Comparing Two Approaches to Help Parents Support Their Children's Social and Emotional Recovery after a Serious Accidental Injury

Comparing Two Approaches to Help Parents Support Their Children's Social and Emotional Recovery after a Serious Accidental Injury - NCBI Bookshelf An official website of the United States government Here's how you know The .gov means it's official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you're on a federal government site. The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely. Log in Show account info Close Account Logged in as: username Dashboard Publications Account settings Log out Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation Bookshelf Search database Books All Databases Assembly Biocollections BioProject BioSample Books ClinVar Conserved Domains dbVar Gene Genome GEO DataSets GEO Profiles GTR Identical Protein Groups MedGen MeSH NLM Catalog Nucleotide OMIM PMC Protein Protein Clusters Protein Family Models PubChem BioAssay PubChem Compound PubChem Substance PubMed SNP SRA Structure Taxonomy ToolKit ToolKitAll ToolKitBookgh Search term Search Browse Titles Advanced Help Disclaimer NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health. Comparing Two Approaches to Help Parents Support Their Children's Social and Emotional Recovery after a Serious Accidental Injury Marizen Ramirez , MPH, PhD, Briana Woods-Jaeger , PhD, Corinne Peek-Asa , MPH, PhD, Joseph E. Cavanaugh , PhD, Kristel Wetjen , RN, Cassidy Branch , MA, Javier E. Flores , MS, Lisa Roth , BS, Mallory Bolenbaugh , BS, Elisabeth A. Stelson , MSW, MPH, Patricia Espe-Pfeifer , PhD, Kimberly A. Randell , MD, MSc, Vidya Chande , MD, Henry W. Ortega , MD, Charles Jennissen , MD, Graeme Pitcher , MD, Linda Snetselaar , PhD, and Justin Kenardy , PhD. Author Information and Affiliations Authors Marizen Ramirez , MPH, PhD, 1,2 Briana Woods-Jaeger , PhD, 3 Corinne Peek-Asa , MPH, PhD, 1 Joseph E. Cavanaugh , PhD, 1 Kristel Wetjen , RN, 4 Cassidy Branch , MA, 5 Javier E. Flores , MS, 1 Lisa Roth , BS, 1 Mallory Bolenbaugh , BS, 1 Elisabeth A. Stelson , MSW, MPH, 1 Patricia Espe-Pfeifer , PhD, 6 Kimberly A. Randell , MD, MSc, 3 Vidya Chande , MD, 7 Henry W. Ortega , MD, 8 Charles Jennissen , MD, 6 Graeme Pitcher , MD, 6 Linda Snetselaar , PhD, 1 and Justin Kenardy , PhD 9 . Affiliations 1 University of Iowa, College of Public Health, Iowa City 2 University of Minnesota, School of Public Health, Twin Cities 3 Children's Mercy Hospital, Kansas City, Missouri 4 University of Iowa Hospitals & Clinics, Iowa City 5 University of Iowa, Public Policy Center, Iowa City 6 University of Iowa, Carver College of Medicine, Iowa City 7 Blank Children's Hospital, Des Moines, Iowa 8 Children's Hospitals and Clinics of Minnesota, St Paul 9 University of Queensland, School of Psychology, Queensland, Australia Washington (DC): Patient-Centered Outcomes Research Institute (PCORI) ; 2020 Jul . Copyright and Permissions Copyright © 2020. University of Iowa. All Rights Reserved. This book is distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License which permits noncommercial use and distribution provided the original author(s) and source are credited. (See https://creativecommons.org/licenses/by-nc-nd/4.0/ Structured Abstract Background: Unintentional injury is the leading cause of hospitalization for children. Children can recover physically but may develop psychological distress and struggle in school. Objectives: We compared 2 approaches that parents could use to help their children recover emotionally and socially from accidental injuries. One is Link for Injured Kids (Link), a program of psychological first aid, which guides parents on how best to speak with their child after an injury, identify signs of distress, and obtain more care for their child if needed. The second approach is an educational booklet given to families as they leave a hospital after a child's injury. The aims of this research were (1) to assess the adoption and implementation of materials for both programs, (2) to compare the effectiveness of the 2 programs on outcomes impacted by the interventions, and (3) to examine differences in the comparative effectiveness of the 2 interventions in rural and urban families. Methods: A comparative effectiveness trial was conducted at 4 children's hospitals in urban and rural settings across the Midwest. Eligible participants were dyads consisting of a parent and their child who was between the ages of 10 and 17 years and had been admitted to the hospital because of unintentional injury. Upon recruitment, dyads were randomly assigned to 1 of 2 treatment groups: (1) Link or (2) trauma education. Parent-child dyads each completed questionnaires at baseline. Parents also answered questions about their children. Parents assigned to the Link group participated in a 20- to 45-minute video and in-person training in motivational interviewing-based communication skills; parents assigned to the trauma education group received an educational booklet. Parent-child dyads completed follow-up questionnaires at 6 weeks, 3 months, and 6 months posthospitalization. At 6 weeks, parents in each intervention group received telephone booster training. Parents from each respective intervention group were asked to report how many times they used Link skills or reviewed trauma education materials at every follow-up point. Based on responses, parents were then categorized into implementation levels (ie, full, partial, or none). Using descriptive statistics and repeated measures modeling, we compared the differences in implementation; scores on parent and child stress measures, including those for posttraumatic stress disorder (PTSD) and depression (primary outcomes); and quality of life (QOL) and child behaviors (secondary outcomes) between intervention groups. An interaction between intervention and location (rural or urban) was examined. Sensitivity analyses found no differences in outcomes when adjusted for baseline symptoms. Results: Of a total of 795 eligible participants, 314 children and their parents enrolled in the study for a 39.5% participation rate. About 10% were racial/ethnic minority children, and 30% came from rural settings. The overall average length of hospital stay was 3.08 days for children in the study. More than half of parents trained in Link reported being partial implementers (54%-58%) throughout the study period. About 60% of trauma education parents fully implemented the intervention, but this decreased to 43% at 3 months and 38% at 6 months. Across both intervention groups, implementation decreased over time ( P = .0025). However, there were no differences between the Link and trauma education interventions in the improvement over time in child symptoms of PTSD and depression (the primary outcomes) or in nonspecific stress and pediatric QOL (the secondary outcomes), even after adjustment for baseline symptoms. Extent of implementation was not associated with differences in the effectiveness of the 2 interventions. Link was associated with mild improvement of selected child behaviors compared with trauma education (child conduct [β = .971; 95% CI, .94-1.00] and peer behaviors [β = .965; 95% CI, .93-1.00]). Additionally, Link was found to be slightly more effective in improving conduct behaviors in rural compared with urban children ( P = .04). Conclusions: Link was not as fully implemented by parents as trauma education. Trauma education and Link were equally associated with improvements in child symptoms of PTSD and depression (the primary study outcomes) and in nonspecific stress and QOL (the secondary outcomes). Link, which is relatively resource intensive, was only slightly better in improving problem behaviors than trauma education. However, future research is needed to verify these findings and explore the potential for Link as an intervention for injured children with preexisting or baseline behavioral problems. Limitations: Our study, which is limited to the samples recruited from 4 Midwestern US children's hospitals, had moderate retention rates (66%-76%) and involved self-reported psychosocial measures rather than clinician-diagnosed psychiatric disorders. Fidelity of intervention delivery by parent to child was not directly measured. Background Personal and Societal Burden of Unintentional Injury in Children Unintentional injury constitutes the greatest single cause of morbidity and mortality among children in the United States. 1 Every year, over 9 million children seek emergency care for injuries, and an average of 12 000 children die from their injuries. Of those seeking emergency care, 225 000 children necessitate further hospitalization. 1 , 2 Many injuries leave behind more than just physical scars, and the psychological toll can be substantial for children and families. The injury event, pain and physical symptoms, hospitalization, medical interventions utilized, and a lengthy and socially isolating recovery period can contribute to a number of chronic psychological conditions, of which posttraumatic stress disorder (PTSD) is the most prominent. 3 , 4 PTSD is a cluster of symptoms (re-experiencing, avoidance, arousal and reactivity, and negative thought patterns) that persist at least 1 month after exposure to a traumatic event. 5 PTSD prevalence estimates among injured children range widely due to the nature of the injury; demographic, environmental, and historical factors of the child; and difficulty providing posttrauma care to families after hospital discharge. 6-9 Although the majority of children recover psychologically from traumatic injury, studies estimate that 18% to 34% of injured children develop PTSD 1 to 2 months post injury; after 3 to 6 months, the prevalence will decrease to 6% to 23%. 10-12 Even at the most conservative estimate of incidence (18%), approximately 40 500 children could develop PTSD after an injury every year. 10 , 13 Children with a history of trauma, who are female, and whose parents experience depression or PTSD from their child's injury are at greater risk for developing PTSD. 8 , 9 , 11 , 14 , 15 Studies have also found that children who are recovering from a traumatic injury experience diminished quality of life (QOL) in various aspects of their lives and are also at risk for developing depression. 9 , 16 PTSD and other posttraumatic comorbid disorders can severely impact overall health and well-being long after the burns, cuts, and bones have healed. 16 , 17 Throughout the duration of a hospital-based study's 2-year follow-up of injured youth, patients who developed psychological distress had significantly lower mobility and physical and social activity levels than those who did not. 18 School performance can also be impacted, as youth with PTSD symptoms are more likely to miss or drop out of school, suffer from decreased concentration and executive functioning, and have poor grades. 19-23 If left unaddressed, these psychological wounds follow children into adulthood, placing them at elevated risk for mental health and social functioning problems throughout their lives. 6 , 22 Supporting children who experience mental health problems after trauma can be daunting for any family, but this is especially true for families in rural areas who face challenges associated with access, availability, and acceptability of mental health services. 9 , 24-28 Rural families are also subject to disparities in specialty pediatric and psychological services, primarily due to insufficient health infrastructures. 29 , 30 Many rural hospitals lack standardized protocols to identify traumatized children at risk for poor psychosocial outcomes. 31 , 32 With no standard of care in place, caregivers are sent home without information about the psychological support that their injured children may require. The economic burden associated with trauma-induced psychological disorders is substantial. In the general population, PTSD accounts for high medical and mental health service utilization and is one of the costliest psychological conditions, largely attributed to patients' repeat contact with the health care system to relieve under-recognized symptoms. 33 The costs to rural hospitals can be even higher than for their urban and suburban counterparts because they are more likely to see repeat patients for underlying mental health problems. 34 Treating and Preventing Adverse Symptoms After Injury Early recognition of problematic symptoms and intervention is needed after exposure to trauma to reduce acute and long-term impacts. 35-37 However, interventions engaging health care providers have shown minimal effectiveness in reducing posttraumatic stress. Stallard's 2006 trial of a single-session debriefing program provided to pediatric patients with road traffic injuries was found to have no effect on PTSD symptoms. 38 “Stepped collaborative care” was used by Zatzick et al and Kassam-Adams et al to support injured patients at risk for psychological sequelae, but this approach likewise failed to reduce symptoms. 39 , 40 Trauma-focused cognitive behavioral therapy has long been recognized as an evidence-based best practice for a variety of childhood traumas, but its effectiveness has not been evaluated for childhood injury. 29 , 36 , 41 , 42 One potential limitation of these programs is that they usually involve formal health care providers as interventionists, who in real-world settings may have limited contact with patients, especially after discharge. Another promising approach focuses on parents and guardians as primary interventionists in the posttrauma care of children. A study by Kenardy et al in Australia found that when delivered to parents, a series of booklets entitled So You Have Been in an Accident , which provides trauma education and resources, was associated with reduced PTSD symptoms for both children and their parents. 