healthcare among children and adolescents (Antshel, 2015; Polanczyk et al., 2015). When left untreated, problems may exacerbate and put children at risk of adverse outcomes later on in life, such as school dropout, delinquency, substance use and depression (Franke et al., 2018; Reef et al., 2011). Ideally, treatment should prevent this escalation of problems, reduce the need for long and intensive treatments, and lower societal costs (Drummond et al., 2015). Behavioral parent training is the first psychosocial treatment of choice for reducing behavioral difficulties in preschool and school aged children (Akwa GGZ, 2019; Brestan \& Eyberg, 1998; Eyberg et al., 2008; Kaminski \& Claussen, 2017; National Institute for Health and Care Excellence \[NICE\], 2017; National Institute for Health and Care Excellence \[NICE\], 2019; Nederlandse Vereniging voor Psychiatrie, 2013). However, its use in clinical practice is limited by a scarcity of certified therapists, long waiting lists, and the typically long and sometimes perceived as rigid nature of behavioral parent training programs, which may lead to parents not starting or finishing behavioral parent training (Chacko et al., 2016; Evans et al., 2018; Weisenmuller \& Hilton, 2021). To reduce these barriers, this study evaluates a brief and individually tailored behavioral parent training program for children with behavioral difficulties (the "PAINT-GGZ" program, developed by the Psychosocial ADHD \& disruptive behavior INTerventions \[PAINT\] consortium). The brief training includes optional booster sessions for families who need additional support. The trial will generate knowledge on the (cost)effectiveness of brief parent training with optional booster sessions for children with behavioral difficulties. Brief programs are scarce, mostly do not provide booster sessions, and are mainly evaluated in prevention settings (Kolko \& Lindheim, 2014; Tully \& Hunt, 2015). This project fills these knowledge gaps. 2. OBJECTIVES The first objective of this study is to examine the short- and longer-term (i.e., one year) effectiveness of the brief and individually tailored behavioral parent training program with optional booster sessions compared to care as usual (CAU) on daily measured child behavioral difficulties (primary outcome) and on a range of secondary outcomes (see Outcomes). Additionally, we will explore whether a number of general demographic and clinical characteristics, parental attachment, parental psychopathology, parental reward responsivity, and reward and punishment sensitivity of the child moderate the short- and longer-term intervention effects on our primary outcome. The second objective is to investigate the cost-effectiveness of the program compared to CAU. As we assess cost-effectiveness based on measures at baseline and one year after the intervention, we can only draw conclusions about the cost-effectiveness on the longer term. 3. STUDY DESIGN AND SETTING In this two-arm multi-center randomized controlled trial, parents of children who experience behavioral difficulties in the home setting and were referred to a child mental healthcare center, are randomly assigned (1:1 ratio) to either (a) three sessions of brief behavioral parent training with optional booster sessions, or (b) CAU, as regularly provided by the involved mental healthcare centers. There are no restrictions regarding the type (e.g., psychoeducation, parent counseling and support, individual child therapy, medication) or duration of CAU, only the brief parent training investigated in this trial is not allowed. Both parents and children in the brief behavioral parent training group are not allowed to receive CAU until the first posttreatment assessment (T1). After that, CAU is allowed and parents can also make use of booster sessions with the therapist who provided the training. To ensure our findings reflect the real-world impact of brief behavioral parent training (i.e., to test effectiveness rather than efficacy), the trial is fully embedded in routine mental healthcare. The participating centers are academic and non-academic youth mental healthcare centers of various sizes. Therapists who provide the brief behavioral parent training have a master's degree in psychology or child, family and education studies and have had postmaster education in cognitive behavioral therapy (at least 50 hours) or have a registered higher vocational education degree as social worker in cognitive behavioral therapy. The therapists are trained in the brief parent training by supervisors qualified in cognitive behavior therapy during a one-day training of seven hours, in which therapists practice the parent training by roleplay and evaluate on their performance. Therapists also receive weekly supervision by these qualified supervisors when providing the brief parent training. Additionally, the supervisors provide feedback to each therapist on the audio-taped first session regarding protocol fidelity, intervention integrity, and the process of providing the training (e.g., engaging parents, didactic skills). There are no in- or exclusion criteria for clinicians in the CAU condition. To prevent contamination of the two research arms, therapists never provide both the brief behavioral parent training and CAU to parents participating in this study and are instructed not to share information about the contents of the brief training. 