Mechanisms and Outcomes of Children and Adolescent Psychotherapy

Study Sample Phase I: Retrospective big-data study To assess October 7th implications, data derived from Clalit Health Services CHS databases will be used to assess October 7th's effects on parents' mental health and children's mental and physical health. General population of children 0-18 years old and their parents will be explored over a time span of a year (2000 - 2024). To assess rates of dropout from psychotherapy, data derived from Clalit Health Services (CHS) databases will be similarly minded. The data will include information on children and adolescents (6-18) psychotherapies from 2015 - 2024, as well as their caregivers. Children and adolescents will be divided into two groups based on their age (6-11 and 12-18). Inclusion criteria will be children and adolescents who participated in outpatient psychotherapy in CHS databases. Participants will be excluded in case of mental or physical hospitalization during psychotherapy. Phase II: Prospective exploration of ongoing psychotherapies This phase will explore possible pre-treatment and within-treatment predictors of adolescents' psychotherapy outcome and dropout. The phase will include three groups of participants: patients (adolescents between the ages 11-18), patient's caregivers, and therapists in the child and adolescent clinic in Shalvata Mental Health Center (MHC). Inclusion criteria for patients and caregivers include participation in psychotherapy in the Crisis Intervention Unit in Shalvata MHC. Inclusion criteria for therapists will be as follows (1) a professional license to practice psychotherapy (2) conducting psychotherapy in the child and adolescent clinic in Shalvata MHC. Patients and caregivers will be excluded in case of not being able to or willing to sign informed consent, or if there is an immediate risk, as evaluated by the treating physician. Procedure Phase I: Retrospective data phase To assess national trends in child and adolescent health since the 7th of October, Clalit anonymous database will be utilized. Data will be retrieved from databases according to the study protocol. The data timespan is from 2015 to 2024. To examine dropout rates among children and adolescents, and explore potential pre-treatment predictors, data will be extracted from the electronic health records of Clalit Health Services. Validation Process This is the first study to utilize psychotherapy data from the electronic health records of Clalit Health Services (CHS database). Therefore, before conducting data mining for the specific research questions, the data must first undergo a process to ensure validity and reliability. This involves crossing the raw patient identified data exported from the hospital system with anonymized data from the CHS database. The research team will work with the research room of the CHS database and the computing unit at Shalvata Mental Health Center. Only the main researchers and research room team will be exposed to raw patient identified data and only through the reliability phase. Following this reliability stage, all data will be anonymized and unidentified, according to research room regular procedure. Then, the research team will utilize just the unidentified data for the following analysis. Measure exported are elaborated in the measures section (measures phase 1). Phase II: Prospective exploration of ongoing psychotherapies To explore possible pre-treatment and within-treatment predictors of adolescent's psychotherapy outcome and dropout. Participants will be recruited from the Child and Adolescent clinic in Shalvata MHC. Patients will undergo a two-session intake with the assigned therapist, who will also provide the psychotherapy. After the first session, participants will be informed about the study and asked to sign informed consent. Children and adolescents under the age of 16 will be given an assent. Patients and caregivers will be assured that choosing not to participate in the research or stopping their participation will have no effect on their received treatment. Upon agreeing to participate, patients, therapists and caregivers will be assessed for pre-treatment variables as well as for within therapy factors (alliance and outcome and process expectations). Information on the patient's symptoms will be collected from children and caregivers. In case of any distress following their participation in the study, participants will be guided to inform and reach their therapists or other unit staff. Assessments will be conducted five times throughout therapy, at baseline week 1, at the end of weeks 4 at after the last session (week 7). Self-report questionnaires will be distributed via Qualtrics. Parental consent procedure: We request a special approval to allow one parent to sign the consent form when only one parent attends the intake session with the child (in cases where parents are married). In such instances, it is customary to ask the attending parent to inform the non-attending parent about their child's participation in the study. If the attending parent refuses to notify the other parent for any reason, we will inform them that the research team will contact the other parent directly via phone to communicate the child's participation in the study. For divorced parents, we will have the attending parent sign the consent form during the session. Within 3-5 days, we will contact the non-attending parent by phone. The researcher will record the non-attending parent's name on the consent form and note that the consent was obtained remotely. The child will only participate in the study after the second parent provides their consent. For divorced parents, a dedicated table will be used to document the process of contacting the second parent. The researcher responsible for making the contact will sign the table to confirm that the process was completed. Measures and Data Fields Phase I: Retrospective data phase The following data fields will be examined: Demographic variables (both parents and children): age, sex, sector, socioeconomic status score, district, residency, patient decease date if exists. Mental health variables (both parents and children): diagnosis, age of diagnosis, date of psychotherapy beginning and end, psychotherapy no-show dates, psychotherapy end reason (if available), medication purchase, mental health doctor visits and dates, mental health psychotherapy visits and dates, paramedical visits and dates (occupational therapy, speech therapy), profession type, mental health emergency department (ED) visits, mental health admission date, mental health admission reason. General health variables (both parents and children): diagnosis, age of diagnosis, medication purchase, doctor visits and dates, profession type, ED visits, admission date, admission reason, Charlson score total, allied health profession visits. In the data phase, participants will be considered as dropouts when failing to continue therapy. Specifically, in case of at least three missed sessions, with at least two sessions prior and no more than two sessions after (e.g., possible termination and goodbye) their non-attendance. If therapists did not continue to work in the clinic