Randomized Controlled Trial of Comet Via the Internet or in Group Format.

Parent Training Targeting Externalizing Behaviors in Children in Primary Care: A Randomized Non-inferiority Study of the Comet-program Delivered Via the Internet or in Group Format.

The Comet-program is a Swedish parent training program developed to target externalizing behaviors in children between 3-11 years. The program is normally delivered in group format in primary and specialized care and has already been evaluated in several studies. The internet-based version of the program has also been evaluated separately, but in this study the two formats will be directly compared in a randomized non-inferiority study. Parent ratings, child ratings as well as blinded clinical assessments will be conducted before the interventions, after the interventions (2-3 month after start of intervention), and at follow-up (12 mont after first assessment). The primary research question is: Will the internet-based format of Comet be at least as effective as the group format in reducing externalizing behaviors? Secondary research questions concern possible differential effects of the two formats on cost-effectiveness, parenting behaviors, parent mental health, applicability and consumer satisfaction, and the children's quality of life.

Background: The Comet-program is a Swedish parent training program developed to target externalizing behaviors in children between 3-11 years. The program content is behaviorally oriented and the emphasis lies on increasing parenting behaviors that reinforces pro-social behaviors in children, as well as limiting the reinforcement of anti-social behaviors. The program is manualized and normally delivered in group format with parents to 6-8 children in each group. The group sessions (11 sessions á 2.5 hours) are led by two practitioners and include information, role-playing, and home-work exercises. The group format has been evaluated in two large randomized controlled studies. The effect sizes on child externalizing behaviors have been medium to large (Cohen´s d). Because of the evidence supporting Comet and other research supporting the effectiveness of similar behavioral parenting programs, Comet has been implemented in primary and specialized care in Sweden. A challenge in primary care is however the implementation costs related to the program (e.g., training of group leaders and the time each group leader need to invest in each patient). Therefore, Comet (or similar programs) are in practice used to a small extent in primary care, despite clear recommendations to offer such treatments for externalizing behaviors. A possible solution would be to offer an internet-based version of the program, which in two large randomized controlled trials have shown medium effect sizes on externalizing behaviors in children. In the internet-based version, the patients (parents) work online with seven modules over a period of nine weeks. Every module contains information, video-clips, quizzes and homework exercises. Every week the patient communicates with a therapist, that gives brief guidance and feedback online. The internet-based format of the program has however been implemented at a very small scale in regular care. One reason is probably reluctance on part of practitioners as well as patients in replacing face-to-face contact with online-contact. The management system used in health-care in Sweden is another obstacle, since only face-to-face contacts are included. Hence, the clinics are not financially compensated for any patient contacts online. To overcome these obstacles, the present study will compare the regular Comet program in group format to a modified version of the internet-based version. In the modified version, the online contact between the patient and the therapist will be replaced with three individual face-to-face sessions (45 min each). This will reduce the total amount of time invested per patient to 135 mins, compared to 410-550 mins per patient in the group format. Research questions: The primary research question in the study is: (1) Will the internet-based version of Comet be at least as effective as the group format in reducing externalizing behaviors in children? The secondary research questions will be to investigate if the two formats of Comet will differ in terms of: (2) cost effectiveness, (3) consumer satisfaction/applicability for patients as well as practitioners, (4) effects on parent behaviors and mental health, (5) effects on the quality of the relationship between parents and children, (6) effects on the quality of life experienced by the children, and (7) if the application of parenting strategies included in Comet mediates change in child externalizing behaviors. Method: To investigate these research questions, parents seeking help in primary care clinics for externalizing behaviors in their child will be recruited to the study. Six clinics will be involved in recruitment of participants and offering the interventions. Clinical psychologists on every clinic will therefore be trained in both formats of the Comet program. Enrolled participants will be randomized at each clinic to either take part in the group format or the internet-based format of Comet. The participants randomized to the latter format will start their treatment as soon as they have completed the first assessment, while participants in the group format will have to wait for the scheduled group to start (generally within 1-3 months after enrollment and first assessment). The post assessment will take place when participants have completed the intervention and the follow-up assessment will be 12 months after the first assessment point. In addition, parents in each condition will answer a few questions on a weekly basis during their interventions. Assessments at pre, post and follow-up will include parent ratings of child behaviors, parent behaviors, parent mental health, and consumer satisfaction/applicability (only at post assessment). At pre and post assessment blinded clinicians will also make a diagnostic assessment of Oppositional Defiant Disorder (ODD) and Attention Deficit/Hyperactivity Disorder (ADHD) symptoms of the participating children. When the clinician meets the child for these assessments, the child will also be answering questions (structured interview) regarding their experienced quality of life. Finally, the psychologist involved in the treatments will continuously take notes in a log-book for each participant. The log-book will contain information on parent attendance, adherence and adverse events. A selection of the treating psychologists will also be interviewed to assess their experienced applicability of the two formats of Comet. The weekly questions will measure child externalizing behavior and the extent to which the parents have completed homework exercises (parenting strategies). Since the primary research question is if the internet-based version of Comet is at least as effective as the group format, the power calculation was based on non-inferiority design. The first step is to decide the non-inferiority margin, which is the least acceptable difference between the conditions in order to conclude that the internet-based version is non-inferior. A conventional method for deciding this margin is to use 50% of the effect sizes found in previous studies of the already established intervention. When applying this method for the primary outcome measure, the calculation showed that 134 participants were necessary to include in order to achieve a power of 0.80. In other words, if there is truly no difference between the group format and the internet-based format, then 134 patients are required to be 80% sure that the lower limit of a one-sided 95% confidence interval (or equivalently a 90% two-sided confidence interval) will be above the non-inferiority limit. Most research question will be analyzed using methods from non-inferiority design, i.e., analyzing the difference between adjusted means at post and follow-up with regard to the non-inferiority margin. Primarily, these analyses will be performed on participants who completed the interventions and assessments (per protocol), since intention-to-treat analyses reduces power and increase the risk falsely concluding non-inferiority. For the second research question (cost-effectiveness), the clinical significance in each condition will be calculated and compared (i.e., the proportion of participants that have "recovered", "improved", are "unchanged" or have "deteriorated). Thereafter, the cost per patient for each treatment will be calculated and used to achieve an estimate of cost per recovered/improved participant. This method has previously been applied in a cost-effectiveness analysis of Comet and other parenting programs. The analyses of practitioner satisfaction and experienced applicability will mainly be descriptive, due to lack of power (to few practitioners involved in treatment) and due to fact that part of the data is qualitative (interviews). Finally, the analyses of mediation (research question 7), will be analyzed using Latent Growth Curve Modeling, which is suitable for data that is measured repeatedly during an intervention.