43 Child patients and parents allocated to the intervention treatment group received an informational booklet that outlined common reactions to trauma, self-help instruction, and contacts for support services. Children in the intervention group had decreased anxiety symptoms between 1 and 6 months posttrauma, whereas children in the control group showed increases in anxiety for the same time period. Similarly, parents in the intervention group displayed a greater decrease in posttraumatic distress than the control group during the 1 month period post trauma. Other psychoeducational interventions that similarly delivered information to parents have been found to improve behaviors, 44 reduce posttraumatic symptoms of anxiety, 45 and provide parents with positive coping strategies for their injured child. 46 Berkowitz et al developed the Child and Family Traumatic Stress Intervention that provides family social support and coping skills through increased parent-child communication and parent-supported behavioral skills. 37 Although found to reduce PTSD, Berkowitz's program requires 4 sessions with mental health professionals, which is likely resource prohibitive for families in many rural settings. Regardless of the availability of health resources, many parents and children discuss the impact of traumatic injury on the child's mental health and well-being during the recovery period. These spontaneous, naturally occurring conversations specific to the child's injury have shown positive effects on emotional recovery post trauma. 47 However, parents are often unsure about how best to approach these sensitive conversations, and many of these conversations become instructive or admonishing rather than reflective and supportive of the child's experience. 48 To help equip parents with the tools to initiate these crucial conversations, Link for Injured Kids (Link) was developed as an intervention to train parents in motivational interviewing (MI) skills to help them facilitate emotionally supportive, open conversations with their child. 49 Teaching MI skills to parents has been shown to promote a variety of child health behaviors such as medication adherence, exercise, and weight loss, and decrease maladaptive behaviors such as smoking and sexual risk activities. 50 , 51 Link emerged from Listen, Protect, Connect (LPC), a child-focused program developed to support children who experienced a variety of individual and community crises. 51 , 54-56 In a quasi-experimental 1 group pretest/posttest study, middle- and high school-aged victims of trauma who received LPC from school nurses experienced reductions in depressive and PTSD symptoms and increased connectedness with an adult at school. Based on the need for a similar intervention for parents of children who have experienced traumatic injury, LPC was adapted into Link for pediatric emergency settings. 49 From this a priori body of work, 2 promising approaches were identified to support parents and children after a traumatic injury. One is “So You've Been in an Accident,” a low-resource intervention that involves delivery of trauma education. The second is Link, a more intensive intervention that involves MI skills to build rapport between parent and child. In the current study, we compared the effectiveness of these 2 interventions through a 2-arm randomized trial. Rationale for a Comparative Effectiveness Study In practice, very little, if any, posttrauma support is provided to parents and children after an injury hospitalization. Health professionals and parents recognize that children can suffer a broad range of adverse posttrauma reactions, but most of these symptoms emerge after discharge. Parents and providers strongly suggest the provision of interventions in the hospital setting before discharge to support an injured child's process of emotional recovery. 49 From an ethical standpoint, our comparative effectiveness approach is justified as a design that ensures families receive some level of support. In summary, previous literature indicates that traumatic injuries are common in children. These injuries can be severe and can lead to maladaptive psychosocial effects. In response to a need to provide intervention to injured children and their parents, we conducted a comparative effectiveness study to assess the following aims: (1) assess the adoption and implementation of materials for both Link and the trauma education program by Kenardy et al; and (2) compare the effectiveness of the 2 programs on outcomes impacted by the interventions, including PTSD, depression, coping, family social support, QOL, school outcomes, and contacts with mental health professionals. Participation of Patients and Other Stakeholders Balance of Stakeholder Perspectives The following section details the contribution of stakeholder and patient partners to the design and conduct of research and dissemination of research findings. This study was conducted at 4 children's hospitals in the Midwest: University of Iowa Stead Family Children's Hospital, Blank Children's Hospital, Children's Mercy Kansas City, and Children's Hospitals and Clinics of Minnesota. Researchers engaged patient and stakeholder partners at strategic points throughout the research study to maximize their input. Before, during, and after participant recruitment and data collection, the research team engaged stakeholders in 67 unique events related to study activities. Table 1 outlines the balance of stakeholder perspectives in relation to the topic area and time period of the study. Table 1 Stakeholder Involvement by Topic and Study Period. Stakeholders Institutional stakeholder partners were individuals associated with the 4 study hospitals who had professional experience related to the study aims, and included clinical experts, health care providers, and members of hospital advisory councils. Community stakeholder partners were individuals from the surrounding community who were associated with community agencies, local health organizations, and organizations representative of the population of interest. Patient partners included patients, family members, caregivers, and former patient and parent participants. Note that patient/parent stakeholders were involved during periods when they were not actively participating in the study. Additionally, those involved in intervention development meetings did not become study participants. Methods, Modes, and Intensity of Engagement Stakeholder and patient partners were engaged through (1) focus groups, (2) consultation meetings, (3) individual interviews, and (4) a collaborative research meeting comprising researchers from the 4 hospital sites, parent-child patient partners, and clinical experts. Before commencing study recruitment at Blank Children's Hospital, we conducted 3 on-site focus groups (2 health care provider groups, 1 patient parent group) to orient institutional stakeholders to the study aims and to gather information about relevant experiences from patient partners. Health care providers were recruited into focus groups by the trauma coordinator at the hospital. Potential patient-parent participants were identified by reviewing past year's trauma admissions and inviting parents to participate in focus groups. These focus groups were intended to solidify themes identified during previous focus groups at the University of Iowa before the start of the study. During these focus groups, participants discussed the psychosocial impacts of childhood injury and resources, or lack thereof, to support families. Discussions then focused on the content of the Link materials and the feasibility of conducting a research study in a pediatric trauma level 1 setting. After focus groups were analyzed for major themes, intervention materials were revised. As a result of the focus groups, the research team developed approaches to participant recruitment from both parent and clinician perspectives. For example, 1 parent recommended a method for approaching potential participants based on his son's hospitalization: Patients and parents should be approached just before discharge and the interaction should be brief. Clinicians had similar views on this topic, but added that the approach would need to be highly individualized based on the type of injury. These recommendations were later used by the research team when determining recruitment methods. Focus group analyses also yielded recommendations from parents and clinicians that were later used to revise study materials that were distributed to participants at the time of recruitment. In addition, feedback from stakeholders reaffirmed the need to provide follow-up booster training. Major themes from the health care provider focus group highlight this aspect: I think repeated support and repeated education, multiple times and multiple ways—they're just so shut down sometimes, they're just consumed so much that it goes right past them the first time, even the second time. Repeated education—so that they know how to cope, what to do when they go home. Finally, ideas for follow-up were integrated, including identifying the optimal mode for study communication and the use of an appointment card for scheduling intervention booster phone calls and follow-up questionnaires. Focus groups were not conducted at all sites before starting study activities due to varying recruitment start dates at each site. Because study activities were underway at the flagship site in advance of these activities, focus groups and presentations to these stakeholders were intended to improve efficiency of the research staff in implementing procedures at their respective sites, as well as to modify study materials specific to each site. Additional consultation meetings with various stakeholders were conducted during recruitment and data collection periods at all sites. Institutional stakeholder partners were composed of individuals from various organizations, and included former patients and family members, staff that serve in hospital advisory roles (family advisory and youth advisory councils, nurse unit councils, and a trauma-informed care workgroup—a group of practitioners specializing in traumatic injury research), and clinicians who encounter trauma patients (such as charge nurses, hospitalists, unit medical directors, unit nurses, and hospital social workers). The purpose of these meetings was to familiarize stakeholders with the project, streamline the recruitment process, and obtain support from individuals who would be working in collaboration with interventionists to ensure that recruitment and interventions were implemented efficiently. Throughout the study period, researchers held collaborative research team meetings approximately every 2 months. Meetings were attended by research team members, institutional stakeholders, and community stakeholders from all sites. These meetings provided a platform for research team members to update stakeholders about study progress, elicit feedback and suggestions for improvement during the data collection process, discuss additional forms of data that would add to study results, and plan for collaboration on manuscripts. Following data collection, researchers conducted individual interviews with 17 individual patient partners, including former patient participants from the Link study group (4 parents and 4 children), former patient participants from the trauma education study group (4 parents and 3 children), and a parent and child who had had an injury experience but were not study participants. At the end of the data collection period, a list of parent-child dyads who had completed all follow-up study activities from each study group was generated. Researchers then contacted dyads via email or phone and asked if they would be interested in being research consultants and providing feedback about the study. In one instance, a parent who had not participated in the study, but whose spouse and child had, expressed interest in serving as a consultant, which resulted in 1 of the Link children having both parents provide feedback. Additionally, researchers interviewed 5 community stakeholder partners with diverse backgrounds: a mental health practitioner from a social service agency, a former school nurse, a school social worker, a social worker embedded in a police station, and a school coordinator providing before and after school and emergency care services. These stakeholder partners were identified through established networks at the children's hospitals or personal contacts between investigators and community partners. Patient partners were interviewed up to 2 times and community stakeholder partners were interviewed once. During the interviews, researchers presented an overview of the study, gathered feedback on study procedures, and elicited ideas for the adoption of research into practice. All interviews were conducted at the end of the data collection period and after study participation had ended for patient participants. All 22 engagement partners (patient, parent, and community partners) were compensated for their time contribution. After data collection ended at all sites, the research team hosted an all-site collaborative research and community engagement stakeholder meeting for institutional and community stakeholders and parent and patient partners. At this meeting, parent, patient, and stakeholder partners (n = 30) were presented an overview of the research project and preliminary study results. Throughout the remainder of the meeting, stakeholder, parent, and patient partners discussed both dissemination and adoption of research evidence into practice using nominal group technique. 