Collaborative Care for Adolescents With Persistent Postconcussive Symptoms: A Randomized Trial
BACKGROUND AND OBJECTIVES: Postconcussive and co-occurring psychological symptoms are not uncommon after sports-related concussion and are associated with functional impairment and societal costs. There is no evidence-based treatment targeting postconcussive symptoms in children and adolescents. The goal of this study was to test a collaborative care intervention model with embedded cognitive–behavioral therapy, care management, and psychopharmacological consultation. We hypothesized that patients in collaborative care would demonstrate greater reductions in postconcussive, depressive, and anxiety symptoms and improvement in functioning over the course of 6 months, compared with usual care control.
METHODS: Patients aged 11 to 17 years with persistent symptoms ≥1 month after sports-related concussion were randomly assigned to receive collaborative care (n = 25) or care as usual (n = 24). Patients were assessed before randomization and after 1, 3, and 6 months. Groups were compared over time via linear mixed effects regression models.
RESULTS: Adolescents assigned to collaborative care experienced clinically and statistically significant improvements in postconcussive symptoms in addition to functional gains at 6 months compared with controls. Six months after the baseline assessment, 13.0% of intervention patients and 41.7% of control patients reported high levels of postconcussive symptoms (P = .03), and 78% of intervention patients and 45.8% of control patients reported ≥50% reduction in depression symptoms (P = .02). No changes between groups were demonstrated in anxiety symptoms.
CONCLUSIONS: Orchestrated efforts to systematically implement collaborative care treatment approaches for slow-to-recover adolescents may be useful given the reductions in postconcussive and co-occurring psychological symptoms in addition to improved quality of life.
- CBT —
- cognitive–behavioral therapy
- CI —
- confidence interval
- HBI —
- Health and Behavior Inventory
- PedsQL —
- Pediatric Quality of Life Inventory
- PHQ-9 —
- Patient Health Questionnaire
- PROMIS-PA8 —
- Patient-Reported Outcomes Measurement Information System
What’s Known on This Subject:
Persistent postconcussive symptoms confer significant impairment in a subset of youth after concussion. Co-occurring psychological symptoms are common and may prolong recovery. Lack of empirical data in support of treatment of postconcussive symptoms places the clinician in a challenging position.
What This Study Adds:
Adolescents with persistent postconcussive symptoms who received collaborative care, including integrated care management and cognitive–behavioral therapy, displayed a significant reduction in postconcussive symptoms and functional improvement after 6 months as compared with adolescents in usual care.
Sports-related concussions have galvanized the American public and represent an important public health problem, especially for children and adolescents. Although symptomatic recovery is typically reached within 2 weeks of concussive injury, an estimated 13.7% of school-aged youth remain symptomatic at 3 months.1 Over time, refractory headache, fatigue, dizziness, and inattention can confer significant functional impairment including executive dysfunction and poor academic performance.2–4 Affective symptoms, including depression and anxiety, commonly co-occur with cognitive and somatic symptoms and may prolong recovery from postconcussive symptoms, as has been demonstrated in patients with traumatic brain injury and depressive symptoms.5–9 Although it is widely accepted that a variety of biological, psychological, and social factors together contribute to increased risk of protracted symptom duration, the etiology and trajectory of persistent postconcussive symptoms remain poorly understood.10 Taken together, the complexities of managing persistent postconcussive symptoms in conjunction with comorbid psychological symptoms create a significant burden for injured children and adolescents, their families, and schools.2,3,11
Current clinical paradigms emphasize education about the symptoms of concussion, anticipatory guidance around a gradual return to cognitive and physical activity, psychoeducation, and reassurance of a projected complete recovery.12–15 These clinical strategies may be adequate for children at the time of acute symptom presentations but remain unsubstantiated by rigorous clinical trials.12,16 No evidence-based guidelines exist regarding care of youth with persistent symptoms (≥1 month) from concussion.17 This lack of empirical data places the treating clinician in a challenging position where watchful waiting remains the key component of standard care, leaving a gap in active intervention options.18