Parent-Child relations, Parenting, Non-inferiority trial, Externalizing behaviors, Randomized controlled trial, Cost-effectiveness

Participants are randomized to the standard intervention (Parent training in group format) or the experimental condition (Internet-based parent training).

Independent clinicians will perform diagnostic assessments at pre and post measurements. The clinicians will be blind to the experimental condition the participants they assess.

The standard format of the intervention that is well established in primary and specialized care in Sweden. Parents receive the education/training in small groups led by two group leaders.

A rating scale completed by parents measuring externalizing behaviors in children. The scale contains two subscales which both will be reported independently as outcomes in this study: The intensity scale (range 36-252) and the problem scale (range 0-36). In both cases higher values indicate more externalizing behaviors.

A rating scale completed by parents measuring both externalizing and internalizing behaviors in children, as well as the impact of such behaviors. The subscale of hyperactivity/attention problems will be used as a primary outcome in this study (range 0-10, where higher values represent more hyperactivity/inattention).

A structured diagnostic interview conducted by clinicians with parents and (if possible) the child. Only the sections for ADHD and ODD will be used in the present study. Outcomes that will be reported are diagnostic status of both ADHD and ODD (yes/no), the number of symptoms (range 0-20 for ADHD and 0-9 for ODD), and the clinician rated severity (range 0-8 for each diagnosis, where higher value represents more severe problems).