4. TIME FRAME Year 1: start recruitment, screening and inclusion of parents and children, start running interventions Years 2-4: ongoing contacts with mental health care organizations; recruitment, screening, and inclusion of children; run interventions and conduct measurements; supervision of therapists; fidelity checks; other project management activities; data-analysis. The last inclusion of patients is planned in year 4. Year 5: further data-analysis, dissemination of results, implementation activities. For this trial we have installed a parent advisory board (four members, one of them represents "Balans", a Dutch association for parents of children with developmental problems). The board is involved in this trial since the grant application. They provided feedback and ideas for the current project and meet biannually in the first and last year of the project and annually in years two, three, and four. In the last year, they will be asked to give their ideas about the interpretation of the results. The members of the board receive reimbursements for all meetings. 5. STUDY POPULATION The sample will consist of parents of children (2-12) with behavioral difficulties who have been referred to one of the participating Dutch mental healthcare centers. Based on our previous study that found medium-sized short term effects on daily ratings of behavioral difficulties compared to waitlist (range of Cohen's d = .35 to d = .66, Hornstra et al., 2021) and studies showing medium-sized parenting intervention effects on behavioral difficulties using parent-report rating scales (Bakker et al., 2017; Daley et al., 2014; McCart et al., 2006; Van Aar et al., 2017; Van den Hoofdakker et al., 2007), we estimate the effect of our brief behavioral parent training program as compared to CAU to be at least small (d = .25). A power analysis has been performed using G\*Power software (Faul et al., 2007). Based on an effect size of d = .25 (f = .125), two groups and three repeated measures (T0, T1, T2 for short term effectiveness), with r = .60 between-measurement relation, a power of .80 and α = 0.05, we need at least 42 participants per group, resulting in a total of 84 participants. Given that the data are clustered, we increased the sample size with 10% (Twisk, 2010), resulting in a total of 93 participants. This is in line with the number of participants in similar studies comparing behavioral treatments to CAU (e.g., Van den Hoofdakker et al., 2007). 6. TREATMENT OF SUBJECTS During a period of one year, parents in the parent training condition will participate in a brief three-session parent training program with optional additional booster sessions and/or CAU after the parent training. The first two sessions will take 120 minutes and the third session will take one hour. These three sessions will be planned in three to four consecutive weeks. In between sessions, parents are asked to read some materials and practice the behavioral techniques they have learned with their child at home. After the brief parent training program, care as usual and/or additional booster sessions can be requested if deemed necessary by the parents or therapist. The additional booster sessions will take one hour each and can be requested maximally once every four weeks up to one year after the brief parent training program. The individual, three-session parent training program uses the behavioral techniques that were identified as effective for behavioral difficulties in our microtrial (Hornstra et al., 2021) and the broader literature (Kaminski et al., 2008; Leijten et al., 2019; Leijten et al., 2021). The program combines the stimulus control and contingency management techniques that, separately, reduced behavioral difficulties in the previous microtrial. Importantly, using functional analyses, the techniques are tailored to the specific problem behaviors that individual parents experience with their child. This individualized approach perfectly fits with the advice of our parent advisory board, that indicated that targeting individual parental needs is crucial for parents. To prevent relapse and reduce the need for more intensive treatments (including psychotropic medication, such as methylphenidate or antipsychotics), a booster session on maintenance training can be added at several time points (maximum once every four weeks) for parents wishing to receive additional support. Our parent advisory board considered the optional booster sessions as very important and helpful, as some parents may experience relinquishing of their parenting skills after finishing parent training, for example in stressful times (see also Furlong \& McGilloway, 2014). Parents in the control condition will receive care as usual, which in Dutch mental health care typically consists of psychoeducation, parent counseling and support, medication, (longer) parent training programs, family therapy, and/or cognitive behavior therapy or other therapy for the child. The load imposed on these parents depends on the type of care they and/or their child receive. This usual care will also be provided to parents who do not participate in this study. The treatment in the control condition does therefore not impose any additional load compared to not participating in this study. 