Cognitive–behavioral therapy (CBT) has been recommended as a first-line treatment of persistent postconcussive symptoms based on studies with adults after concussion.13,19 Robust evidence has documented that CBT is effective in the treatment of adolescent depression, anxiety, and chronic pain.20,21 Psychopharmacological management of co-occurring mood disorders requires close coordination and collaboration between physicians with specialized clinical training and experience in traumatic brain injury and child psychiatry.22 Zatzick et al23 previously established safety of psychotropic medication provision with proper supervision in adolescents with traumatic brain injury. However, CBT, care coordination, and psychopharmacological treatment have not yet been harnessed within a clinical intervention to target postconcussive symptoms and co-occurring depressive and anxiety symptoms.24,25
Collaborative care is a health care delivery model that integrates medical and behavioral health care for patients with chronic disorders in which both physical and psychological symptoms are present, such as sequelae related to traumatic brain injury.26,27 Large-scale randomized clinical trials have established the effectiveness of collaborative care interventions that combine care management, CBT, and evidence-based pharmacotherapy in treating pediatric and adult primary care patients with depressive, anxiety, and posttraumatic stress disorders.27–29 Given the constellation of somatic, cognitive, sleep, and postconcussive symptoms, collaborative care may be ideal because it can facilitate linkages and coordination across pediatric primary care, subspecialist sports medicine and rehabilitation, and school-based service delivery sectors. However, no previous study has examined the efficacy of collaborative care treatment of postconcussive symptoms in adolescents.
The current investigation was a randomized controlled trial conducted among adolescents who reported persistent postconcussive symptoms after sports-related concussion. The study was designed to test whether receiving a CBT intervention embedded in a collaborative care treatment model would reduce postconcussive, anxiety, and depressive symptoms. We also hypothesized that intervention patients would demonstrate improved health-related quality of life. The intervention was designed as a stepped-care model whereby adolescents were first provided with care management and CBT; patients’ symptoms were monitored, and for those who remained symptomatic, psychopharmacologic consultation was made available.
Design, Setting, and Participants
Participants were adolescents ages 11 to 17 years with sports- or recreation-related concussion. The sample was 65% female, with the following racial distribution: 75.5% white, 14.3% other, 8.2% Asian or Pacific Islander, and 2.0% black. Approximately 8% of the sample identified as Hispanic or Latino. All concussions were diagnosed by sports medicine or rehabilitation medicine specialists trained in concussion diagnosis and management. Informed consent and assent were obtained before data collection, and study procedures were approved by the Seattle Children’s Institutional Review Board. Recruitment occurred between October 2014 and July 2015.
During active recruitment, a research associate reviewed upcoming visits for patients with a sports-related concussion in the Sports Medicine and Rehabilitation Medicine Clinics at Seattle Children’s and the Sports Concussion Program at Harborview Medical Center, clinics that receive referrals from emergency and primary care physicians for management of postconcussive symptoms. If patients were symptomatic ≥1 month after the concussion, a research associate subsequently informed the family of the clinical trial and conducted an eligibility screening by phone. Parents were asked to assess the frequency of postconcussive symptoms their children experienced during the preceding week and the week before the injury by using the Health and Behavior Inventory (HBI).30 Patients were eligible to participate if they experienced an onset or increase in frequency of ≥3 postconcussive symptoms after a concussion within the past year compared with preinjury parent report. Patients and ≥1 parent or guardian were required to be able to read and speak English and to live within commuting distance of the regional children’s hospital where assessments and treatment took place. Patients were excluded if they had sustained more severe brain injuries, as documented by any abnormalities demonstrated on computed tomography or MRI. Patients who were actively suicidal or abusing substances or who reported bipolar or psychotic disorders were also excluded. Patients with other preexisting psychological disorders were considered eligible.