A rating scale completed by parents measuring parenting behaviors and strategies. The parents both rate the frequency of different activities (range 21-105 - higher values represent better outcome) and indicate if they find the activity/strategy problematic (range 0-21 - higher values represent worse outcome).

A rating scale completed by parents measuring the level of warmth and conflict in the relationship between the parent and child. There is one subscale for each of these constructs which will be reported separately. Warmth in the relationship has a range of 5-25, where higher values represent warmer relationship. Conflict has a range of 4-28, where higher values represent more conflicts.

A rating scale completed by parents measuring (own) perceived stress. The range of the scale is 0 to 40, with higher values representing more stress.

A rating scale completed by parents measuring (own) depression. The range of the scale is 0 to 27, with higher values representing more depressive symptoms.

A structures interview that a clinician conducts with a child, measuring the experienced quality of life. The subscales pertaining relationship to parents (range 0-14, high values representing better relationship) and own wellbeing (range 0-12, higher values representing more wellbeing) will be used in the present study.

A rating scale completed by parents measuring both externalizing and internalizing behaviors in children, as well as the impact of such behaviors. The subscales related to child well-being (emotional problems and peer problems) will be used as a secondary outcome in this study. The range of both subscales are 0 to 10, where higher values represent more problems.

A rating scale completed by parents measuring the satisfaction with the offered intervention - range 8 to 32, where higher values represent higher satisfaction.

A structured interview will be conducted with a selection of the clinicians about their experience with the interventions in the study, in terms of satisfaction and applicability.

The parents will each week throughout the intervention rate the present impact of their child's externalizing behaviors by completing the impact subscale of the Strengths and Difficulties Questionnaire (range 0-15, where higher values indicate more severe impact of the externalizing behaviors).

Inclusion Criteria: Primary reason for contact with health care is child externalizing behaviors. Exclusion Criteria: Not sufficiently fluent in written Swedish to be able to take part in an internet-based (text-based) intervention. No access to computer/ipad/phone Social problems that makes investigation or intervention from the social services necessary.

This has not yet been decided by the research team, but the information will be updated as soon as a decision has been reached.

Attkisson CC, Zwick R. The client satisfaction questionnaire. Psychometric properties and correlations with service utilization and psychotherapy outcome. Eval Program Plann. 1982;5(3):233-7. doi: 10.1016/0149-7189(82)90074-x.

Bullinger M, Brutt AL, Erhart M, Ravens-Sieberer U; BELLA Study Group. Psychometric properties of the KINDL-R questionnaire: results of the BELLA study. Eur Child Adolesc Psychiatry. 2008 Dec;17 Suppl 1:125-32. doi: 10.1007/s00787-008-1014-z.

Enebrink P, Hogstrom J, Forster M, Ghaderi A. Internet-based parent management training: a randomized controlled study. Behav Res Ther. 2012 Apr;50(4):240-9. doi: 10.1016/j.brat.2012.01.006. Epub 2012 Jan 30.

Essau, C. A., Sasagawa, S., & Frick, P. J. (2006). Psychometric Properties of the Alabama Parenting Questionnaire. Journal of Child and Family Studies, 15(5), 595-614.

Eyberg, S. M., & Pincus, D. (1999). Eyberg Child Behavior Inventory (ECBI) and Sutter-Eyberg Student Behavior Inventory-Revised (SESBI-R). Professional manual. Odessa, FL: Psychological Assessment Resources.

Flight L, Julious SA. Practical guide to sample size calculations: non-inferiority and equivalence trials. Pharm Stat. 2016 Jan-Feb;15(1):80-9. doi: 10.1002/pst.1716. Epub 2015 Nov 25.

Folkhälsomyndigheten. (2013). Så arbetar kommuner, landsting, kommun- och regionförbund och ideella organisationer med föräldrastöd. Östersund & Stockholm.