afterwards, patients will not be considered dropouts (a possible job termination by the therapist). Phase II: Prospective exploration of ongoing psychotherapies The following measures will be used to assess predictors of psychotherapy outcome and dropout. Demographics. Age, sex, sector, residency, number of children, education, occupation, marital status, referral source, participation in individual psychotherapy, and mental and general health diagnosis will be measured. The Pediatric Symptom Checklist- 17 (PSC-17: is a 17-item self-report scale designed to assess internalizing, externalizing, and attention problems in children and adolescents aged 4 to 16 years. Parents or caregivers will complete the checklist by indicating whether each item is "never," "sometimes," or "often" true for their child. A parallel version will be filled out by children and adolescents (PSC-17-Y).The PSC was translated into Hebrew in 2012 by Carmit Frisch and Professor Sara Rosenblum. There is a validation study of the 17-item version in Hebrew indicating high reliability using Cronbach's α: 0.92 for the total score, 0.87 for the internalizing, 0.86 for externalizing, and 0.84 for attention problems sub-scales. The DSM-5 Level 2 Measure-Somatic Symptom-Parent/Guardian. Somatic symptoms will be assessed with a 15-item somatic symptoms severity scale (PHQ-15) from the Patient Health Questionnaire. The measure is completed by the parent and each item asks the parent to rate the severity of the child's (ages 6-17) somatic symptoms during the past 7 days. Each item on the PHQ-15 is rated on a 3-point scale (0=not bothered at all; 1=bothered a little; 2= bothered a lot). PHQ-15 has been translated into several languages, and previously used in Israel. Previous internal consistency was reported to be Cronbach's α= 0.80. Hopkins Symptoms Checklist-11 (HSCL-11: The HSCL-11 is a short form of the revised 90-item symptoms checklist. It contains 11 items that participants rate on a 4-point Likert scale. It aims to assess symptomatic distress including anxious and depressive symptoms. The HSCL-11 has been widely used worldwide, and previously shown high reliability in its Hebrew version Cronbach's α= 0.96. The Working Alliance Inventory- Short Revised (WAI-SR). WAI-SR is a self-report questionnaire that aims to assess the quality of the therapeutic relationship and agreement on goals during therapy. Assessments are made from both the patient, parent, and therapist perspective. The scale has been previously used in children adolescents (ages 11-17), and is commonly used in this age group. Previous research has also used the WAI measurement to assess caregiver-therapist alliance. It consists of 12 items that are rated on a 5-point Likert scale. Furthermore, previous research on adolescent therapeutic alliance has shown high internal consistency (Cronbach's alpha=0.91 ). Credibility/Expectancy Questionnaire (CEQ). CEQ is a self-report questionnaire that aims to examine the credibility of the treatment, as well as outcome expectations. It consists of 6 items that are rated on a 9-point (for questions 1,2,3,5) or 11-point (for questions 4 and 6) Likert scale. CEQ has been previously examined for parents of patients and was found to be a reliable measurement (CEQ-P) . The questionnaire was also previously used with adolescents and showed good internal consistency (Cronbach's α=0.79-0.90). In the current study, CEQ will be filled out by parents, patients and therapists. Expectations of the Active Process in Psychotherapy Scale (EAPPS). EAPPS is a 32-item self-report questionnaire that aims to assess the expectations of the active process that is assumed to produce change in the psychotherapeutic process. It consists of seven factors: (1) positive therapist-client relations, (2) verbal processing of therapist-client relations, (3) exploration of unexpressed content, (4) the ability to share sensitive content openly and securely, (5) working through specific emotional problems, (6) therapy fosters resilience, (7) providing tools for cognitive control. EAPPS will be filled out by parents and therapists only. The scale was found to have satisfactory internal consistency Cronbach's α ranging between 0.66 to 0.85. Exposure to Traumatic Events. Parents will be asked to describe their own and their children's level of exposure to traumatic events during the war. They will use nine self-report items consisting of a 5-point Likert scale questions to indicate their exposure to various aspects of the war: being present during a terrorist assault, experiencing the death or kidnapping of a close person, experiencing war-related injury, knowing someone who was injured as a result of war events, experiencing missile alarm sirens, personal or close familial involvement in military or civil service presence in combat zones, and exposure to videos, pictures and written stories of the 7th of October attack via social media. Parents will be asked to state whether their child was also exposed to these events using yes/no questions. Dropout Questionnaire Therapists will be asked a single question about their expectation on the possibility of their patient's dropout from therapy. Rating will be assessed on a 7-point Likert scale. In case of therapist's report of dropout after treatment has begun, they will be asked about the motives of treatment termination. At first, therapists will be asked in a multiple choice about the reasons they assume led to an early termination from their perspective and from the patient's and caregiver perspectives. 10 questions will be asked to explore whether the decision to terminate treatment was made unilaterally. The distribution of measurements across phase III is shown in the appendix. Statistical Strategy For the retrospective data analysis phase, aimed to examine the predictive role of children's and parents' pre-treatment variables (as elaborated in the measures section) on therapy dropout, univariate and multivariate logistic regressions will be conducted. Similar models will be used to explore the odds ratio for developing a mental illness or showing other signs of a decline in children's mental health state (such as mental health ER visits, psychiatric medication prescriptions, new diagnoses, etc.). The analyses will utilize Kaplan-Meir survival analyses and Cox regressions. To examine the role of patient-caregiver, patient-therapist, and caregiver-therapist congruency in alliance and expectations on likelihood to dropout and outcome (prospective phase), multilevel structural equation modeling (MSEM) will be used (Preacher et al., 2010) to account for the nested data structure (nested data within dyads nested within therapists). MSEM analysis will be conducted in Mplus 8.4 program. Supervised machine learning approach will be conducted to assess the integrative model of pre-treatment and within-treatment factors combined. The analytical plan includes the following steps: a priori sample size estimation for both training and validation of models, reporting on pre-process decisions, developing a training sample and external cross-validation. Model training will be conducted in R version 4.1.1 using package caret v6.0-90 and caret Ensemble v2.0.1.