55 Attendees produced a list of viable dissemination options, as well as ranked-ordered settings (ie, schools and health care settings) for future adaptations of each intervention. Partners were later interviewed about potential barriers and facilitators to disseminating and implementing the intervention in these settings. Examples of How Engagement Activities Changed the Research As a result of engagement with various health care providers and clinical experts before, during, and after the study, the research team was better able to develop methods to adequately recruit the targeted study population. Important feedback also allowed our team to fine-tune training materials to provide useable resources to potential participants. The data collection process expanded substantially to include a medical abstraction protocol in which researchers reviewed patient participant medical records to obtain various health indicators in order to better categorize participant injuries and extract consistent data from the medical records across sites. Finally, through the engagement process, the research team elicited valuable ideas related to dissemination of study results and adaptations of the interventions to different populations and settings. Methods Study Overview Engagement activities led to the design of the intervention trial, which evaluated early intervention strategies that were available to caregivers, specifically parents/guardians, to reduce negative psychosocial outcomes experienced by children postinjury. The aims of the trial were to compare the implementation of Link and trauma education (aim 1) and to compare the effectiveness of the 2 programs in improving psychosocial outcomes (aim 2) resulting from traumatic unintentional injury causing hospitalization among children. We examined whether these effects differed for rural and urban patients. We hypothesized that there would be minimal differences between rural and urban patients. Study Design Randomized Comparative Effectiveness Trial We conducted a randomized controlled trial to compare the effectiveness of 2 early intervention programs designed to reduce the psychosocial impacts of stress experienced among injured children who were admitted to 1 of the 4 participating children's hospitals. One University of Iowa team member not involved in recruitment used a random number generator to assign blocks of 10 consecutively enrolled participants to 1 of 2 treatment groups: Link or trauma education. To reduce the risk of bias by researchers during recruitment procedures (eg, study personnel may have guessed upcoming treatment assignments), we placed intervention assignments in a sealed envelope; these assignments were then revealed following consent. Study procedures occurred from January 2015 through February 2018. Study Setting This trial was conducted at 4 children's hospitals in the Midwest: University of Iowa Stead Family Children's Hospital (level 1 pediatric trauma center), Blank Children's Hospital (level 2 pediatric trauma center), Children's Mercy Kansas City (level 1 pediatric trauma center), and Children's Minnesota (level 1 pediatric trauma center). These hospitals were selected because they represent the sites where the most severely injured children could receive definitive care and, if hospitalized, would be recruited into the study. The 4 children's hospitals represent both rural and urban settings in the Midwest. Participants: Forming the Study Cohort Three hundred parent-child dyads were targeted to participate in the study; a description of the sample size calculation can be found in the “ Analytical and Statistical Approaches ” section. The research teams at each hospital reviewed electronic patient admission records daily to identify eligible patients and an appropriate time to approach them before discharge. All children between the ages of 10 and 17 years who were admitted for new unintentional injuries (not a complication from a previous injury) were identified for potential inclusion in the study. Children were excluded from eligibility if (1) they were admitted for intentional injuries or complications of an unintentional injury; (2) they had a diagnosis of psychosis or a severe intellectual deficit that would limit the ability to complete questionnaires; (3) they had a suicide attempt in the last year; (4) law enforcement was involved in the hospitalization (including child protective services involvement in suspected abuse cases); (5) they were pregnant; (6) they were in residential treatment or foster care; or (7) they did not speak English. Children with intentional injuries were excluded from participation because the intervention programs had not been tested on hospitalized victims of assault or self-harm, and the psychological response to these events may be different. Complications of an unintentional injury were excluded because these admissions often occur sometime after the initial traumatic incident, thus representing a different course of progression of posttraumatic stress. Intervention materials were only available in English and had not yet been translated to non-English-speaking populations. A screening log was kept using a REDCap database for both ineligible and eligible child participants. Eligible child patients and their parents were approached simultaneously by study personnel to determine interest in participation. Patients with severe neurocognitive impairments as determined by review of medical records and consultation with clinical staff were ineligible for participation. Often, only 1 parent was available when researchers approached patients for enrollment; therefore, researchers aimed to enroll only 1 parent. Description of Interventions Link for Injured Kids Parent participants received a 20- to 45-minute training (time frame was determined by parent training needs) on the Link program delivered by an interventionist. Interventionists were research team members who held a bachelor's degree and received specialized and ongoing training from a certified member of the Motivational Interviewing Network of Trainers. During the parent training, the parent watched a Link program video (viewed on a study iPad or computer), and was provided a brief training in key MI skills (ie, open questions, reflective listening skills, linking to additional resources) and in using the Kessler-6 Psychological Distress Scale (K-6). 56 The K-6 has been used as a brief 6-question scale to measure nonspecific distress in traumatized populations. A packet of study materials was given to the parent following the training session and contained (1) a Link pocket card and separate brochure containing information about basic Link skills and a copy of the K-6 screener, (2) a magnet containing a list of medical resources/crisis hotlines, and (3) a copy of the Link program video on DVD. The Link video contained demonstrated examples of the key intervention skills described previously and example scenarios of parent-child conversations. Parents were instructed to use MI skills during conversations with their child to assess psychological, emotional, physical, and cognitive reactions to trauma, assess nonspecific emotional distress using the K-6 screener, and refer their child to additional resources as needed. At 6 weeks postdischarge, the interventionist conducted a telephone booster session with the Link-trained parent. During this booster session, the interventionist reviewed key intervention skills from the initial training and conducted a brief telephone survey to assess the parent's utilization of the Link program skills and their ratings of helpfulness of key skills. Trauma Education Parent participants were given the educational booklet So Your Child Has Been in an Accident … A Book for Parents About Dealing With Accidents by a research team member (including, but not limited to, interventionists trained in the Link program). The booklet contained information about (1) children's normal reactions to accidents, (2) how long reactions last, (3) what parents can do to help their child, (4) coping strategies for parents, and (5) a list of medical resources/crisis hotlines. The content of the booklet was applicable, but not limited, to all types of accidents including falls, sports injuries, and motor vehicle accidents. 43 At 6 weeks postdischarge, the research team conducted a brief telephone survey to assess the parent's perceptions regarding the trauma education program, and an additional copy of the booklet was mailed upon request. Study Outcomes The primary outcomes ( Tables 2 and 3 ) analyzed in the study were child symptoms of PTSD using the Child PTSD Symptom Scale (CPSS) 57 , 58 and child symptoms of depression using the Center for Epidemiologic Studies Depression Scale (CES-D). 59 Secondary outcomes for both child and parent participants included assessments of pediatric QOL, 60 family communication, 61 childhood emotional/behavioral problems using the Strengths and Difficulties Questionnaire (SDQ), 62 social support using the Multidimensional Scale of Perceived Social Support (MSPSS), 63 and coping using the How I Coped Under Pressure Scale (HICUPS). 64 Additionally, assessments of psychological distress such as the K-6 screener 56 were included as secondary outcomes for child participants. Table 2 Study Outcomes: Child Participants. Table 3 Study Outcomes: Parent Participants. Follow-up Time Frame for the Study Study procedures began in January 2015 and ended in February 2018. Parent and child participants were each invited to answer 3 follow-up questionnaires over the course of the study: one at 6 weeks, one at 3 months, and one at 6 months postenrollment. This follow-up schedule was chosen to represent time points important for (1) diagnosis of PTSD (6 weeks) and (2) time schedules commonly used in the literature to monitor progression of psychosocial symptoms over time (3 and 6 months). Similar to the baseline questionnaires, the follow-up questionnaires measured depressive symptoms, symptoms of PTSD, social support, family communication, strengths and difficulties, and QOL using the validated survey instruments described previously. Additionally, the follow-up questionnaires collected information about any mental health or community resources participants may have used since their previous questionnaire. Data Collection and Sources Participants were contacted at each follow-up time point to complete a questionnaire via US mail or internet link. Those completing questionnaires via mail were mailed a questionnaire and instructed to return it to the research team using an enclosed postage-paid envelope. Those completing the questionnaires via internet were emailed an automated survey reminder that contained an internet link to an electronic questionnaire. All questionnaire data were stored in a REDCap database. Data from online surveys were automatically entered into the REDCap database, while paper surveys returned by mail or completed by phone were entered by a study team member into the REDCap database. The research team contacted each participant via phone on the date that questionnaires were sent to inform them that questionnaires would be arriving, and to provide an opportunity to complete the questionnaires by phone at the time of the call. The research team contacted participants up to 12 times at each survey time point, or up to 1 year following consent, to remind them to complete questionnaires. After a month of no response at each time point, the questionnaires were re-sent via mail or email. Participants who did not complete a questionnaire at a specified follow-up time point were classified as “missing” for that time point and allowed to continue participation by completing questionnaires at subsequent follow-up time points. Participants who informed the research team that they did not wish to continue participation were classified as “withdrawn.” Self-administered Surveys Participants completed questionnaires at baseline, 6 weeks, 3 months, and 6 months. Participants were compensated up to $200 per dyad for the completion of all questionnaires. The questionnaires were composed of basic demographics (asked at baseline only; eg, sex, age, grade, race/ethnicity, education, income, and employment), previous/recent stressors, coping, family social support, family communication, communication about reactions to trauma, and additional mental health resources sought. Medical Abstraction The research team completed a medical record abstraction for each child participant, related only to the child's unintentional injury hospital admission. The data collected from the medical record included the following: Basic demographics (ie, date of birth, sex, weight, height, and body mass index [BMI]) Mechanism of injury the child sustained (collected from either International Classification of Diseases, Ninth Revision ( ICD-9 ) or Tenth Revision ( ICD-10 ) external cause of injury codes depending on ICD coding schemes used by each of our hospitals) Comorbidities (ie, history of depression, anxiety, and attention-deficit/hyperactivity disorder) Glasgow Coma Scale measures (to assess level of consciousness) Initial vital signs and pain ratings Details of hospital admission (ie, mode of arrival to hospital, length of stay, days in intensive care unit, days on mechanical ventilation, and trips to the operating room) Toxicology screening results Medications administered in the first 48 hours of admission Blood products administered for the duration of admission Injury characteristics (ie, abbreviated injury score, injury severity score, ICD-9 or ICD-10 nature and body location of injury codes, and ICD-9 or ICD-10 procedure codes for entire duration of admission) Rural Indicators We used Rural Urban Commuting Area (RUCA) codes 67 to classify families by residence zip code, as rural/urban residence was to be tested as a prespecified possible modifier of intervention effectiveness. RUCA classifies zip codes into urban and rural areas using work commuting data and proximity to cities and towns of different population densities. Specifically, the following codes classified urban locations: 1.0, 1.1, 2.0, 2.1, 3.0, 4.1, 5.1, 7.1, 8.1, and 10.1. Rural locations corresponded to the following codes: 3.0, 4.0, 4.2, 5.0, 5.2, 6.0, 6.1, 7.0, 7.2, 7.3, 7.4, 8.0, 8.2, 8.3, 8.4, 9.0, 9.1, 9.2, 10.0, 10.2, 10.3, 10.4, 10.5, and 10.6. Patient-Centered Implementation Measures To assess the acceptability and usability of the 2 intervention strategies, we measured indicators of implementation informed by the RE-AIM (Reach Effectiveness Adoption Implementation Maintenance) framework, used in implementation evaluation research. 