7. METHODS Parents will provide information on children's age and gender, on their own and children's ethnicity, educational background, and household composition, and on children's previous and ongoing pharmacological and non-pharmacological treatments, medical conditions, and clinical diagnoses (e.g., ADHD, oppositional defiant disorder \[ODD\], conduct disorder \[CD\]). PRIMARY AND SECONDARY STUDY PARAMETERS The primary outcome of the study is the severity of four individual target behavioral difficulties that parents want to address in the training. The secondary outcomes of the study are parent-reported behavioral difficulties, parent-reported child well-being, parent-reported parenting behaviors, masked audio records of mealtime routines to measure parent and child behavior, parent-reported parenting stress, parent-reported parenting self-efficacy, parent-reported quality of the parent-child relation, consumption and costs of mental health care and health state utility values. The investigators will furthermore explore parents' and therapists' evaluations of the brief parent training program. All outcomes are described in more detail below. MODERATORS In addition to the general demographic and clinical characteristics and some baseline characteristics such as the severity of behavioral problems, the candidate moderators of the study are parental attachment as measured with the 12-item version of the Revised Experiences in Close Relationship (ECR-R) questionnaire (Wei et al., 2007), parental psychopathology as measured with the 25 items of the Strengths and Difficulties questionnaire (SDQ) for parents (https://www.sdqinfo.org/Adult/), parental reward responsivity as measured with the 8-item Reward Responsiveness (RR) questionnaire (Van den Berg et al., 2010), and parent-reported child reward responsivity and punishment sensitivity as measured with 22 items of the Sensitivity to Punishment and Sensitivity to Reward Questionnaire for Children (SPSRQ-C) (Luman et al., 2012). STUDY PROCEDURES Parents of children with behavioral difficulties problems who have been referred to one of the participating Dutch mental healthcare centers will be recruited by clinicians within these centers. After the diagnostic assessment of their child and before any treatment has started, the clinicians involved in the diagnostic phase will inform parents about the study and hand out the information letter. Parents who are interested can give their contact information to the clinicians. The researchers will use this contact information to further inform parents about the research through telephone calls. If parents want to participate, the researchers will ask them to provide informed consent. Both legal caretakers (if present) of the child are asked to provide consent. Parents will have at least 14 days to decide upon participation and ask the researchers any questions. If needed, parents can ask for more time to decide upon participation. Parents can leave the study at any time for any reason if they wish to do so without any consequences. To measure the primary and secondary outcomes and candidate moderators of treatment response, parents in both treatment arms will be asked to make audiotapes of their mealtime routines, fill out questionnaires, and answer short daily phone calls (five days in a row) on four measurement occasions. Completing the questionnaires (including those measuring moderator variables) takes 30 to 45 minutes per measurement occasion, making the audiotapes takes about 30 minutes per measurement occasion (i.e., two recordings of 15 minutes each), and making the phone calls takes two to three minutes each. Furthermore, parents in the behavioral parent training arm fill out the ECBI-I questionnaire at the beginning of and two weeks after each booster session, which takes a couple of minutes. The measurements take place at baseline before randomization (T0), one week after the third session for parents in the brief behavioral parent training arm and eight weeks after T0 for parents in the CAU arm (first posttreatment measurement, T1), six months after T1 (second posttreatment measurement, T2) and twelve months after T1 (third posttreatment measurement, T3). 8. STATISTICAL ANALYSIS Data will be analyzed on an intention-to-treat basis. To examine the effects of the intervention on the primary and secondary outcomes, we will conduct multilevel analyses (mixed modeling), taking missing data into account (Twisk et al., 2013). Depending on the fit of the models, up to four hierarchical levels will be distinguished: outcomes (level 1), nested within children (level 2), nested within therapists (level 3), nested within healthcare centers (level 4). Random intercepts at the subject level, therapist level and healthcare center level will be included only if the Likelihood Ratio Test shows a significant improvement of the model fit. Time will be included as short-term (T1) and longer-term (T2, T3) within-subjects factors and condition (brief parent training vs. CAU) as between-subjects factor. The interaction between time and condition will be examined to interpret changes over