Bounce Back

Study 1

Summary

Langley et al. (2015) randomly assigned 74 students within four different schools to either immediate intervention or the waitlist control group. Parent and child reports of posttraumatic stress and depression, and child reports of anxiety symptoms, were assessed at baseline and three months after baseline (posttest).

Langley et al. (2015) found that, as compared to the control group, at the posttest treatment students significantly improved:

• posttraumatic stress symptoms (parent and child reported)
• anxiety symptoms (child reported)
• emotion regulation (parent reported)
• emotional/behavioral problems (parent reported)

In terms of risk and protective factors, Langley et al. (2015) found that compared to the control group at the posttest, treatment students significantly improved on measures of:

• social adjustment (child reported)

Evaluation Methodology

Design:

Recruitment: Participants were recruited from four Los Angeles elementary schools. Students with psychotic disturbances (e.g. acute suicidal behavior) or sexual abuse as an only or primary trauma were excluded from participation given that Bounce Back is provided in a mixed-gender format. Students were invited to be screened; of the 417 screened, 113 were eligible based on past trauma experience and current elevated posttraumatic stress symptoms, and 74 enrolled.

Assignment: The 74 participants were randomly assigned within the four schools to immediate intervention or a delayed (waitlist) group. The waitlist group received the intervention approximately three months following the initial intervention group.

Assessments/Attrition: Participants completed assessments at baseline, 3 months after baseline (posttest), and 6 months after baseline (follow-up). By the time of the 6-month assessment, the waitlist control group had already received the intervention, thus analyses focused on the 3-month period from baseline to posttest. Three students did not complete a posttest assessment (sample n = 71), resulting in an overall attrition rate of 4% at the posttest.

Sample:

The sample was 50% male, predominantly Hispanic (48%) and Caucasian (27%), and fairly low household income (42% below $40,000/year). All schools qualified as Title I, with at least 40% of the students qualifying for free or reduced-price lunch. The average grade level was 2.73 and average age was 7.65.

Measures:

Primary Outcome Measures: Participants' levels of posttraumatic stress symptoms and symptom frequency were measured using the UCLA Posttraumatic Stress Disorder Reaction Index, with both parent and child reports of frequency and severity of PTSD symptoms (α=.88 for parents and α=.83 for children). The Children's Depression Inventory was used to assess children's depressive symptoms (α=.84 for parents and α=.82 for children). The Screen for Child Anxiety Related Emotional Disorders assessed child-reported symptoms corresponding to various anxiety disorders (α=.92).

Secondary Outcome Measures: The Strengths and Difficulties Questionnaire for parents and teachers assessed students' problem areas (emotional, conduct, hyperactivity/inattention, and peer relationships) as well as prosocial behavior (α=.61 for parents and .74 for teachers). The Social Adjustment Scale-Self-Report for Youth evaluated the degree of cohesion and support within a student's interpersonal relationships (α=.80). The Coping Efficacy measure assessed children's satisfaction with handling past and current stressors and their anticipated effectiveness for handling future stressors (α=.77). The Emotion Regulation Checklist, a parent-report measure. was employed to assess how easily children are upset and how quickly or easily they can get past stressors (α=.82 for the negativity subscale and α=.69 for emotional regulation).

Analysis: Linear mixed-effects models were used, with group (immediate treatment and waitlist control) as a between-subjects factor and time (baseline, 3 months, 6 months) as a within-subjects factor, and a group by time interaction to examine differential treatment effects from baseline (pretest) to 3 months (posttest). Main effects of site and group by site interactions were included in the models to control for clustering within schools, and the main effect analyses examined differences between groups from baseline to 3 months. In addition, the Benjamini-Hochberg false discovery rate approach was used to correct for multiple comparisons.

Intent-to-Treat: Mixed-effects models automatically handle missing values via likelihood-based methods, producing unbiased parameter estimates provided observations are missing at random. Because of this and small attrition rates, the authors did not impute missing data. However, they reported that if more than 25% of the items in a given outcome measure were unanswered, the total score for the instrument was treated as missing. If less than 25% were unanswered, the authors computed an expected total based on the average of the completed items.

Outcomes

Implementation Fidelity: Fidelity ratings indicated excellent therapist adherence to the intervention manual (mean 2.91 on a scale of 0 to 3) and excellent overall quality of content implementation (mean 2.94 on a scale of 0 to 3).

Baseline Equivalence: There were no statistically significant differences between treatment and control on 19 socio-demographic variables and outcome measures, except for the highest level of parental education, with the parents in the immediate treatment group being slightly more educated than those of the control group.

Differential Attrition: Langley et al. (2015) did not examine baseline differences for the analytic sample. In terms of differences between those who completed the study and those who did not, "there was no evidence of a baseline difference between the subjects who did and did not complete the study except with regard to ethnicity: six of the 13 African American participants did not complete the study (five of these six because they relocated from the area and changed schools) compared with one of 36 Hispanic participants and none in the other ethnic groups" (p.859).

Posttest: At the posttest (i.e., 3 months after baseline), Langley et al. (2015) found that treatment students, in comparison to control students, showed lower levels of parent- and child-rated posttraumatic stress symptoms, child-rated anxiety, and parent-rated emotional/behavioral problems while also exhibiting improvement in parent-rated emotional regulation.

In terms of risk and protective factors, compared to the control group at the posttest, treatment students significantly improved on measures of self-reported social adjustment.

Long-Term: Not conducted.

Study 2

Summary

Santiago et al. (2018) conducted a cluster randomized controlled trial in which eight schools and 52 students were assigned to immediate intervention or a waitlist control group. Measures assessing PTSD symptoms, anxiety, depression, coping skills and classroom behavior were collected before the intervention (pretest) and three months after baseline (posttest).

Santiago et al. (2018) reported that at the posttest, compared to the control group, students in the treatment group showed improvements in:

• posttraumatic stress symptoms (child reported)

In terms of risk and protective factors, Santiago et al. (2018) reported that treatment students, as compared to control students, showed improvements in:

• coping (parent reported)

Evaluation Methodology

Design:

Recruitment: Teachers and school social workers identified children in the first through fourth grades who might benefit from the program through the school referral process. The inclusion criteria were (a) exposure to trauma (identified using the modified TESI-C-Brief; Ford et al., 2000) and (b) current moderate to severe symptoms of posttraumatic stress (score of 25 on the UCLA-RI; Steinberg et al., 2004). According to the consort diagram presented in Figure 1 (p. 5), a total of 105 students across 8 schools were screened, and 52 were randomly assigned to condition.

Assignment: Using block randomization, 52 children nested in 8 schools were randomly assigned to immediate treatment or to a waitlist control once students were screened (thus, schools were the unit of assignment). Because of these block randomization procedures, schools that participated multiple years were not necessarily in the same condition across all three years of the study (2013 - 2016). The number of schools assigned to each condition was not reported. Treatment and control included six groups each (25 students in treatment and 27 students in control).

Assessments/Attrition: Assessments were conducted at baseline, and again at three months (posttest) and six months after baseline. The waitlist control group began receiving the intervention after the three-month assessment. According to Figure 1 (p. 5), there was no attrition from assignment to the three-month follow-up.

Sample: The mean age of the sample was 7.76 years, 65% were male, and 82% were Latino. Almost half of the students (47%) had two immigrant parents, 16% had one immigrant parent, and 37% had no immigrant parents. More than half of the participants (59%) had a household income of less than $25,000, supporting an average of four individuals. More than half (55%) of parents had an education of less than high school. The mean age of parents/caregivers participating in this study was 35.75 years, 83% were female.

Measures:

Posttraumatic stress symptoms: Both parents and children reported on children's symptoms of PTSD using the University of California-Los Angeles PTSD Reaction Index (UCLA-RI; Steinberg, Brymer, Decker, & Pynoos, 2004), a 20-item questionnaire on posttraumatic stress symptom frequency during the past month. This instrument was administered verbally by research staff and school social workers to children and with paper and pencil for parents. Items on this scale are rated on a 5-point Likert-type scale from 0 (never) to 4 (most of the time) and Cronbach's alpha in this sample ranged from .86 to .92 for parent report and from .74 to .92 for child report across time points.

Depression: Children completed the self-report version of the Children's Depression Inventory (CDI; Kovacs, 1992). An item assessing suicidal ideation was excluded in this study, yielding a 26-item version. For each item, children selected the statement that best described them from among three options, with item scores ranging from 0 to 2. Parents also reported on their child's symptoms by completing a parent version of the CDI at each time point. Cronbach's alpha for parent report in this sample ranged from .88 to .90 and from .85 to .86 for the child report.

Anxiety: Children completed the Screen for Child Anxiety Related Emotional Disorders Child Report (SCARED-C; Birmaher et al., 1999), a 41-item measure that includes five factors: somatic/panic (13 items), generalized anxiety (9 items), separation anxiety (8 items), social phobia (7 items), and school phobia (4 items). Participants rated the items of each factor on a 3-point scale. In the current sample, Cronbach's alpha ranged from .86 to .92.

Coping: Parents reported on children's coping using the Responses to Stress Questionnaire (RSQ; Connor-Smith, Compas, Wadsworth, Thomsen, & Saltzman, 2000), a 57-item measure that assesses responses to stress. The primary control coping factor was used in this study and includes nine items assessing problem solving, emotional expression, and emotion regulation rated on a 5-point scale of 0 (not at all) to 4 (a lot). In the current study, Cronbach's alpha ranged from .70 to .77.

Classroom behavior/school functioning: Teachers reported on children's school functioning using the Strengths and Difficulties Questionnaire (SDQ) Teacher Report (Goodman, 1997). The SDQ is a 25-item questionnaire with five subscales assessing school functioning: emotional symptoms, conduct problems, hyperactivity/inattention problems, peer problems, and prosocial behavior. A total difficulties score is calculated by summing the first four subscales, with a higher score indicating more problems. The Cronbach's alpha for the current sample ranged from .80 to .86.

Analysis: Repeated-measures analyses of variance (ANOVAs) were conducted using IBM SPSS Statistics. For the main effect analysis, a time (within-subject factor) by group (between-subject factor) interaction was examined (baseline to 3 months). Only one of the eight pretests were included in the model: since parent-reported posttraumatic stress symptoms was significantly higher for the treatment group at baseline, this pretest was adjusted for in the analysis. The main effect analysis also included hierarchical linear modeling (HLM) to account for the nested nature of the data (child nested within group/school). Though not presented in the current study, the authors reported that HLM results were replicated and are available upon request.

Intent-to-Treat: The authors reported on page 5 that "missing data ranged from 2% to 12% for child report and 10-23% for parent report across measures/time. However, data were imputed; thus, all cases were analyzed." The authors also reported that "Little's (1988) missing completely at random test was conducted, and data were determined to be missing at random. Subsequently, missing data were imputed using maximum likelihood multiple imputation procedures (using SPSS 24.0 Expectation Maximization program)" (p. 5).

Outcomes

Implementation Fidelity: Clinicians reported good fidelity of their delivery of Bounce Back (mean 2.53 on a scale of 0 - 3). In terms of dosage, 90-98% of children attended scheduled sessions or received a make-up over the 10 group sessions. Most of the children (84%) had caregivers who attended at least one parent session. Parent involvement ranged from zero to four sessions completed.

Baseline Equivalence: There were no significant differences between treatment and control groups across six demographic characteristics (gender, race/ethnicity, income, grade, age/years, and primary caregiver education). However, there were differences in one of 8 pretest measures (in favor of the control group). That is, parent report of posttraumatic stress symptoms was significantly higher for the treatment group at baseline, and (as mentioned in the analysis section) was adjusted for in the analysis. Teacher report of classroom behavior problems was higher for the treatment group (p=.05) but was not adjusted in the analysis.

Differential Attrition: No attrition.

Posttest: At the posttest (i.e., 3 months after baseline), Santiago et al. (2018) found an effect on two of seven outcomes. Results showed that students in the treatment group demonstrated greater reductions in child-reported PTSD (effect size = .11) and improvements in parent-reported child coping (effect size = .13), compared with the control group. There were no significant group differences in parent-reported PTSD, parent- and child-reported depression, child-reported anxiety, and teacher-reported classroom behavior.

Long-Term: Not conducted.