Randomization to collaborative or usual care treatment groups occurred in a 1:1 ratio according to random assignments that were computer generated by a study biostatistician. Patients were stratified by the frequency of postconcussive symptoms (0–21 vs 22–60) reported by their parent within the past week via the 20-item HBI. The cutoff of 21 for stratification purposes was chosen based on a median split of HBI scores from a sample of youth with mild traumatic brain injury who had ≥3 symptoms, from data provided by the primary author of measure (K. Yeates, PhD, personal communication, 2014). Research associates conducting baseline and follow-up assessments were blinded to intervention or control group status.31 The unblinded care manager called families to inform them of their randomization assignment after the baseline assessment.
Collaborative Care Intervention
Patients randomly assigned to the intervention received care management, CBT, and possible psychopharmacological consultation over the course of the 6-month treatment period. Intervention patients were treated for the duration of their symptoms, and treatment was terminated upon symptom resolution or at the end of the 6-month follow-up. Modular CBT targeting postconcussive depressive and anxiety symptoms was delivered by 1 of 5 study therapists (4 PhD-level psychologists and 1 licensed therapist). CBT focused on techniques to encourage increased appropriate activation, including pacing of activities and the use of coping skills, relaxation strategies, sleep hygiene, and positive thinking in the face of symptoms. Health care coordination needs were carried out by the care manager, who communicated with the consulting psychiatrist, primary care physicians, and subspecialists to ensure that individual physical, psychological health, and health care–related concerns were actively addressed. Care management also included advocacy and coordination with the patient’s school regarding accommodations and motivational interviewing to maintain engagement in treatment. A multidisciplinary case conference included a bachelors-level care manager (E.S.), a supervising CBT interventionist (C.A.M.), a pediatrician (F.P.R.), and a supervising MD psychiatrist skilled in the postinjury collaborative care model (D.Z.); this team met weekly to discuss patients randomly assigned to the intervention condition. An expert in pediatric psychopharmacology (R.H.) provided pharmacologic consultation in cases of severe or recalcitrant symptoms that did not respond to psychotherapy when warranted at any time in treatment.
Usual Care Control Condition
Adolescent patients and their parents in the control condition underwent the baseline and follow-up assessments at 1, 3, and 6 months after enrollment. They received usual care organized by the referring clinic concussion specialist. Usual care in our system consisted of a visit initially to a sports medicine trained physician. For youth who were symptomatic beyond 4 to 6 weeks, patients were referred to pediatric rehabilitation medicine. They also usually underwent MRI, with diffuse tensor imaging sequences, and testing by a neuropsychologist. Patients having difficulty with sleep were usually prescribed a hypnotic medication, and patients may have also been provided with a subthreshold exercise program by a physical therapist. Usual care was documented by reviewing the medical chart and parent report of health care utilization outside the hospital over the course of the 6-month study.
The following set of measures captured the degree of adolescent physical and psychological impairment present at baseline and 1, 3, and 6 months, based on both youth and parent report.
Primary Outcome Measures
The HBI was administered to both youth and parents to assess the presence of postconcussive symptoms. The HBI is a 20-item questionnaire that assesses postconcussive symptoms on a 4-point scale, ranging from “never” to “often,” and yields total scores in cognitive and somatic domains, with higher scores indicating more symptomatic patients. The scale includes youth-report and parent-report versions with established reliability and validity in youth with sports injury.30,32
The Patient Health Questionnaire (PHQ-9), which includes 9 questions based on Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition major depression criteria, was administered to assess depressive symptoms.33 It has been found to have high sensitivity (89.5%) and acceptable specificity (73%) for the diagnosis of major depression in adolescent populations.34Anxiety Symptoms
The PROMIS-PA8 (version a) asks about the frequency of 8 anxiety symptoms over the preceding 7 days on a 5-point scale, with higher scores indicating more symptoms. The scale includes youth-reported and parent-reported versions with established validity and reliability among adolescents.35–37Health-Related Quality of Life
The Pediatric Quality of Life Inventory (PedsQL) is a 23-item questionnaire that assesses physical, emotional, social, and school functioning, with higher scores indicating greater quality of life.38 The scale includes youth-reported and parent-reported versions with demonstrated validity and reliability, including good to excellent internal consistency (Cronbach’s α = 0.88 for child and 0.90 for parent report).38–41 It has successfully been used in youth traumatic brain injury research previously.42 A 4- or 5-point difference is clinically meaningful.39
Secondary Outcome Measure
Satisfaction With Care
An item from the Client Satisfaction Questionnaire assessed satisfaction with care: “In an overall, general sense, how satisfied are you with the care you/your child have received?”43Other Measures
A medical chart review at the end of the study follow-up period provided demographic and health care utilization information. Additional mental health care utilization throughout the course of enrollment was also assessed in a questionnaire administered to both treatment groups.