time between conditions. To examine whether the candidate moderator variables moderate the intervention effects, these variables will also be added as interaction effects to the multilevel analyses. To assess the cost-effectiveness of the brief behavioral parent training compared to CAU, we will calculate the Incremental Cost Effectiveness Ratio (ICER) from a societal perspective. We will perform both a cost-utility analysis (based on QALYs) and cost-effectiveness analyses (based on the daily ratings of behavioral difficulties and based on ECBI-I scores). An ICER is calculated by dividing the difference in the total costs by the difference in the total effect (either QALYs or the daily ratings of behavioral difficulties and ECBI-I scores, respectively). In these analyses, the costs of healthcare will be used and interpolated to the previous 12 months. We will compare each ICER based on relevant willingness to pay thresholds. Bootstrapping will be used to calculate the reliability of our estimates. 9. ETHICAL CONSIDERATIONS Parents in the parent training condition will receive a brief parent training program that uses the behavioral techniques that were identified as effective for behavioral difficulties in our recent microtrial (Hornstra et al., 2021) and the broader literature. The risk of any physical or mental harm of the study is very low and not to be expected higher than what may be expected in daily life of the parents and their children. We will exclusively use conventional methods and interventions and the study is embedded in mental healthcare settings. The children of participating parents have been referred for behavioral difficulties and parents who visit the participating clinical institutions seek treatment for these problems. Adverse reactions of the parent training will be monitored (Allan \& Chacko, 2018; Linden, 2013). After the parent training, parents wishing to receive additional guidance can request extra booster sessions and/or care as usual, depending on their needs. Tailoring the care that will be provided to parents' needs will minimize the chance of any physical or mental harm. The children with behavioral difficulties (i.e., minors) will not be directly involved in the study. Specifically, they will not be present during the brief parent training sessions and all measures will be parental measures. Furthermore, diverse populations and participants will not be denied treatment. To debrief participants, after the study is finished parents who wish to learn about the results of the study can request this information from the researchers. We inform parents about the possibility to ask for this information in the information letter. 10. ADMINISTRATIVE ASPECTS, MONITORING AND PUBLICATION HANDLING AND STORAGE OF DATA AND DOCUMENTS Digital data, both raw and processed data are stored at a study specific secured folder within the network of Accare, the institution at which the research is carried out. This study specific folder can only be accessed by authorized personnel, who are involved in this study. All questionnaire data and processed audio files are stored in one, pseudonymized file, that is accompanied by a 'readme' text file that contains a code book explaining the meaning of all variables. A separate logbook-file will be created documenting all decisions that are made during the process from raw to processed data. Raw data containing identifiable information is kept strictly separate from the processed data and can only be accessed by two of the supervisors (BJvdH., TJD), research assistants and PhD candidate (RSvD). After the research project will be completed (i.e., data collection, data analysis and publishing of research articles), all digital data will be transferred to a study specific folder for long-term storage. The folder can only be accessed by authorized personnel. All research data will be stored for at least 15 years after the data collection has been completed. Upon request, the processed, pseudonymized data will be available for fellow researchers and made available (restricted access) for use in future research projects, to which all participants have consented before enrolling in the study. Requests for re-use of data will be evaluated to check whether the research question falls within the scope of the informed consent. MONITORING AND QUALITY ASSURANCE Intervention fidelity (i.e., the percentage of items covered in each session) is assessed with a fidelity checklist using an adapted version of the procedures of Abikoff and colleagues (Abikoff et al., 2013; Abikoff et al., 2015). This implies that all intervention sessions (including the booster sessions) are audiotaped. The audiotapes of 20% randomly selected sessions will be scored on intervention fidelity by independent evaluators. After each session, therapists also have to complete a fidelity checklist in which they are asked which items they covered. Therapists also receive weekly supervision by these qualified supervisors when providing the brief parent training. Additionally, the supervisors provide feedback to each therapist on the audio-taped first session regarding protocol fidelity, intervention integrity, and the process of providing the training (e.g., engaging parents, didactic skills).