67 , 68 Implementation refers to the extent to which an intervention is delivered as intended. Study participants who were randomly assigned to trauma education were asked, “Since your last survey, how many times did you use what you learned in the trauma education booklet (‘So You've Been in an Accident')?” Response options included “I didn't read it,” “I read it, but didn't use it,” and “I used it X times” (where X ranged from “1” to “7 or more”). Individuals were classified as nonimplementers if they reported “I didn't read it,” partial implementers if they reported “I read it, but didn't use it,” and full implementers for all other response options. Participants randomly assigned to Link were asked the following 3 questions: “Since your last survey, did you ever use the ‘How/What/Tell Me More' Skill?”; “Since your last survey, did you ever use the ‘Reflections' Skill?”; and “Since your last survey, did you ever link your child with outside sources of support?” Yes/No response options were provided for each of these 3 questions. Individuals were classified as nonimplementers if they responded “No” to all 3 of the listed questions, partial implementers if they reported “Yes” to at least 1 but not more than 2 of the 3 questions, and full implementers if they responded “Yes” to all 3 questions. Booster Call Assessments During 6-week follow-up phone calls, parents in each study group were asked to estimate the number of times they referred to written materials, the number of times they used intervention strategies while communicating with their children, and the length of those conversations. Parents were asked to provide an overall rating of how helpful these strategies were, using a 1 to 10 scale with 1 being not at all helpful and 10 being very helpful. Finally, parents were asked open-ended questions about the coping strategies offered to their children and the challenges and facilitating factors in using the intervention materials. Intervention Fidelity Testing To ensure intervention fidelity among the research team, Link training and booster sessions were audio recorded with the participant's consent, and 20% of audio files were reviewed using the Behavioral Change Counseling Index (BECCI) by a research team member for adherence to the MI style of communication. The BECCI measures practitioner competence before, during, and after training in behavior change counseling. 69 For this study, an adapted BECCI instrument was used for scoring, and included ratings of 9 constructs of MI scored on a continuous scale of 0 to 4 (where 0 = “not at all” and 4 = “a great extent”). Interventionists were provided continued mentoring in MI skills throughout the course of the study. Child Reported Measures of Communication During the 3- and 6-month surveys, child participants estimated the number of times parents communicated with them about their feelings after the traumatic injury, their comfort level while communicating with their parents, and the perceived helpfulness of these conversations (scale 1-10, where 1 = “not at all successful” and 10 = “very successful”). Analytical and Statistical Approaches Sample Size Calculation Focusing on PTSD symptoms as the primary outcome, we calculated sample sizes based on a moderate minimally important effect size of 0.5 SD (mean difference = 4.06) comparing trauma education and Link. With α = .05, a sample size of 134 parent-child dyads in each of the groups (N = 268) would have 100% power to detect a 0.5-SD true difference. To account for potential loss to follow-up, we also oversampled by 12% for a targeted sample size of 300 dyads, of which 144 were projected to come from a rural community. The projected value of 144 dyads was based on the estimated proportion of pediatric rural trauma patients treated in the 2 children's hospitals in the state of Iowa (ie, approximately 48%). The estimated proportion is derived from preliminary trauma data collected by the Iowa Department of Health. This subsample size of 144 rural children would be large enough to detect a difference in the impact of our interventions in a strictly rural sample. Aim 1: Implementation Evaluation We first defined a categorical measure of implementation level (none, partial, and full implementers) based on parent-reported use of the Link intervention. For the trauma education group, we created implementation levels based on the number of times the provided materials were used. For trauma education, parents who reported having never read the provided materials were categorized as nonimplementers; having read but not used the material defined the partial implementation group, and having read and used the material at least once characterized the full implementation group. For Link, we determined implementation based on the number of times parents talked to their injured child using the skills they learned. Individuals categorized as nonimplementers were those who reported using none of the skills, partial implementers were those who reported using at least 1 of the skills, and full implementers were those who reported using all of the skills. Descriptive statistics were used to determine the counts in each category at each time point for each intervention group. To capture the ordinal nature of the outcome and implementation level, as well as accommodate subject-level clustering, we used a generalized linear mixed model (GLMM) with a multinomial distribution and cumulative logit link function. This model included a covariate for group assignment (Link or trauma education), measurement of time, and an interaction between the two. Time was represented in models as the number of weeks post baseline. We did this due to the general lack of alignment between dates of survey completion and planned time points (ie, 6 weeks, 3 months, and 6 months). Additionally, this representation of time was consistent with that used for analyses performed for aim 2. Aim 2: Outcome Evaluation Each outcome variable of interest—PTSD, depression, psychological distress, QOL, and the SDQ measures—was first assessed for normality through use of descriptive statistics. Through this assessment, it was determined that the outcome data were extremely skewed, either right or left. In order to accommodate the observed nonnormality of the data, we used GLMMs. For each outcome, we specified a log link and γ distribution for the respective GLMM. We used a log link primarily for computational tractability, while specification of the γ distribution enabled modeling of the right-skewed data. For those outcome data that were left-skewed, the corresponding scaling was reversed (eg, value of 0 recoded as 4 and vice versa) so that the skewness was flipped. Given that the domain of the γ distribution excludes zero, all outcome scales whose scaling included zero were shifted upward by 1 unit. Posttransformation, the scales for each outcome were as follows: CPSS (scale range, 1-52), CES-D (scale range, 1-37), child QOL (scale range, 1-101), child family communication (scale range, 10-50), child SDQ (scale range, 4-44), child SDQ subscales (scale range, 1-11), MSPSS (scale range, 1-5), HICUPS (scale range, 4-68), and K-6 (scale range: 6-30). Each posttransformation outcome scale was such that improved outcomes were indicated by lower scores. The transformed parent scales were similar in range and interpretation to those corresponding to child participants. Only data collected postbaseline were used in the fitting of models. We excluded baseline data to induce linearity in each outcome's trend. Most outcomes exhibited a sharp uptick in the first measurement postbaseline (eg, 6 weeks) followed by a steady decrease over the remaining time points. Given our use of GLMMs, this nonlinear trajectory could not be properly captured. Inclusion of baseline measures as covariates was considered, but they were found not to explain any additional variability in the data. Model fitting included a covariate for group membership (Link or trauma education). Because outcomes are expected to change over follow-up, time was included as an explanatory variable, as well as the interaction between group membership and time. Inclusion of the interaction allowed for the determination of a differential effect between treatment groups over time for each outcome. Due to the general lack of alignment between dates of survey completion and planned time points (ie, 6 weeks, 3 months, and 6 months), time was represented in models as the number of weeks postbaseline. Based on evaluation of the Akaike information criterion, a commonly used statistical model selection criterion, we found this continuous measure time (ie, weeks postbaseline) to be favored over categorical measures of time (eg, first survey, second survey, etc). We also conducted subanalyses, specified a priori, to examine heterogeneity of effects for rural/urban patients. To examine the heterogeneity of effects for rural/urban patients across treatment groups over time, we fit models containing a 3-way interaction between intervention, time, and rural/urban status. These models also included all lower-order interactions. Statistical significance of the 3-way interaction term is indicative of a differential time effect between treatment groups that is moderated by whether patients reside in rural or urban locations. Missingness was attributed to 2 sources: partially complete surveys and missing surveys. For those surveys that were partially complete, responses to select questions among those that defined our outcome measures were missing. In contrast to these partially complete surveys, missing surveys were those in which a participant failed to submit any response to any question on a given survey. Among the child population the total missingness from partially and completely missing surveys was 28.16% (depression), 27.22% (PTSD), 34.98% (QOL), 23.63% (psychological distress), 29.95% (SDQ), 24.4% (SDQ-Emotional Symptoms), 24.49% (SDQ-Conduct Problems), 25.09% (SDQ-Hyperactivity), 25.43% (SDQ-Peer Relationship Problems), and 23.81% (SDQ-Prosocial Behavior). For the parent population these values were 25.00% (depression), 29.98% (PTSD), 37.87% (QOL), 22.51% (psychological distress), 26.33% (SDQ), 23.75% (SDQ-Emotional Symptoms), 22.92% (SDQ-Conduct Problems), 23.92% (SDQ-Hyperactivity), 23.75% (SDQ-Peer Relationship Problems), and 23.01% (SDQ-Prosocial Behavior). To address missingness arising from the partially complete surveys, single-chain Monte Carlo Markov chain multiple imputation with 300 iterations (200 burn-in) and Jeffrey's prior was used. For each group of questions that defined a particular outcome, observed data within the question group were used to impute the missing responses. Five imputed data sets were generated for each outcome and used in the fitting of the GLMMs described previously. Sensitivity analyses comparing results from the imputed data to those obtained through a complete case analysis were conducted. We used the GLMM framework to accommodate missingness attributed to the missing surveys. We assumed these data were missing at random. This assumption was justified through testing for differences in baseline characteristics between those observed and those missing at each time point for each intervention group and finding a rejection rate compatible with the type I error rate of 0.05. Study Conduct/Protocol and Amendments In 2016, 3 additional sites were added to the study protocol to recruit the targeted sample of participants (N = 300 dyads). The University of Iowa served as the coordinating site for this research and provided oversight to participating sites that were added. All study procedures conducted at the University of Iowa Stead Family Children's Hospital and Children's Mercy Hospital Kansas City were reviewed and approved by the University of Iowa IRB (ID number 201111728, reliance number 14110518). All study procedures conducted at Blank Children's Hospital were reviewed and approved by UnityPoint Health-Des Moines IRB (ID number IM2014-029). All study procedures conducted at Children's Minnesota were reviewed and approved by Children's Minnesota IRB (number 1602-010). This study is registered at ClinicalTrials.gov (ID NCT02323204 ; unique protocol ID 201111728). Results Study Participants Figure 1 details the flow of participants for the study. A total of 1688 children between the ages of 10 and 17 years were admitted to 1 of the 4 children's hospitals because of a possible injury during the study recruitment period (January 2015 to July 2017) and were screened for eligibility into the study. A total of 1374 children were excluded, of whom 893 did not meet study criteria (“ see “ Participants: Forming the Study Cohort ” for ineligibility criteria), 263 declined participation, and 243 were missed recruits (eg, missed, no staff available, and/or transferred to another medical facility). Exclusion categories were not mutually exclusive; screened patients often met multiple exclusion criteria. Figure 1 CONSORT Flow Diagram. A total of 314 children and their parents enrolled in the study, out of a total of 795 eligible for inclusion ( Table 4 ). This corresponded to a 39.5% participation rate. Using randomization, we allocated 155 (49.4%) dyads to the Link intervention and 159 (50.6%) dyads to the trauma education intervention. For the Link group, 145 (93.5%) children and 146 (94.2%) parents completed baseline training and questionnaires. For the trauma education group, 154 (96.9%) children and 156 (98.1%) parents received the trauma education materials and completed the baseline questionnaire. Baseline questionnaires were not submitted by 3 children (1 Link and 2 trauma education), whose parents received intervention training before the child completing the baseline questionnaire. The intervention was delivered with the intent to collect the information before discharge; however, these child participants ultimately never completed baseline surveys after multiple attempts at contact from the research team. They were included in analysis because they provided data from follow-up questionnaires and medical records. Table 4 Completion Rates of Administered Surveys. Booster Trainings At follow-up, 103 parents (72%) completed Link booster training and 120 (76.9%) parents completed trauma education placebo booster training ( Table 4 ). Adherence to MI communication styles during initial and booster training sessions was found to be high, with a mean BECCI score of 25.8 and a range of 13 to 34, of 36 possible points. A total of 36 training sessions were rated (20% of 179 recorded intervention sessions) by at least 1 rater, and 9 training sessions (20% of the 36 intervention sessions scored by the first rater) rated by 2 raters. Given that the scoring metric was continuous, we computed an intraclass correlation coefficient (ICC) to assess the interrater reliability between the 2 raters. In the computation of the ICC, we assumed 2-way mixed effects, consistency, and single rater measurement. 