Moderate to large effect sizes have been observed in a general sample of youth with mild to severe traumatic brain injury.44 We calculated that a sample size of 40 would provide adequate power to observe a large treatment effect (partial η2 = 0.14). In the current study, we recruited 49 adolescents to compensate for potential dropouts.
We used t tests and χ2 analyses to compare groups at baseline, using P < .05 as our threshold for statistical significance. To determine whether patients in the 2 groups manifested different patterns of change in symptoms and functioning over the 6 months after enrollment, we used linear mixed effects regression models.45,46 For all dependent variables, we fit models containing time categories (baseline, 1 month, 3 months, and 6 months), treatment group, and treatment group by time interactions. The final models presented include no covariates given the sample size. All outcome analyses were intent-to-treat and used continuous variables. Effect sizes were calculated with Cohen’s d.47 Relative risk ratios were calculated to compare collaborative care and usual care groups with respect to high levels of postconcussive symptoms and depression symptom response. We conducted χ2 analyses to examine differences between groups in treatment satisfaction.
A total of 49 adolescents (mean age = 15.0 years) enrolled in the study. Adolescents randomly assigned to the intervention and control conditions were not statistically different with regard to demographic variables, preinjury health concerns, or clinical profiles at baseline, as indicated with dichotomized variables (Table 1). A total of 57.5% of the adolescents incurred sports-related concussion, and 42.5% sustained recreation-related concussion (eg, being hit by a ball on the playground, trampoline injury), with a higher proportion of youth assigned to collaborative care sustaining sports concussions (P = .03). The investigation attained >98% follow-up of the participants at 1, 3, and 6 months (Fig 1).
Of the 25 patients randomly assigned to the intervention, 23 completed the full course of collaborative care treatment over 6 months (1 withdrew from treatment before initiating treatment and remained in the study, 1 withdrew from the study). The median number of CBT sessions received was 8 (range 0–12); 88% of patients received ≥4 sessions. One-third of patients received stepped-up care in the form of psychopharmacologic consultation and subsequent medication. Patients receiving collaborative care on average had 6.2 (SD = 4.6) outpatient concussion-related visits (ie, sports medicine, rehabilitation medicine, neurology).
Among 24 patients assigned to the control condition, 8% saw a psychiatrist, 33% reported visits with a mental health professional, 58% sought treatment from their primary care physician, and 21% sought treatment from another physician. Patients receiving usual care on average had 5.8 (SD = 2.3) outpatient concussion-related visits, including sports medicine, rehabilitation medicine, neurology, and pain medicine visits (range 2–23 visits).
Intervention patients had clinically and statistically significant improvements in postconcussive symptoms and health-related quality of life outcomes compared with controls (Fig 2 and Table 2). The overall group × time interaction term indicated more improvement in youth-reported postconcussive symptoms over time among intervention than control adolescents (d = 0.70). At 6 months after the baseline assessment 13% of intervention patients and 41.7% of control patients reported high levels (HBI ≥ 22) of postconcussive symptoms (relative risk = 0.03). Statistically significant improvements in health-related quality of life were demonstrated by both child and parent report. For functional status at 6 months, PedsQL child report improved a mean of 27.7 in the intervention group and 17.8 for controls (mean difference, 9.8; 95% confidence interval [CI], 0.9 to 18.8), and by parent report PedsQL improved by 33.3 points in the intervention group and 19.3 in the control group (mean difference, 14.0; 95% CI, 4.6 to 23.1).
