Effect of a Videoconference-Based Online Group Intervention for Traumatic Stress in Parents of Children With Life-threatening Illness

Key Points Question Is an acceptance and commitment therapy–based group intervention, delivered using videoconferencing, effective in reducing posttraumatic stress symptoms in parents of very ill children? Findings This randomized clinical trial found that videoconference-based acceptance and commitment therapy (compared with a waiting list) was effective in reducing posttraumatic stress symptoms in parents of very ill children. Meaning This study supports the use of acceptance and commitment therapy as an approach to reduce posttraumatic stress symptoms in parents of very ill children, following an acute or unexpected illness or diagnosis, and finds that a videoconferencing platform can be used effectively to access hard-to-reach populations.


Introduction
Parents of children diagnosed with a life-threatening illness or injury are faced with significant psychosocial demands that may challenge their own psychological well-being. Although most parents adapt with time, a proportion suffer from psychiatric conditions, including posttraumatic stress disorder, as a direct result of their child's illness. 1,2 Others experience subthreshold, but clinically significant, psychiatric symptoms that can lead to longer-term mental health problems. 3,4 Parental mental health problems may have implications for long-term psychological, behavioral, and emotional problems for the child, [5][6][7][8] because high levels of distress can impair the parent's capacity to respond to the demands of their child's illness 9 and can affect the home environment after discharge. 10 A systematic review 9 of family adjustment to childhood cancer described the complex effects of child illness on family life, including the lack of time for nonessential activities, the needs of the child who is ill taking priority over those of parents and siblings, and extended periods of family separation during treatment. Importantly, these clinically significant distress reactions in parents have been found across many illnesses, including children who have experienced trauma, 11,12 those admitted to pediatric or neonatal intensive care units, [13][14][15] and those newly diagnosed with type 1 diabetes.
Interventions targeting these early distress reactions in parents are critical, because these initial reactions have been found to be associated with long-term outcomes. 3 This is consistent with the Pediatric Medical Traumatic Stress Model, 16,17 which provides a useful framework in which to understand important associated factors and the development and trajectory of psychological disorders in parents associated with child medical trauma, as well as identifying the optimal time for treatment.
Despite the emotional vulnerability of this population, evidence-based, preventive mental health interventions are lacking. The few studies 18,19 that have attempted to develop and test interventions have demonstrated limited success, with some evidence supporting problem-solving and family therapy approaches in parents of children with cancer.
Our study builds on promising pilot results 20,21 to evaluate the effectiveness of Take a Breath, a brief group intervention for parents of children with a range of life-threatening illnesses and injuries.
It is based on acceptance and commitment therapy (ACT) 22 and comprises key elements of acceptance, mindfulness, values clarification, and goal setting. Given that intrusive thoughts, avoidance, and high levels of emotional arousal are among the most common distressing symptoms in parents of children with a serious illness or injury, 23,24 the group-based ACT model was selected because it provides peer support and modeling of coping strategies and helps normalize the challenges faced by parents. The ACT-based intervention is delivered using videoconferencing to maximize accessibility for a highly stressed and geographically dispersed population.
This study aimed to determine the efficacy of the ACT-based program in reducing posttraumatic stress symptoms (PTSS) in parents. It was hypothesized that the program would result in a significant reduction in PTSS in the intervention group compared with a control group assigned to a waiting list for the intervention. The study also investigated whether the ACT-based program resulted in improvements in the psychological skills addressed in the intervention, such as mindfulness, valuesbased living, and psychological flexibility. Secondary outcomes explored whether there were improvements in other aspects of parental well-being, specifically, the parent's experience of their child's illness and family functioning. We hypothesized that the intervention group would show significantly better functioning within these areas compared with the waiting list group.

Method Study Design
Ethics approval for this randomized clinical trial was granted by the Royal Children's Hospital Human reporting guideline. The full protocol is available in Supplement 1 and has been published elsewhere. 25 Eligible participants were randomized to either an intervention or a waiting list control group.
Parents completed questionnaires at 3 time points: within 4 weeks of their child's diagnosis or admission (T1, screening), 2 weeks before intervention commencement, or approximately 4 to 6 months after the diagnosis or admission (T2, preintervention), and immediately after intervention completion (T3, postintervention).