70 The obtained ICC was 0.838, suggesting good agreement between raters for these 9 observations. For this report, our focus was not on the measures of communication or intervention helpfulness data. Rather, we focused on the degree of intervention implementation in aim 1 as well as the presented primary and secondary outcomes for aim 2. Follow-up Questionnaires Completion rates for follow-up questionnaires for Link intervention dyads were 75.3% for parents and 71.7% for children at 6 weeks, 67.8% for parents and 62.8% for children at 3 months, and 65.8% for parents and 66.2% for children at 6 months. For the trauma education group, completion rates were 76.3% for parents and 72.7% for children at 6 weeks, 71.8% for parents and 73.4% for children at 3 months, and 71.8% for parents and 73.4% for children at 6 months. Study Population Characteristics Both Link and trauma education groups exhibited similar characteristics at baseline ( Table 5 ). The mean age (SD) was 14.04 years (2.16) among Link children and 13.82 years (2.57) among children randomly assigned to the trauma education group. In both Link and trauma education, the child population was predominantly male (Link, 64.1%; trauma education, 61%). This is in contrast to the distribution of sex in the parent population, which was primarily female (Link, 78.1%; trauma education, 78.8%). With respect to participant ethnicity, the majority of the study population identified as non-Hispanic or non-Latino (Link, child: 93.8%; Link, parent: 93.8%; trauma education, child: 90.3%; trauma education, parent: 94.2%). Parent-child dyads in each intervention group identified primarily as White (Link, child: 86.2%; Link, parent: 87.7%; trauma education, child: 85.7%; trauma education, parent: 87.8%), and the distributions of race across intervention groups were similar. We defined urbanicity based on reported zip codes corresponding to enrolled participants' primary residence. RUCA codes were used to categorize each zip code as corresponding to either a rural or urban location. Based on this measure, our study population resided predominantly in urban locations (Link, child: 64.8%; Link, parent: 65.1%; trauma education, child: 62.2%; trauma education, parent: 61.7%). At baseline, average CPSS (PTSD) and CES-D (depression) scores were relatively low (which is indicative of a better outcome) on the study-defined scale (CPSS scores range from 1 to 52; CES-D scores range from 1 to 31), but the variability in score was high. Comparing intervention groups within the child-parent populations, we observed similar means (SD) at baseline (eg, PTSD—Link, child: 9.97 [7.91]; trauma education, child: 9.95 [8.31]; Link, parent: 8.21 [7.87]; trauma education, parent: 7.61 [7.62]). For these comparisons, all P > .05. The patterns exhibited by the primary outcomes of interest at baseline were mirrored by the secondary outcomes in that there were no differences between intervention groups for each of the secondary outcomes (again, all P > .05). Baseline characteristics between those observed and those with missing data were compared at each time point. Chi-square and Wilcoxon rank sum tests were used for categorical and continuous variable comparisons, respectively. A total of 90 tests were performed: 3 (each corresponding to a survey time point) for each of the 15 baseline characteristics for both the child and parent data sets. A rejection rate compatible with the type I error rate of 0.05 was found (ie, 4/90 rejected), suggesting a lack of difference in baseline characteristics between those observed and lost to follow-up at a given time point. Table 5 Demographics and Outcome Measures at Baseline. Aim 1: Implementation of Link and Trauma Education Table 6 characterizes the degree of reported implementation (DRI) at each survey time point, for each intervention group. The DRI at each survey time point reflected the degree of implementation between the current and previous survey time points (eg, DRI for the 6 week survey describes implementation since enrollment). Data on implementation was reported only through surveys administered to parents enrolled in the study. Although the majority of enrolled participants reported their level of implementation, some individuals did not report implementation. Percentages displayed in Table 6 are therefore with respect to those who reported implementation, rather than the entirety of individuals who received each treatment. Based on the 6-week survey, the majority of Link participants who reported implementation were classified as partial implementers (58.1%), whereas parents who received the trauma education intervention were primarily classified as full implementers (60%). This trend remained consistent over the course of the study. Looking across each of the administered surveys, we observed a slight reduction in the degree of reported implementation postbaseline for each intervention group. This is evidenced by an increase in the counts (percentages) of individuals classified as nonimplementers and decrease in the counts (percentages) of individuals classified as full implementers. To fully characterize the relationship between time, intervention group, and the degree of reported implementation, we fit a cumulative logit mixed model. Reported implementation was defined as the response, and time, intervention, and the interaction between them were used as covariates. A representation of the fitted model is provided in Table 7 . Use of a cumulative logit model was motivated by the ordinal nature of implementation. Table 6 Degree of Reported Implementation by Intervention. Table 7 Degree of Reported Implementation: Cumulative Logit (CL) Model Results. Model estimates indicate a decrease in implementation over time (time = −0.04; 95% CI, −0.07 to −0.02; P = .0025) and generally lower implementation in the Link group compared with the trauma education group (intervention = −1.07; 95% CI, −1.92 to −0.22; P = .01). However, there is a lack of evidence to support a differential time effect by intervention (intervention × time = 0.01; 95% CI, −0.04 to 0.05; P = .79). Comparing intervention groups at each administered survey time point, we observe that the odds that a Link participant was a full implementer as opposed to a partial or nonimplementer is 0.35, 0.37, and 0.39 times that of a trauma education participant at 6 weeks, 3 months, and 6 months, respectively. Given the proportional odds assumption made in the fitting of the cumulative logit model, these odds ratios remain unchanged when considering full or partial implementers vs nonimplementers. Aim 2: Outcome Evaluation Table 8 provides sample means (SD) computed from data provided by each of the 3 surveys administered postbaseline (6 weeks, 3 months, and 6 months). Computed means for each intervention group exhibit a decreasing trend across surveys for each of the primary outcomes of PTSD and depression. However, as observed in the means computed at baseline, the data are highly variable. Computed means for each secondary outcome generally exhibit a decreasing trend, with a few exceptions (eg, SDQ-Peer Relationship Problems—trauma education, parent: 2.46, 2.67, 2.76). For each study population, child and parent, we fit GLMMs with a log-link and γ distribution function to model each outcome. A γ distribution was specified for fitted models to accommodate the right-skewed distributions of each outcome. A log-link was specified for computational tractability. Time (in weeks), intervention, and the interaction between the two were included as covariates. Table 9a reports the time effects associated with each study intervention, based on the imputed child-reported data. Analyses were conducted with and without imputation, but the results did not differ meaningfully. Table 8 Mean (SD) of Each Outcome Measure at Each Administered Survey Time Point. Table 9a Estimated Time Effect by Intervention—Imputed Child-Reported Outcomes. For the primary outcome, both interventions trended toward improvement over time. However, there is no evidence of a differential effect by intervention (PTSD = 0.987; 95% CI, 0.94-1.04; P = .6298; Depression = 1.006; 95% CI, 0.96-1.05; P = .7893). For the secondary behavioral outcomes, assessed via the SDQ measures, Link also trended toward improvement over time, while trauma education had mixed effects across each SDQ measure. Given this disparity between Link and trauma education trends, we do find evidence of a difference in time effect by intervention when looking at SDQ-Conduct (0.971; 95% CI, 0.94-1.00; P = .0286) and SDQ-Peer Relationship Problems (0.965; 0.93-1.00; P = .03) specifically. The differences in effect for these outcomes indicate slightly greater improvement in the Link group over time compared with the trauma education group. Within the trauma education group, the largest effect observed was for the QOL measure (0.872) and the smallest was for the SDQ-Emotional Symptoms and Hyperactivity measures (each 0.996). These estimates can be translated to outcome reductions of 56% and 2.4%, respectively, over a 6-month period. For the Link group, the largest effect observed was also for the QOL measure (0.863) and the smallest for the SDQ-Hyperactivity measure (0.995). These translate to 6-month reductions of 58.7% and 3%, respectively. Figure 2 displays the model-predicted trajectories for each intervention group. The figures show the range over which change was observed, rather than the full scale of each measure. For depression and PTSD (displayed in the upper panels of the figure), a gradual reduction in score was observed with each intervention, and the predicted lines are mostly parallel. Note that for each outcome of interest, a reduction is indicative of improvement. The bottom panels, which display the trends for SDQ-Conduct Problems and SDQ-Peer Relationship Problems, demonstrate a gradual increase in score for trauma education (which is indicative of a worsening in condition), whereas Link exhibits a slightly decreasing (improving) trajectory. Figure 2 Trajectory of Select Imputed Child-Reported Outcomes . Parent-Reported Outcomes for Children and Themselves Results were generally consistent between imputed parent-reported data ( Table 9b ) and imputed child-reported data with a single exception. As was the case for the imputed child-reported SDQ-Conduct Problems, there is evidence of a differential time effect by intervention group (0.967; 95% CI 0.94-0.99; P = .0201). However, we did not find evidence of a differential time effect for SDQ-Peer Relationship Problems (0.982; 0.95-1.01; P = .2568), as was the case for the child study population. Table 9b Estimated Time Effect by Intervention—Imputed Parent-Reported (of Child) Outcomes . Table 10 displays the time effects corresponding to outcomes that pertain to the parents, rather than their children, for each intervention group. Both Link and trauma education suggest improvement over time for PTSD (CPSS), depression (CES-D), and psychological distress (K-6), but the rate of improvement over time does not differ between each intervention group (CPSS = 0.973; 95% CI, 0.93-1.02; P = .23; CES-D = 0.979; 95% CI, 0.94-1.02; P = .2951; K-6 = 1.012; 95% CI, 0.97-1.05; P = .5381). As we did for the child analyses, we used both imputed and nonimputed parent data for the parent analyses. Results did not differ meaningfully. Table 10 Estimated Time Effect by Intervention—Imputed Parent-Reported Outcomes . Heterogeneity of Treatment Effect: Rural/Urban Status Model estimates of the 3-way interaction term between treatment, time, and rural status were obtained for each outcome using each set of imputed data (child- and parent-reported). With 109 of 110 (child/parent) participants residing in rural areas, we achieved approximately 85% of our targeted sample size (128 rural participants). Given a rural sample short of our intended recruitment, subgroup analyses investigating the heterogeneity of treatment effect were underpowered relative to the planned power of 81%. Table 11 provides imputed model estimates of the 3-way interaction between time, intervention, and rural status for each of the parent- and child-reported outcomes. For child-reported data, the 3-way interaction term was significant only for the SDQ-Peer Relationship Problems outcome (estimate = 0.017; 95% CI, 0.00-0.03; P = .04), suggesting that the effect was differential between urban and rural children for this outcome. Based on these results, there was a slightly stronger effect for rural than urban children. For the parent data, there was no evidence in support of a differential rural/urban treatment effect. Results between imputation and nonimputation analyses did not differ meaningfully. Table 11 Time-Intervention-Urbanicity Interaction Model Estimates. Effectiveness of Link by Level of Implementation We performed post hoc analyses that assessed the effect of increased intervention