A majority of intervention patients (78.3%) demonstrated >50% reduction in depressive symptoms, a metric used in previous collaborative care trials of treatment of depression in adolescents,29 compared with 45.8% of patients in the usual care group (relative risk 1.71 [1.05–2.79]). Depressive symptom outcomes demonstrated a pattern of symptomatic improvement for adolescents who received the collaborative care intervention compared with controls over time, although differences in frequency of symptoms between the 2 groups at 6 months did not achieve statistical significance.
Adolescents who received collaborative care treatment and their parents reported high levels of satisfaction with care (Table 3). A total of 82.6% of intervention patients were “very satisfied” with their care, as compared with 45.8% of patients in the control group (χ2 = 6.9, P = .001). A total of 87% of parents of intervention patients and 29.2% of parents of patients in the control group were “very satisfied” with care (χ2 = 16.0, P < .0001). There were no reported adverse events.
This study provides evidence of the effectiveness of a collaborative care intervention in reducing postconcussive symptoms in a sample of adolescents with prolonged recovery from sports-related concussion. Although patients in both groups showed symptom reduction in the first 3 months, only those who received collaborative care demonstrated sustained improvements through 6 months of follow-up. The large effect size for postconcussive symptom reduction and quality-of-life improvement reported in the collaborative care group compared with the usual care group was greater than the mean effect size for integrated pediatric medical–behavioral health care interventions according to a recent meta-analysis.48 Parents of adolescents randomly assigned to collaborative care reported improvement in their children’s health-related quality of life that was concordant with that reported by adolescents. Adolescents who received the collaborative intervention were more satisfied with their treatment than those who received usual care, suggesting that this patient population is amenable to behavioral treatment and coordination services offered through care management.
A clinical profile of adolescents with a prolonged postconcussive recovery course emerged, indicating that on average patients were moderately depressed at baseline while displaying anxiety levels on par with mean population levels. At baseline, 84% of enrolled patients reported high postconcussive symptoms, 41% reported high depressive symptoms, and 8% reported high anxiety levels. Low levels of anxiety symptoms in the adolescent patient population consistent with the general population may have limited the power to detect significant treatment effects on anxiety outcomes. Clinically significant reduction in depressive symptoms was established at 6 months in the collaborative care intervention group as compared with patients receiving care as usual. No significant group differences were observed over time for anxiety or parental mental health.
This is the first randomized controlled treatment trial of a collaborative care intervention for persistent postconcussive symptoms in the pediatric population. An integrated, personalized treatment approach using a collaborative care model of health care delivery targeted the full range of postinjury symptoms and comorbidities through CBT and continuous case management services that included advocacy in the school setting as well as stepped-up care coordination and medication consultation to achieve symptom remission. Additional strengths of the study include use of a robust battery of measures that longitudinally tracked symptoms and impairment throughout a 6-month time period, and the high retention of participants.
There were also limitations to the study. The sample was small and drawn from a regional population, including a large proportion of Caucasian girls, which may limit generalizability. Adolescents were enrolled in the study at variable times since injury, introducing heterogeneity of symptom duration. For this trial, we allowed patients to enroll at any point in their symptom trajectory (≥1 month) to reflect a real-world clinical population. Moreover, we used the PHQ-9 to evaluate depressive symptoms, which have some overlap with postconcussive symptoms. We cannot determine which particular components of our intervention were responsible for symptom reduction; nevertheless, a health services approach to providing comprehensive services was intentional and aligned with real-world clinic conditions. Usual care at the regional children’s hospital where the study was conducted included multidisciplinary rehabilitative services at pediatric outpatient clinics. It is worth noting that adolescents randomly assigned to the usual care control group displayed symptom reduction over the course of 6 months. However, adolescents who received collaborative care continued to make symptom reduction gains between 3 and 6 months, whereas the symptoms of adolescents in the control group plateaued after 3 months.