Ghaderi A, Kadesjo C, Bjornsdotter A, Enebrink P. Randomized effectiveness Trial of the Family Check-Up versus Internet-delivered Parent Training (iComet) for Families of Children with Conduct Problems. Sci Rep. 2018 Jul 31;8(1):11486. doi: 10.1038/s41598-018-29550-z.

Goodman R. Psychometric properties of the strengths and difficulties questionnaire. J Am Acad Child Adolesc Psychiatry. 2001 Nov;40(11):1337-45. doi: 10.1097/00004583-200111000-00015.

Hogstrom J, Olofsson V, Ozdemir M, Enebrink P, Stattin H. Two-Year Findings from a National Effectiveness Trial: Effectiveness of Behavioral and Non-Behavioral Parenting Programs. J Abnorm Child Psychol. 2017 Apr;45(3):527-542. doi: 10.1007/s10802-016-0178-0.

Julious SA. Sample sizes for clinical trials with normal data. Stat Med. 2004 Jun 30;23(12):1921-86. doi: 10.1002/sim.1783.

Kling A, Forster M, Sundell K, Melin L. A randomized controlled effectiveness trial of parent management training with varying degrees of therapist support. Behav Ther. 2010 Dec;41(4):530-42. doi: 10.1016/j.beth.2010.02.004. Epub 2010 Oct 1.

McEachern AD, Dishion TJ, Weaver CM, Shaw DS, Wilson MN, Gardner F. Parenting Young Children (PARYC): Validation of a Self-Report Parenting Measure. J Child Fam Stud. 2012 Jun;21(3):498-511. doi: 10.1007/s10826-011-9503-y.

Nordin M, Nordin S. Psychometric evaluation and normative data of the Swedish version of the 10-item perceived stress scale. Scand J Psychol. 2013 Dec;54(6):502-7. doi: 10.1111/sjop.12071. Epub 2013 Oct 5.

Sampaio F, Enebrink P, Mihalopoulos C, Feldman I. Cost-Effectiveness of Four Parenting Programs and Bibliotherapy for Parents of Children with Conduct Problems. J Ment Health Policy Econ. 2016 Dec 1;19(4):201-212.

Schumi J, Wittes JT. Through the looking glass: understanding non-inferiority. Trials. 2011 May 3;12:106. doi: 10.1186/1745-6215-12-106.

Sheehan DV, Sheehan KH, Shytle RD, Janavs J, Bannon Y, Rogers JE, Milo KM, Stock SL, Wilkinson B. Reliability and validity of the Mini International Neuropsychiatric Interview for Children and Adolescents (MINI-KID). J Clin Psychiatry. 2010 Mar;71(3):313-26. doi: 10.4088/JCP.09m05305whi.

Socialstyrelsen. (2009). Socialtjänstens öppna verksamheter för barn och unga - en nationell inventering av metoder. Stockholm.

Stattin H, Enebrink P, Ozdemir M, Giannotta F. A national evaluation of parenting programs in Sweden: The short-term effects using an RCT effectiveness design. J Consult Clin Psychol. 2015 Dec;83(6):1069-1084. doi: 10.1037/a0039328. Epub 2015 May 25.

Titov N, Dear BF, McMillan D, Anderson T, Zou J, Sunderland M. Psychometric comparison of the PHQ-9 and BDI-II for measuring response during treatment of depression. Cogn Behav Ther. 2011;40(2):126-36. doi: 10.1080/16506073.2010.550059.

Trentacosta CJ, Criss MM, Shaw DS, Lacourse E, Hyde LW, Dishion TJ. Antecedents and outcomes of joint trajectories of mother-son conflict and warmth during middle childhood and adolescence. Child Dev. 2011 Sep-Oct;82(5):1676-90. doi: 10.1111/j.1467-8624.2011.01626.x. Epub 2011 Aug 29.

Vårdgivarguiden. (2017). Vägledningsdokument: Utagerande beteende hos barn och unga med psykisk ohälsa.