implementation on each defined outcome. Analyses included fitting models similar to those defined in each main analysis, but with the Link intervention categorized into various dosing groups defined by implementation. These analyses indicated a lack of evidence in support of a dose-response relationship within Link. In addition, analyses were performed in which implementation was added and interacted with the set of covariates contained in the previously defined outcome models. The 3-way interaction defined by implementation, time, and intervention was found to have no statistical significance. Discussion Context for Study Results Traumatic injury can have long-lasting effects on children. The need for psychological care of injured children has been expressed strongly by parents, guardians, and caregivers. In this study, we used the hospital as an ideal setting to test the effectiveness of interventions for severely injured children. Four children's hospitals were the sites for intervention, and the level of acute PTSD in this sample was relatively low, with about 30% of children at baseline having acute PTSD symptoms. These interventions were implemented with their parents/guardians, who would be supervising their recovery when they were released from the hospital. We offered 2 interventions of differing intensities. One was Link, a more involved intervention that required approximately 20 to 45 minutes to train parents in MI techniques, impacts of trauma, screening for stress, and linkage to services. The second intervention was trauma education, a low-cost program involving the provision of reading materials about trauma and its effects and linkage to services. Differences in Implementation Link, the more intensive intervention that required use of MI skills, was reportedly implemented to a lesser extent than trauma education. Link is a complex intervention with multiple components (MI, screening, and referral). Thus, parents were more likely to be partial implementers (54%-59%) rather than full implementers (17%-26%) of Link, and the majority of parents reported using some, but not all, Link components. The flexibility in implementation is reasonable given how different children's symptoms, needs, and communication styles might be. Ultimately, for an intervention that is child-centered, it is by discretion of the parents to determine if they would use the screener or seek additional supports for advanced mental health care. On the other hand, many more parents in the trauma education group reported full use (reading and reviewing) of the intervention materials. Trauma education is therefore a relatively simple intervention to implement, and most parents across the study period (60% at 6 weeks, 43% at 3 months, and 39% at 6 months) reported full utilization of intervention materials. Over time, we also found that implementation did decrease rather equally for each group, perhaps reflecting less perceived need for any intervention by parents. Differences in Outcomes Overall, there were no differences in the effectiveness of Link and trauma education. Both were equally effective in reducing PTSD and depression (primary outcomes), as well as improving secondary outcomes (QOL, nonspecific stress, and overall behaviors/emotional problems). However, Link was found to slightly improve selected problem behaviors, especially conduct and peer behaviors. The comparable reductions in child PTSD and depressive symptoms over time support that no harm was associated with use of these interventions. Findings were consistent in child measures as reported by the child or the parent. These findings are also rather consistent with our previous studies of Link in a school setting 54 and of trauma education in a hospital setting. 43 From our study, we were able to quantify the reduction in psychosocial maladjustment. Each month, children whose parents were trained in trauma education had an approximately 7.6% reduction in PTSD symptoms and a 3.9% reduction in depressive symptoms. Children whose parents were trained in Link had an 8.8% reduction in PTSD symptoms and a 3.3% reduction in depressive symptoms per month. Over our 6 month study period, these reductions could be substantial, leading to >40% reduction in PTSD and >18% reduction in depressive symptoms. While these reductions were substantial, we emphasize that there were no meaningful differences in the effectiveness of these interventions in our sample of injured children. Furthermore, without an untreated control group, we do not know if these reductions could have occurred naturally without any intervention. There was also no substantive difference in secondary outcomes of interest when comparing our 2 intervention groups. We saw similar improvements in pediatric QOL (12.8% improvement per month for trauma education and 13.7% improvement for Link children), and in nonspecific stress (6.1% reduction per month for trauma education and 5.2% reduction per month for Link children). There was no overall difference in the overall SDQ, a global measure of child behaviors, emotions, and relationships. SDQ is considered a screener for risk of psychiatric disorders, and has been indicative of general child emotional and behavioral health. 71 , 72 There was, however, a slight difference in measures of child behavioral health, measured by 2 subscales of the SDQ (specifically the Peer Relationship Problems and Conduct Problems measures). Children whose parents received Link had a 2.6% reduction in problem behaviors, whereas children whose parents received trauma education had no improvements overall. Because we ran multiple tests, it is possible that these were chance findings. Still, these are very modest reductions, and further research is needed to explore these relationships in a potentially larger sample focused specifically on child behaviors. However, it is promising that these results were rather consistent in child- and parent-reported measures of child conduct behaviors. Given these modest findings, there are some potential implications for studies focused on improving child problem behaviors, especially for youth who require adjustment to their postinjury health conditions. Often, injured children exhibit problem behaviors as reactions to trauma, and in some research are reportedly misdiagnosed with attention-deficit/hyperactivity disorder or other behavioral psychopathologies. 75 Future studies may consider the traumatic roots of problematic behaviors, which often escalate to adjustment problems, social isolation, delinquency, and aggression. 76 , 77 We recommend studying the effectiveness of Link for Injured Kids as a potential intervention for problem behaviors that may be caused or exacerbated by traumatic injury. Finally, to our surprise, degree or dosage of implementation did not impact the effectiveness of a program like Link. First, it should be noted, however, that because the interventions differed in how they would be delivered in real-world practice, we could not use the same measures of implementation. Trauma education involved reading the intervention content, while Link involved using communication skills with the injured child. Implementation must be measured through reported behaviors of “reading” vs “talking” that align with anticipated intervention activities. Thus, the head-to-head examination of effectiveness is imperfect. Our finding of no difference in the comparative effectiveness of the interventions by level of implementation is not surprising. The noncomparability of our measures of implementation is a limitation in comparative effectiveness research. Therefore, measurement of implementation is an area for further exploration and study when comparing interventions that differ significantly in terms of their intervention activities. Second, this evaluation may have been confounded by the complexity and flexibility of Link, a child-centered program that is designed to be flexible, and therefore does not require full implementation of all steps. For example, parents may choose not to use the screening tool or refer their child to additional resources based on the initial conversations. Thus, partial utilization of Link was as sufficiently effective as full utilization, and likely led to null impacts of implementation. Generalizability of Findings Our study is limited to the study populations reached at the 4 children's hospitals in the Midwest. Our sample had a reasonable number of rural patients and was predominantly White, which is representative of the general Midwest population, and especially the state of Iowa, where most of our trauma patients were enrolled. We restricted our eligibility to children who suffered an unintentional injury. Thus, we do not know if Link and trauma education could work for children admitted for intentional injury. We also cannot generalize our findings to patients who may sustain less severe injuries such as those seen in emergency department (ED) or outpatient settings. Implementation of Study Results Results from this study suggest that use of trauma education, a low-cost and low-level intensity intervention, is just as effective in reducing PTSD and depression as Link for Injured Kids, a more costly and higher-intensity intervention. This finding has important implications for dissemination and application of Link and trauma education in real-world settings. First, there are costs associated with the time and expertise needed to train families in Link; it is thus not so easy to implement Link in a hospital setting without available resources, particularly when training parents in MI skills. Hence, more parents reported partial rather than full implementation of Link. However, Link is a flexible program that, through natural conversations with traumatized children, may use all or some MI skills (eg, open-ended questions, reflections), a screening tool, and/or referrals to advanced care. On the other hand, trauma education is simple to provide and the only costs associated with its use would be in tailoring resource/referral lists that are specific to a hospital. Offering trauma education to all families of traumatically injured children is a low-cost, feasible intervention in any children's hospital across the country. Subpopulation Considerations Aside from what was observed via the child-reported SDQ-Peer Relationship Problems outcome, we did not find any differences in the comparative effectiveness of both interventions by rurality, although our analyses were slightly underpowered. The one observed result for SDQ-Peer Relationship Problems (estimate = 0.017; 95% CI, 0.00-0.03; P = .04) suggests that, observed over time, the Link intervention is more effective in rural than urban children. The analyses presented thus far are considered population-averaged effects, and provide overall effect estimates by group assignment. There was considerable heterogeneity, and therefore variability, in outcomes. Given this limited data set and the obvious noise in the data, future larger-scale studies are needed to replicate and sharpen these findings. Patient Engagement Stakeholder engagement before to, during, and following the study period was a major source of support for this research endeavor. Community stakeholders both within and outside the hospital setting assisted researchers in refining the research process and ensuring quality data collection. Engagement activities following the study period allowed patient stakeholders and researchers to collaboratively examine and evaluate the research protocol and the preliminary research findings. Patient stakeholders discussed the applicability of the research protocol to a wide variety of other populations and settings, and were eager to share their experiences and thoughts about improving and expanding the Link and trauma education interventions in future studies. Study Limitations Retention rates in the study ranged from 65.8% to 76.3% across the 3 follow-up time points at 6 weeks, 3 months, and 6 months. These rates of retention are comparable to rates in similar types of follow-up intervention studies of pediatric trauma patients in the literature. 43 , 45 , 76 Notably, we found no appreciable differences in sample characteristics between those at baseline and the sample at each follow-up time point, as described previously. It is possible that there are some unmeasured characteristics that are unbalanced across our intervention groups, which could lead to bias. Direct intervention is not applied by the research team to the child; rather, the focus is on training the parent to implement the intervention or on providing education about trauma to the parent. This leaves substantial variability in intervention delivery. Fidelity of intervention delivery was not directly measured; instead, we relied on parents' self-reported intervention implementation. This recall bias could lead to overreporting use of either intervention, especially trauma education, in which reported implementation is simplified into a measure of reading or referring to materials. We initially explored with our parent consultants the possibility of audio recording, specifically Link intervention delivery by parent to child; however, audio recording parent-child conversations was viewed as extremely invasive. We did not diagnose children for psychiatric disorders, which if used, would allow us to measure actual mental health psychopathology. However, our interventions were delivered by parents, who are generally not therapists. We therefore anticipated effects to be confined mostly to symptoms and behaviors, and relying on child- and parent-reported symptoms was thus appropriate. To address potential concerns about recall bias or underreporting/overreporting of outcomes especially, we conducted surveys with both parents and children, and found consistent intervention effects using both parent- and child-reported outcomes. Because this was a comparative effectiveness study, we did not have an untreated control group. Thus, we cannot determine with certainty if reduction in stress/PTSD/depressive symptoms and behavioral problems could have occurred through the natural course of recovery from trauma. We are encouraged by the a priori randomized controlled trial of trauma education that used a no-treatment control group and reported improved anxiety in similarly injured children in Australia. 