Participants
Participants were parents of children with a recently diagnosed life-threatening illness or injury admitted to the oncology or cardiology departments or the pediatric intensive care unit (PICU) at the Royal Children's Hospital, Melbourne, Australia. Admissions into these departments were chosen because of the sudden onset of the child's illness and the significant threat posed to the child's life and future functioning. Participants were recruited from consecutive first admissions to these departments from January 2014 to February 2016. Eligible participants were (1) the primary caregiver (aged Ն18 years) of a child aged 0 to 18 years; (2) parents of a child with a first-time presentation for a new cancer diagnosis, cardiac surgery within the first month of life, or admitted to the PICU for more than 48 hours; (3) parents reporting elevated levels of acute stress symptoms as measured by the Acute Stress Disorder Scale (ASDS) 26 ; and (4) parents who were sufficiently fluent in English to complete questionnaires and participate in the program. Parents were excluded if they presented with a pre-existing psychiatric disorder, they had experienced another trauma (eg, death of family member) in the 2 months before the child's diagnosis, they were no longer the primary caregiver at the time of the intervention, or their child was deemed eligible for palliative care (as determined by the hospital clinical care team) or had died.

Procedure
The child's medical team was consulted to determine study eligibility. Parents were approached on the hospital wards (or by telephone if the child had been discharged) by the research team to discuss participation and to screen for eligibility. Screening included completion of the ASDS 26  Model. 4 In some families, both parents were eligible; however, randomization occurred at the family level. Thus, parents from the same family were always in the same condition. Partners of eligible parents who did not meet eligibility criteria were also invited to participate because they were seen as a key source of support for enacting changes 27 and may have also benefited from the program.
Data from parents who were invited but not eligible were not included in analyses.
Parents were asked to consent to trial participation and were enrolled at T2 before randomization. The Murdoch Children's Research Institute's Clinical Epidemiology and Biostatistics Unit, independently of the research team, generated the randomization list with a computerized randomization plan generator using the method of randomly permuted blocks. Three randomly allocated lists were created, 1 for each hospital department (cardiology, oncology, and PICU). A researcher independent of participant recruitment and intervention delivery managed the lists and allocated parents to intervention or the waiting list using sequentially numbered, sealed envelopes.
This process ensured that parents and the research team were blind to participant allocation at assessment. Once the intervention was completed, both the intervention and waiting list groups completed their surveys at T3.

Intervention or Treatment
The program is a 6-session, parent-mediated, psychological intervention based on ACT. 25

Facilitator Training and Program Fidelity
A total of 4 facilitators received training in ACT and in the delivery of the intervention from a senior ACT clinician, who is a specialist in trauma-focused psychotherapy. Each session was audio recorded and reviewed fortnightly during team supervision sessions. Fidelity of program delivery was recorded and assessed using session monitoring checklists completed by both facilitators after each session. An audio recording of 1 session from each group was randomly selected and reviewed for fidelity by an independent clinician, using a fidelity monitoring checklist.

Screening Measures
The ASDS and a demographic information questionnaire were completed at T1 only. All other measures were completed at T2 and T3 by parents in both groups. Each measure has been evaluated as a valid and reliable measure in these clinical populations.
The ASDS 26 measures acute stress reactions within 4 weeks after a traumatic event and is highly predictive of the development of longer-term PTSS and mental health problems. 3,28 The ASDS measures symptoms of dissociation, re-experiencing, avoidance, and arousal and can be scored using a total score cutoff (a score Ն56 was recommended for identifying individuals at risk of later developing posttraumatic stress disorder) 26 or in terms of symptom clusters. A score of 9 or higher on the dissociative cluster, as well as a score of 28 or higher on the combined re-experiencing, avoidance, and arousal cluster scores, indicates the presence of clinically significant levels of acute stress disorder symptoms 26 . Both scoring methods were used in this study to determine eligibility, with parents required to be eligible on at least 1 scoring method to be invited to participate. Internal consistency (Cronbach α) was 0.80 for the ASDS total score, 0.71 for dissociation, 0.32 for re-experiencing, 0.75 for avoidance, and 0.64 for arousal.

Primary and Secondary Outcomes
The primary outcome was PTSS in parents as measured by the Posttraumatic Stress Disorder Checklist-Version 5 (PCL-5). 29 It contains 20 items that assessed 20 criteria for posttraumatic stress disorder in the Diagnostic and Statistical Manual of Mental Disorders (Fifth Edition). 30 Parents were asked to complete the measure in relation to their child's diagnosis. The total score (range, 0-80) was used, with higher scores indicating greater PTSS. The PCL-5 has been found to be a valid and reliable measure across many populations. [29][30][31] Internal consistency for the total score in the current study was α = .93. Secondary outcomes measured included depression, anxiety, and stress

Demographic and Medical Information
Information was collected regarding parent age, sex, marital and employment status, level of education, and country of birth, as well as their child's age and sex. Medical diagnosis, date of diagnosis, and treating department were extracted from hospital medical records.