Previous studies have established the role of depression in mediating poor outcomes of children with traumatic brain injury.10,49 This investigation contributes to an expanding literature on combined medical–behavioral health care in the pediatric setting aimed at improving physical and behavioral health outcomes measured by both symptoms and associated health-related quality of life. Our findings indicate that co-occurrence of depressive symptoms in slow-to-recover adolescents is common and suggests that a continuum of care management, CBT, and pharmacotherapy when necessary is a promising approach for reducing the burden of postconcussive symptoms and associated functional impairment. Additional cost analysis is warranted to assess cost savings and health care utilization associated with a collaborative care treatment program for slow-to-recover patients compared with traditional outpatient care. In the absence of empirically based treatment options for this patient population, this study provides evidence for the application of the collaborative care model of integrated care for postconcussive symptoms. In settings where adaptation of the collaborative care model may not be feasible, components of the collaborative care intervention including skill training in behavioral activation, coping, relaxation, or sleep hygiene techniques in addition to closer coordination with schools might be incorporated into provider visits. Larger-scale randomized clinical trials are needed in this understudied and critical aspect of adolescent medicine. Our findings prompt more investigation into the role of affective symptoms in perpetuating physical symptoms secondary to prolonged recovery from sports-related concussion.
We thank Drs Burton, Herring, Jingui, Krabak, Lockhart, Osario, Quitiquit, and Tsao, and F. Bennett, ARNP, for referring patients to the study, and we appreciate the efforts of study therapists Sarah Beals, Heidi Gray, Emily Law, and Daniel O’Rourke.
Seattle Sports Concussion Research Collaborative members included Monique Burton, MD (Department of Pediatrics, University of Washington); Sara Chrisman, MD, MPH (Center for Child Health Behavior and Development, SCRI and Department of Pediatrics and the Harborview Injury Prevention and Research Center, University of Washington); David Coppel, PhD (Department of Neurologic Surgery, University of Washington); Richard Ellenbogen, MD (Department of Neurologic Surgery, University of Washington); Seth Friedman, PhD (Center for Brain Integrative Research SCRI and Department of Radiology, University of Washington); Stanley Herring, MD (Departments of Neurologic Surgery, Orthopedics and Sports Medicine, and Rehabilitation Medicine, University of Washington); Thomas Jinguji, MD (Department of Orthopedics and Sports Medicine, University of Washington); John Lockhart, MD (Department of Pediatrics, University of Washington); Christine MacDonald, PhD (Department of Neurologic Surgery and the Harborview Injury Prevention and Research Center, University of Washington); Thomas McNalley, MD (Department of Rehabilitation Medicine, University of Washington); Marisa Osorio, DO (Department of Rehabilitation Medicine, University of Washington); Celeste Quitiquit, MD (Department of Pediatrics, University of Washington); Ali Rowhani-Rahbar, MD, PhD (Center for Child Health Behavior and Development, SCRI and Departments of Epidemiology and Pediatrics and the Harborview Injury Prevention and Research Center, University of Washington); and Monica Vavilala, MD (Departments of Pediatrics and Anesthesiology and Pain Medicine and the Harborview Injury Prevention and Research Center, University of Washington).
- Accepted July 1, 2016.
- Address correspondence to Carolyn A. McCarty, PhD, PO Box 5371, Suite 400, Mailstop CW8-5A, Seattle, WA 98145-5005. E-mail: firstname.lastname@example.org
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: Supported by the Seattle Sports Concussion Research Collaborative and by the National Center for Advancing Translational Sciences of the National Institutes of Health under CTSA grant UL1TR000423. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Funded by the National Institutes of Health (NIH).
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
- Institute of Medicine (IOM) and National Research Council
- Ransom DM,
- Vaughan CG,
- Pratson L,
- Sady MD,
- McGill CA,
- Gioia GA
- Ponsford J,
- Willmott C,
- Rothwell A, et al
- McCrory P,
- Meeuwisse W,
- Johnston K, et al
- Hopkins K,
- Crosland P,
- Elliott N,
- Bewley S; Clinical Guidelines Update Committee B
- Zou G
- Cohen J
- Copyright © 2016 by the American Academy of Pediatrics