43 Also, in a recent meta-analysis of injured children between 6 and 17 years of age without any interventions, 25% of this heterogeneous sample had symptoms of PTSD. 77 In this review, children hospitalized or seen in an ED for their injuries experienced considerably higher rates of PTSD symptoms (some as high as 37%) than those in our study. 77 In a separate but more recent study of hospitalized injured children, 43% had symptoms 4 to 7 weeks after trauma. 14 The level of PTSD symptoms in our study population was 28% at 6 weeks but was reduced to 17% at 6 months. Our levels are notably lower than those in studies of PTSD symptoms among severely injured children, suggesting, although not testable in this study, a possible protective effect of our intervention in preventing onset of PTSD symptoms. Another possible interpretation is that our findings are specific to our study population; our liberal eligibility criteria differed from other studies that required at least a 1-night hospital stay. We recruited all patients with admissions regardless of length of stay. Third, although the lack of a control group still represents a limitation, we point to priorities expressed by parents, providers, and injured children themselves in previous research and among those engaged in our study to justify our design. This comparative effectiveness study that used 2 interventions (Link was active, while trauma education was passive) responds directly to the needs and preferences of our patient partners, expressed in previous research 49 and by the parents, children, and clinicians engaged directly in our study. Providing no intervention would not align with a patient-centered approach. With the help of stakeholders, our research questions were designed to meet an ultimate goal of finding interventions that support parents and injured children, hasten the recovery process, and prevent and reduce the probability of further psychological, social, and behavioral damage. Finally, we note that this study is one of the first to include a focus on injured rural children. The access to and availability and acceptability of mental health services in rural settings is extremely limited, 9 , 24-28 which can result in increased risk or exacerbation of trauma symptoms and challenges to psychological recovery. While our findings report no differences in the effectiveness of our 2 interventions in reducing PTSD and depressive symptoms between rural and urban settings, the provision of trauma education can still be justified as an inexpensive resource to provide to rural parents upon their child's discharge. It is a benign intervention that does no harm. Future Research There are a number of other future areas for this type of patient-centered research. First, future research on Link is needed to explore its potential effectiveness in improving child behaviors, given our modest findings. Second, similar intervention research could be conducted in other hospital settings such as the ED, outpatient clinics, and community settings such as schools, where many more injured children are seen. Some of these injuries are less severe, however, which would then yield potentially differential effects. Third, for studies that involve testing of Link, careful consideration is needed to identify Link interventionists other than parents to provide direct intervention to injured children, because parents may be challenged in receiving trainings/intervention materials in a high-stress environment. Possible staff who could take on this role in the hospital setting include child life specialists, social workers, nurse's aides, and nurses. Finally, there is some potential in translating our materials to minority populations in both urban and rural settings and developing Link and trauma education for children who suffer from intentional injuries such as interpersonal violence and self-directed violence. Although our study did not find strong evidence of improved stress outcomes among Link patients compared with trauma education patients, stakeholders perceived Link to be potentially valuable. Thus, other extensions of Link in settings beyond the hospital and with other types of interventionists are currently being explored as encouraged by our stakeholders. Drs. Ramirez, Peek-Asa, and Espe-Pfeifer along with Dr Karen Heimer of University of Iowa's Department of Sociology are conducting a study to evaluate how a school-based version of Link might be effective in reducing problem behaviors associated with school violence. Additionally, Dr Ramirez is working with school nurses to develop a school nurse version of Link. Conclusions We compared 2 different types of parent-based interventions to support children's recovery from unintentional injury: Link for Injured Kids, a program that supports parent-child communication using MI skills and includes screening and referral tools; and trauma education, a booklet with information about injuries and referrals to care. Trauma education, a less intensive intervention, was implemented more fully by parents than Link. In terms of outcomes, there were no differences in the overall effectiveness of Link and trauma education on the primary study outcomes (symptoms of PTSD and depression) or secondary study outcomes (nonspecific stress and improved pediatric QOL). However, Link was mildly associated with a secondary outcome: fewer peer and conduct problem behaviors among traumatically injured children. Link was also slightly more effective in improving peer relationship behavior among rural children, but not among urban children. A number of limitations are noteworthy. First, our study did not have a control group to assess whether both interventions were better in improving outcomes than none. Second, our measures are based on self-reports, including measures of fidelity. Third, results are limited to 4 children's hospitals in the Midwest and to all injured children who were admitted (regardless of length of stay) to these hospitals. Finally, study outcomes had a lot of noise or variability, thus making it difficult to detect effects. We recommend that future studies be conducted to further explore how Link might be effective specifically for injured children with behavioral problems. Because Link is a time- and resource-intensive intervention, we are not certain that it should be implemented with overtaxed parents of injured children in hospital settings, based on the findings from this study. Testing of Link in other settings, such as outpatient clinics or schools, involving other types of interventionists are worth further exploration. Trauma education, which was found to be superior to no intervention in previous research, is low cost and benign, and should be offered universally. References 1. Borse NN, Gilchrist J, Dellinger AM, Rudd RA, Ballesteros MF, Sleet DA. CDC Childhood Injury Report: Patterns of Unintentional Injuries among 0-19 Year Olds in the United States, 2000-2006. Centers for Disease Control and Prevention; December 2008. Accessed April 16, 2020. https://www.cdc.gov/safechild/pdf/CDC-ChildhoodInjury.pdf 2. Centers for Disease Control and Prevention, National Center for Injury Prevention and Control. National Action Plan for Child Injury Prevention: An Agenda to Prevent Injuries and Promote the Safety of Children and Adolescents in the United States. 2012. Accessed April 16, 2020. https://www.cdc.gov/safechild/pdf/National_Action_Plan_for_Child_Injury_Prevention.pdf 3. Bisson JI. Post-traumatic stress disorder. BMJ. 2007;334(7597):789-793. [ PMC free article : PMC1852007 ] [ PubMed : 17431265 ] 4. Bisson J, Andrew M. Psychological treatment of post-traumatic stress disorder (PTSD). Cochrane Database Syst Rev. 2007;(3):CD003388. https://doi ​.org/10.1002/14651858 ​.CD003388.pub3 [ PubMed : 17636720 ] 5. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 5th ed. American Psychiatric Publishing; 2013. 6. January AM, Zebracki K, Chlan KM, Vogel LC. Symptoms of depression over time in adults with pediatric-onset spinal cord injury. Arch Phys Med Rehabil. 2014;95(3):447-454. [ PubMed : 24316327 ] 7. Jaques ML, Weaver TL, Weaver NL, Willoughby L. The association between pediatric injury risks and parenting behaviours. Child Care Health Dev. 2018;44(2):297-303. [ PubMed : 28983939 ] 8. Goslin MC, Stover CS, Berkowitz S, Marans S. Identifying youth at risk for difficulties following a traumatic event: pre-event factors are associated with acute symptomatology. J Trauma Stress. 2013;26(4):475-482. [ PubMed : 23861167 ] 9. Kassam-Adams N, Bakker A, Marsac ML, Fein JA, Winston FK. Traumatic stress, depression, and recovery: child and parent responses after emergency medical care for unintentional injury. Pediatr Emerg Care. 2015;31(11):737-742. [ PubMed : 26535495 ] 10. Olofsson E, Bunketorp O, Andersson AL. Children and adolescents injured in traffic—associated psychological consequences: a literature review. Acta Paediatr. 2009;98(1):17-22. [ PubMed : 18795910 ] 11. Stallard P, Velleman R, Baldwin S. Prospective study of post-traumatic stress disorder in children involved in road traffic accidents. BMJ. 1998;317(7173):1619-1623. [ PMC free article : PMC28739 ] [ PubMed : 9848900 ] 12. Aaron J, Zaglul H, Emery RE. Posttraumatic stress in children following acute physical injury. J Pediatr Psychol. 1999;24(4):335-343. [ PubMed : 10431499 ] 13. Borse NN, Gilchrist J, Dellinger AM, Rudd RA, Ballesteros MF, Sleet DA. CDC Childhood Injury Report: Patterns of Unintentional Injuries among 0-19 Year Olds in the United States, 2000-2006. Centers for Disease Control and Prevention; December 2008. Accessed July 23, 2012. https://www.cdc.gov/safechild/pdf/CDC-ChildhoodInjury.pdf 14. Brocque RML, Hendrikz JK, Kenardy JA. The course of posttraumatic stress in children: examination of recovery trajectories following traumatic injury. J Pediatr Psychol. 2010;35(6):637–645. [ PubMed : 19541598 ] 15. Schreier H, Ladakakos C, Morabito D, Chapman L, Knudson MM. Posttraumatic stress symptoms in children after mild to moderate pediatric trauma: a longitudinal examination of symptom prevalence, correlates, and parent-child symptom reporting. J Trauma. 2005;58(2):353-363. [ PubMed : 15706200 ] 16. Schneeberg A, Ishikawa T, Kruse S, et al. A longitudinal study on quality of life after injury in children. Health Qual Life Outcomes. 2016;14(1):120. doi: 10.1186/s12955-016-0523-6 [ PMC free article : PMC5000468 ] [ PubMed : 27561258 ] [ CrossRef ] 17. Michaels AJ, Michaels CE, Moon CH, et al. Posttraumatic stress disorder after injury: impact on general health outcome and early risk assessment. J Trauma. 1999;47(3):460-466; discussion 466-467. [ PubMed : 10498298 ] 18. Holbrook TL, Hoyt DB, Coimbra R, Potenza B, Sise M, Anderson JP. Long-term posttraumatic stress disorder persists after major trauma in adolescents: new data on risk factors and functional outcome. J Trauma. 2005;58(4):764-769; discussion 769-771. [ PubMed : 15824653 ] 19. Broberg AG, Dyregrov A, Lilled L. The Göteborg discotheque fire: posttraumatic stress, and school adjustment as reported by the primary victims 18 months later. J Child Psychol Psychiatry. 2005;46(12):1279-1286. [ PubMed : 16313428 ] 20. Lipschitz DS, Rasmusson AM, Anyan W, Cromwell P, Southwick SM. Clinical and functional correlates of posttraumatic stress disorder in urban adolescent girls at a primary care clinic. J Am Acad Child Adolesc Psychiatry. 2000;39(9):1104-1111. [ PubMed : 10986806 ] 21. Staley M, Anderson L, Greenhalgh D, Warden G. Return to school as an outcome measure after a burn injury. J Burn Care Rehabil. 1999;20(1 Pt 1):91-94; discussion 90. [ PubMed : 9934643 ] 22. Barzilay R, Calkins ME, Moore TM, et al. Association between traumatic stress load, psychopathology, and cognition in the Philadelphia Neurodevelopmental Cohort. Psychol Med. 2018:1-10. doi:10.1017/S0033291718000880 [ PubMed : 29655375 ] [ CrossRef ] 23. Mikkonen J, Moustgaard H, Remes H, Martikainen P. The population impact of childhood health conditions on dropout from upper-secondary education. J Pediatr. 2018;196:283-290.e4. doi:10.1016/j.jpeds.2018.01.034 [ PubMed : 29551321 ] [ CrossRef ] 24. Domino ME, Lin CC, Morrissey JP, et al. Training psychologists for rural practice: exploring opportunities and constraints. J Rural Health. 2019;35(1):35-41. [ PubMed : 29664202 ] 25. Talbot JA, Ziller EC, Szlosek DA. Mental health first aid in rural communities: appropriateness and outcomes. J Rural Health. 2017;33(1):82-91. [ PubMed : 26817852 ] 26. Henning-Smith C, Evenson A, Corbett A, Kozhimannil K, Moscovice I. Rural Transportation: Challenges and Opportunities. University of Minnesota Rural Health Research Center; November 2017. Accessed April 16, 2020. https: ​//3pea7g1qp8f3t9ooe3z3npx1-wpengine ​.netdna-ssl.com/wp-content ​/uploads/2019 ​/01/1518734252UMRHRCTransportationChallenges.pdf 27. Probst JC, Barker JC, Enders A, Gardiner P. Current state of child health in rural America: how context shapes children's health. J Rural Health. 2018;34(suppl 1):S3-S12. [ PMC free article : PMC5373918 ] [ PubMed : 27677973 ] 28. Salam RA, Arshad A, Das JK, et al. Interventions to prevent unintentional injuries among adolescents: a systematic review and meta-analysis. J Adolesc Health. 2016;59(4S):S76-S87. [ PMC free article : PMC5026686 ] [ PubMed : 27664598 ] 29. Clark JJ, Sprang G, Freer B, Whitt-Woosley A. ‘Better than nothing' is not good enough: challenges to introducing evidence-based approaches for traumatized populations. J Eval Clin Pract. 