Statistical Analysis
To detect a difference of 7.5 (SD, 13.94) in the primary outcome measure (PCL-5), between the 2 treatment groups, with a significance level of 2-sided P < .05 and power of 0.80, 72 participants were required per group. Allowing for an estimated 35% lost to follow-up, it was projected that a total of 210 participants was needed to be randomized across the 2 groups.  43 These analyses were all stratified by illness group. All analyses were performed using Stata statistical software version 15.1 (StataCorp). Two-sided statistical significance was set at P < .05. To not unduly penalize the data, multiple comparison adjustment was not applied. Data analysis was performed from July to September 2018.

Results
The recruitment flow of participants can be seen in the Figure. In Preliminary analyses also explored the differences between participants who consented and participated, and those who consented but did not participate across these same sociodemographic Fidelity of program delivery was rated at 98% across all sessions. There was a total of 26 groups, with group sizes ranging from 3 to 8 parents. Table 2 shows the mean scores for all outcome measures at both the T2 and T3 time points, across both groups. Analyses compared T3 group means, adjusting for T2 scores as well as illness group, with the latter being nonsignificant in all Abbreviations: T2, 2 weeks before intervention commencement, or approximately 4 to 6 months after the diagnosis or admission; T3, immediately after intervention completion. a Cohen d values are for postintervention scores between groups, adjusted for preintervention scores.
b Mean comparison between groups at follow-up using generalized estimating equation models (clustering on dyad), adjusting for illness group and preintervention score. Concerning the investigation of the differences between the intervention and waiting list groups from T2 to T3, significant differences were found on the primary outcome measure ( Functional outcomes are displayed in Table 3, which report differences between the intervention and waiting list groups in the psychological skills taught in the intervention, from T2 to T3. The intervention group displayed improvements that were of a significantly larger magnitude in experiential avoidance (Acceptance and Action Questionnaire-II; specifically that parents were less flexibility. Together, these results suggest that participation in the intervention improved parents' capacity to better manage the demands of caring for a sick child and potentially ameliorated the negative impact of parental PTSS on the child. 5 The application of an ACT-based intervention is highly novel in this population, with past studies using traditional cognitive behavioral therapy 18,19,43,44 or problem-solving therapy. 18,24 Although we found significantly greater improvements in PTSS for parents receiving the 6-session intervention, there was no difference between groups for our secondary mental health measures of anxiety, depression, and stress. It is possible that a longer intervention is required to successfully address these additional mental health concerns; however, the program was not designed to address these symptoms. Past intervention studies in this field have all focused on the parents of children with cancer. [14][15][16]21,41 The ACT-based intervention was beneficial regardless of the child's illness or injury, providing support for intervention approaches that target parents across the wider hospital system.

JAMA Network Open | Pediatrics
The findings also support the use of videoconferencing as a promising approach to enhance accessibility, which is a key barrier to the provision of evidence-based treatments for this population. 45 Approximately half of our participants lived in rural or regional Victoria, Australia, or in another state. Use of the videoconferencing platform provided access to these parents who would not have been able to attend a traditional face-to-face service. Engaging parents within a tertiary hospital setting has been recognized as particularly challenging. 20,23,44,46 Our approach also demonstrates the feasibility of using videoconferencing for a group program to provide opportunity for peer support and the normalization of experiences.

Limitations
Implementation of this trial presented many challenges, and our study has some limitations.
Recruiting parents when their child was seriously ill was difficult, and it took 2 years to recruit the sample despite working within a large tertiary hospital. As a result, our modest sample size meant that we were only able to detect large differences between the intervention and control groups. In addition, attrition rate of parents initially screened and enrolled but then lost to follow-up before commencement of the program was 64%. Primary reasons for attrition at this stage were time constraints or feeling that intervention was not needed. Fathers and parents whose primary language was not English were overrepresented in the study dropouts. This finding is not uncommon in intervention studies and further highlights the need to ensure that recruitment and intervention materials reflect diversity and are engaging to all. 23,46 Other limitations that may have biased the findings include that the study only used self-reported measures and that, at the time of postintervention assessment at T3, the participants were no longer blinded as to group allocation.
Additionally, multiple comparison adjustments were not applied to secondary and functional outcomes, potentially inflating the type I error rate. Furthermore, significant baseline differences in mental health symptoms between the intervention and waiting list groups indicate that randomization did not successfully allocate equivalent groups to treatment groups. It is unclear why this occurred; hence, replication of differential postintervention improvements using a sample with equivalent groups at baseline is required to substantiate the important findings reported here.

Conclusions
The findings of this randomized clinical trial support the use of an ACT approach to reduce PTSS in parents of children who are seriously ill, across various illness groups, indicating that Take a Breath has potential for wide use within the pediatric hospital system. The findings also suggest that a videoconferencing format can be used in a clinically effective way to access geographically dispersed populations and engage their participation in a group program.