2012;18(2):352-359. [ PubMed : 21143348 ] 30. Pignatiello A, Teshima J, Boydell K, Minden D, Volpe T, Braunberger PG. Child and youth telepsychiatry in rural and remote primary care. Child Adolesc Psychiatr Clin N Am. 2011;20(1):13-28. [ PubMed : 21092909 ] 31. Steele M, Shapiro J, Davidson B, et al. Survey comparing criteria used by rural and urban primary care physicians for referrals to child and adolescent psychiatrists and children's mental health agencies in Ontario. J Can Acad Child Adolesc Psychiatry. 2010;19(4):284-289. [ PMC free article : PMC2962541 ] [ PubMed : 21037919 ] 32. Meit M, Knudson A, Gilbert T, et al. The 2014 Update of the Rural-Urban Chartbook. Rural Health Reform Policy Research Center; 2014. Accessed April 16,2020. https://ruralhealth ​.und ​.edu/projects/health-reform-policy-research-center ​/pdf/2014-rural-urban-chartbook-update.pdf 33. Zatzick DF, Rowhani-Rahbar A, Wang J, et al. The cumulative burden of mental, substance use, and general medical disorders and rehospitalization and mortality after an injury. Psychiatr Serv. 2017;68(6):596-602. [ PMC free article : PMC5550030 ] [ PubMed : 28142384 ] 34. Schroeder SM, Peterson ML. Identifying variability in patient characteristics and prevalence of emergency department utilization for mental health diagnoses in rural and urban communities. J Rural Health. 2018;34(4):369-376. [ PubMed : 29058344 ] 35. Pfefferbaum B, Varma V, Nitiéma P, Newman E. Universal preventive interventions for children in the context of disasters and terrorism. Child Adolesc Psychiatr Clin N Am. 2014;23(2):363-382, ix-x. [ PMC free article : PMC3964365 ] [ PubMed : 24656585 ] 36. Cobham VE, March S, De Young A, et al. Involving parents in indicated early intervention for childhood PTSD following accidental injury. Clin Child Fam Psychol Rev. 2012;15(4):345-363. [ PubMed : 22983482 ] 37. Berkowitz SJ, Watson PJ, Brymer MJ. Early preventive interventions for adolescents exposed to a potentially traumatic event. Minerva Pediatr. 2011;63(3):201-215. [ PubMed : 21654600 ] 38. Stallard P, Velleman R, Salter E, Howse I, Yule W, Taylor G. A randomised controlled trial to determine the effectiveness of an early psychological intervention with children involved in road traffic accidents. J Child Psychol Psychiatry. 2006;47(2):127-134. [ PubMed : 16423143 ] 39. Zatzick D, Russo J, Lord SP, et al. Collaborative care intervention targeting violence risk behaviors, substance use, and posttraumatic stress and depressive symptoms in injured adolescents: a randomized clinical trial. JAMA Pediatr. 2014;168(6):532-539. [ PubMed : 24733515 ] 40. Kassam-Adams N, Garcia-España JF, Marsac ML, et al. A pilot randomized controlled trial assessing secondary prevention of traumatic stress integrated into pediatric trauma care. J Trauma Stress. 2011;24(3):252-259. [ PMC free article : PMC8935905 ] [ PubMed : 21594900 ] 41. de Arellano MA, Lyman DR, Jobe-Shields L, et al. Trauma-focused cognitive-behavioral therapy for children and adolescents: assessing the evidence. Psychiatr Serv. 2014;65(5):591-602. [ PMC free article : PMC4396183 ] [ PubMed : 24638076 ] 42. Cohen JA, Mannarino AP. Trauma-focused cognitive behavior therapy for traumatized children and families. Child Adolesc Psychiatr Clin N Am. 2015;24(3):557-570. [ PMC free article : PMC4476061 ] [ PubMed : 26092739 ] 43. Kenardy J, Thompson K, Le Brocque R, Olsson K. Information-provision intervention for children and their parents following pediatric accidental injury. Eur Child Adolesc Psychiatry. 2008;17(5):316-325. [ PubMed : 18350366 ] 44. Zehnder D, Meuli M, Landolt MA. Effectiveness of a single-session early psychological intervention for children after road traffic accidents: a randomised controlled trial. Child Adolesc Psychiatry Ment Health. 2010;4:7. doi: 10.1186/1753-2000-4-7 [ PMC free article : PMC2841102 ] [ PubMed : 20181120 ] [ CrossRef ] 45. Cox CM, Kenardy JA, Hendrikz JK. A randomized controlled trial of a web-based early intervention for children and their parents following unintentional injury. J Pediatr Psychol. 2010;35(6):581-592. [ PubMed : 19906829 ] 46. Marsac ML, Kassam-Adams N, Hildenbrand AK, Kohser KL, Winston FK. After the injury: initial evaluation of a web-based intervention for parents of injured children. Health Educ Res. 2011;26(1):1-12. [ PubMed : 20858769 ] 47. Berkowitz SJ, Watson PJ, Brymer MJ. Early preventive interventions for adolescents exposed to a potentially traumatic event. Minerva Pediatr. 2011;63(3):201-215. [ PubMed : 21654600 ] 48. Alisic E, Gunaratnam S, Barrett A, et al. Injury talk: spontaneous parent-child conversations in the aftermath of a potentially traumatic event. Evid Based Ment Health. 2017;20(4):e19-e20. doi: 10.1136/eb-2017-102736 [ PMC free article : PMC5750414 ] [ PubMed : 29030502 ] [ CrossRef ] 49. O'Neal EE, Plumert JM, Peterson C. Parent-child injury prevention conversations following a trip to the emergency department. J Pediatr Psychol. 2016;41(2):256-264. [ PMC free article : PMC4884903 ] [ PubMed : 26275976 ] 50. Ramirez M, Toussaint M, Woods-Jaeger B, et al. Link for injured kids: a patient-centered program of psychological first aid after trauma. Pediatr Emerg Care. 2017;33(8):532-537. [ PMC free article : PMC5548509 ] [ PubMed : 26428077 ] 51. Borrelli B, Tooley EM, Scott-Sheldon LA. Motivational interviewing for parent-child health interventions: a systematic review and meta-analysis. Pediatr Dent. 2015;37(3):254-265. [ PubMed : 26063554 ] 52. Erickson SJ, Gerstle M, Feldstein SW. Brief interventions and motivational interviewing with children, adolescents, and their parents in pediatric health care settings: a review. Arch Pediatr Adolesc Med. 2005;159(12):1173-1180. [ PubMed : 16330743 ] 53. Kataoka S, Langley AK, Wong M, Baweja S, Stein BD. Responding to students with posttraumatic stress disorder in schools. Child Adolesc Psychiatr Clin N Am. 2012;21(1):119-133, x. [ PMC free article : PMC3287974 ] [ PubMed : 22137816 ] 54. Wong M. Interventions to reduce psychological harm from traumatic events among children and adolescents: a commentary on the application of findings to the real world of schools. Am J Prev Med. 2008;35(4):398-400. [ PubMed : 18779033 ] 55. Ramirez M, Harland K, Frederick M, Shepherd R, Wong M, Cavanaugh JE. Listen protect connect for traumatized schoolchildren: a pilot study of psychological first aid. BMC Psychol. 2013;1(1):26. doi: 10.1186/2050-7283-1-26 [ PMC free article : PMC4269971 ] [ PubMed : 25566374 ] [ CrossRef ] 56. McMillan SS, King M, Tully MP. How to use the nominal group and Delphi techniques. Int J Clin Pharm. 2016;38(3):655-662. [ PMC free article : PMC4909789 ] [ PubMed : 26846316 ] 57. Kessler RC, Andrews G, Colpe LJ, et al. Short screening scales to monitor population prevalences and trends in non-specific psychological distress. Psychol Med. 2002;32(6):959-976. [ PubMed : 12214795 ] 58. Foa EB, Johnson KM, Feeny NC, Treadwell KR. The Child PTSD Symptom Scale: a preliminary examination of its psychometric properties. J Clin Child Psychol. 2001;30(3):376-384. [ PubMed : 11501254 ] 59. Foa EB, Cashman L, Jaycox L, Perry K. The validation of a self-report measure of posttraumatic stress disorder: the Posttraumatic Diagnostic Scale. Psychol Assess. 1997;9(4):445-451. 60. Poulin C, Hand D, Boudreau B. Validity of a 12-item version of the CES-D used in the National Longitudinal Study of Children and Youth. Chronic Dis Can. 2005;26(2-3):65-72. [ PubMed : 16251012 ] 61. Varni JW, Seid M, Kurtin PS. PedsQL 4.0: reliability and validity of the Pediatric Quality of Life Inventory version 4.0 generic core scales in healthy and patient populations. Med Care. 2001;39(8):800-812. [ PubMed : 11468499 ] 62. Olson DH, Gorall DM, Tiesel JW. Faces IV package: Administration. Life Innovations, Inc; 2004. Accessed April 16, 2020. https://www ​.nctsn.org ​/measures/family-adaptability-and-cohesion-scale 63. Goodman R. Psychometric properties of the strengths and difficulties questionnaire. J Am Acad Child Adolesc Psychiatry. 2001;40(11):1337-1345. [ PubMed : 11699809 ] 64. Bruwer B, Emsley R, Kidd M, Lochner C, Seedat S. Psychometric properties of the Multidimensional Scale of Perceived Social Support in youth. Compr Psychiatry. 2008;49(2):195-201. [ PubMed : 18243894 ] 65. Ayers TS, Sandler IN. Manual for the Children's Coping Strategies Checklist and the How I Coped Under Pressure Scale. Program for Prevention Research, Arizona State University; 1999. Accessed July 20, 2020. https://www ​.yumpu.com ​/en/document/read/12240213 ​/manual-for-the-childrens-coping-strategies-checklist 66. Henson RK. Understanding internal consistency reliability estimates: a conceptual primer on coefficient alpha. Meas Eval Couns Dev. 2001;34(3):177-189. 67. Cromartie J, Bucholtz S. Defining the” rural” in rural America. Amber Waves. 2008;6(3):28. 68. Glasgow RE, Dickinson P, Fisher L, et al. Use of RE-AIM to develop a multi-media facilitation tool for the patient-centered medical home. Implement Sci. 2011;6:118. doi: 10.1186/1748-5908-6-118 [ PMC free article : PMC3229439 ] [ PubMed : 22017791 ] [ CrossRef ] 69. Glasgow RE, Vogt TM, Boles SM. Evaluating the public health impact of health promotion interventions: the RE-AIM framework. Am J Public Health. 1999;89(9):1322-1327. [ PMC free article : PMC1508772 ] [ PubMed : 10474547 ] 70. Lane C, Huws-Thomas M, Hood K, Rollnick S, Edwards K, Robling M. Measuring adaptations of motivational interviewing: the development and validation of the Behavior Change Counseling Index (BECCI). Patient Educ Couns. 2005;56(2):166-173. [ PubMed : 15653245 ] 71. McGraw KO, Wong SP. Forming inferences about some intraclass correlation coefficients. Psychol Methods. 1996;1(1):30-46. 72. Goodman R, Scott S. Comparing the Strengths and Difficulties Questionnaire and the Child Behavior Checklist: is small beautiful? J Abnorm Child Psychol. 1999;27(1):17-24. [ PubMed : 10197403 ] 73. Mathai J, Anderson P, Bourne A. Use of the Strengths and Difficulties Questionnaire as an outcome measure in a child and adolescent mental health service. Australas Psychiatry. 2003;11(3):334-337. 74. Szymanski K, Sapanski L, Conway F. Trauma and ADHD—association or diagnostic confusion? A clinical perspective. J Infant Child Adolesc Psychother. 2011;10(1):51-59. 75. Schwarz ED, Perry BD. The post-traumatic response in children and adolescents. Psychiatr Clin North Am. 1994;17(2):311-326. [ PubMed : 7937362 ] 76. Oral R, Ramirez M, Coohey C, et al. Adverse childhood experiences and trauma informed care: the future of health care. Pediatr Res. 2016;79(1-2):227-233. [ PubMed : 26460523 ] 77. Berkowitz SJ, Stover CS, Marans SR. The child and family traumatic stress intervention: secondary prevention for youth at risk of developing PTSD. J Child Psychol Psychiatry. 2011;52(6):676-685. [ PMC free article : PMC3096712 ] [ PubMed : 20868370 ] 78. Kahana SY, Feeny NC, Youngstrom EA, Drotar D. Posttraumatic stress in youth experiencing illnesses and injuries: an exploratory meta-analysis. Traumatol. 2006;12(2):148-161. Acknowledgments Research reported in this publication was funded through a PCORI Award (CER-1306-02918). The opinions in this publication are solely the responsibility of the authors and do not necessarily represent the views of PCORI, its Board of Governors, or Methodology Committee. Research reported in this publication was supported by the National Center for Advancing Translational Sciences of the National Institutes of Health under award number U54TR001356. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Research reported in this report was funded through a Patient-Centered Outcomes Research Institute® (PCORI®) Award (#CER-1306-02918). Further information available at: https://www.pcori.org/research-results/2013/comparing-two-approaches-help-parents-support-their-childrens-social-and Institution Receiving Award: University of Iowa Original Project Title: Evaluation of Parent-Based Interventions to Support Children after Traumatic Injury PCORI ID: CER-1306-02918 ClinicalTrials.gov ID: NCT02323204 Suggested citation: Ramirez M, Woods-Jaeger B, Peek-Asa C, et al. (2020). Comparing Two Approaches to Help Parents Support Their Children's Social and Emotional Recovery after a Serious Accidental Injury . Patient-Centered Outcomes Research Institute (PCORI). https://doi.org/10.25302/07.2020.CER.130602918 Disclaimer The [views, statements, opinions] presented in this report are solely the responsibility of the author(s) and do not necessarily represent the views of the Patient-Centered Outcomes Research Institute® (PCORI®), its Board of Governors or Methodology Committee. Copyright © 2020. University of Iowa. All Rights Reserved. This book is distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License which permits noncommercial use and distribution provided the original author(s) and source are credited. (See https://creativecommons.org/licenses/by-nc-nd/4.0/ Bookshelf ID: NBK619785 PMID: 41411459 DOI: 10.25302/07.2020.CER.130602918 Share Views PubReader Print View Cite this Page Ramirez M, Woods-Jaeger B, Peek-Asa C, et al. Comparing Two Approaches to Help Parents Support Their Children's Social and Emotional Recovery after a Serious Accidental Injury [Internet]. Washington (DC): Patient-Centered Outcomes Research Institute (PCORI); 2020 Jul. doi: 10.25302/07.2020.CER.130602918 PDF version of this title (1.2M) In this Page Background Participation of Patients and Other Stakeholders Methods Results Discussion Conclusions References Acknowledgments Other titles in this collection PCORI Final Research Reports Related information NLM Catalog Related NLM Catalog Entries PMC PubMed Central citations PubMed Links to PubMed Recent Activity Clear Turn Off Turn On Comparing Two Approaches to Help Parents Support Their Children's Social and Emo... Comparing Two Approaches to Help Parents Support Their Children's Social and Emotional Recovery after a Serious Accidental Injury Your browsing activity is empty. Activity recording is turned off. Turn recording back on See more... Follow NCBI Twitter Facebook LinkedIn GitHub NCBI Insights Blog Connect with NLM Twitter Facebook Youtube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov