BACKGROUND: Mild traumatic brain injury injuries (mTBIs), including concussions, represent >2 million US pediatric emergency department visits annually. Post-mTBI mental health symptoms are prominent and often attributed to the mTBI. This study examined whether individuals seeking post-mTBI mental health care had previous mental health diagnoses or a new onset of such disorders, and determined if mental health care utilization differed by race/ethnicity.
METHODS: Retrospective cohort study, using the Medicaid Marketscan claims national dataset (2007–2012). Utilization of mental health services 1 year before and 1 year after mTBI was compared between children with and without mental health diagnoses before injury. Primary outcome was receipt of post-mTBI outpatient mental health care.
RESULTS: A total of 31 272 children 20 years or younger were included, 8577 (27%) with mental health diagnoses before their mTBI and 22 695 without one. After injury, children without previous mental health disorders increased mental health services utilization; however, most (86%) postinjury mental health care was received by children with previous mental health disorders. Having a mental health diagnosis pre-mTBI was the most important risk factor for receiving post-mTBI mental health care (odds ratio 7.93, 95% confidence interval 7.40–8.50). Hispanic children were less likely to receive post-mTBI mental health care.
CONCLUSIONS: mTBI was associated with increased utilization of mental health services but most of these services were received by children with previous mental health disorders. Our documentation of racial/ethnic disparities in mental health care utilization reemphasize the importance of providing individualized, culturally, and linguistically competent care to improve outcomes after mTBI for all children.
- AIS —
- Abbreviated Injury Scale
- aRR —
- adjusted risk ratio
- CI —
- confidence interval
- ED —
- emergency department
- ICD-9 —
- International Classification of Diseases, Ninth Revision
- mTBI —
- mild traumatic brain injury
- TBI —
- traumatic brain injury
What’s Known on This Subject:
Mental health symptoms after mild traumatic brain injury (mTBI) are prominent and often attributed to the TBI. It remains unclear if utilization of post-mTBI mental health care is related to development of new symptoms or previous illness. No information on mental health care utilization among minority children exists.
What This Study Adds:
mTBI is associated with increased utilization of mental health services; however, most children seeking mental health care after mTBI had a previous mental health diagnosis. Hispanic children were less likely to receive post-mTBI mental health care.
Concussion and other mild traumatic brain injury (mTBI) have garnered increasing attention in the past few years.1 Emergency department (ED) visits for this problem have increased dramatically2 and recent estimates show that there are nearly 2 million concussions annually to youth from sports and recreational injuries alone.3
Although most individuals with concussion and mTBI recover rapidly, numerous studies have documented that many individuals continue to have various symptoms that persist for weeks to months after injury.1,4–6 Mental health symptoms, such as depression, anxiety, and attentional disorders, are prominent, and their presence is attributed to mTBI.5,7 In addition, most treatments for mTBI, such as cognitive rehabilitation and subthreshold exercise programs, are not directed to any accompanying mental health problems.8,9
Many studies have shown that the strongest predictors of persistent postconcussive symptoms in youth are previous mental health problems, especially depression and anxiety.10 Additionally, previous research has shown that children from racial and ethnic minorities are less likely to receive behavioral services11 and mental health care.12 However, to date, there is no information on receipt of such services for minority children after an mTBI.
The purpose of this study was to examine whether individuals seeking care for mental health problems after an mTBI had previous mental health diagnoses or have a new onset of such disorders after injury. In addition, we sought to determine whether there were racial and ethnic disparities in the receipt of post-mTBI mental health services. Our hypothesis was that most youth seeking care for mental health problems after mTBI had such problems before injury. We also hypothesized that children from minority racial and ethnic backgrounds would be less likely to use mental health services after their injury.
This retrospective cohort study used a large national data set of Medicaid claims, Medicaid Marketscan, previously used for pediatric traumatic brain injury (TBI) studies and described elsewhere.13 In brief, it contains individual-level inpatient and outpatient medical claims; each claim includes date, place, and type of medical service and associated diagnoses. It also provides demographic information on the patient at the time of service, and data on Medicaid enrollment.
We examined data from all 14 participating states in the Medicaid Marketscan dataset, for 2007 to 2012. All data were de-identified, including the names of the states; therefore, the study was considered exempt by the University of Washington institutional review board.
Participants were patients 20 years or younger with a diagnosis of a TBI between 2008 and 2011, who did not require hospitalization, and who were continuously enrolled 1 year before and 1 year after the injury. TBI was defined by the International Classification of Diseases, Ninth Revision (ICD-9) codes 800.0 to 801.99, 803.0 to 804.99, and 850.0 to 854.19, from the outpatient Marketscan dataset. In the Marketscan dataset, outpatient services are those services provided in outpatient facilities, such as doctor’s offices, hospital outpatient facilities, emergency rooms, or other outpatient facilities.
To identify the index mTBI, all outpatient claims with a TBI diagnosis were identified and only the first claim was considered. Once the index TBI was identified, we tracked patients in the inpatient dataset and excluded those who were hospitalized after their initial visit to ensure that our cohort included exclusively patients with mild TBI not requiring inpatient care. TBI severity was ascertained by using Abbreviated Injury Scale (AIS) scores, calculated by using the Stata (Stata Corp, College Station, TX) ICD-9 program for injury categorization.14 We excluded patients who had AIS scores ≥3.
Definition of Previous Mental Health Disorder
Patients were categorized according to their mental health status before the index mTBI. We defined previous mTBI mental health disorders by ICD-9 diagnostic codes for visits in the 12 months before the mTBI: 290 to 294 (organic psychotic conditions); 300 to 316 (neurotic disorders, personality disorders, and other nonpsychotic mental disorders); and 296 (episodic mood disorders). We excluded from our cohort patients with ICD-9 codes 295(schizophrenic disorders), 297 to 299 (delusional disorders, nonorganic psychoses, and pervasive developmental disorders), and patients with ICD-9 diagnostic codes 317 to 319 (intellectual disabilities) because the nature and progression of their disease could confound results. A flowchart of our cohort is presented as a supplemental figure (Supplemental Fig 3)
Definition of Outpatient Mental Health and Other Services
The primary outcome was receipt of post-TBI outpatient mental health services. These included provider claims (psychiatrist, clinical psychologist, and psychiatric nurse) and service claims (psychiatric and psychological tests and treatments, including individual, family, and group psychotherapies) coded in the claims data by provider, service, and revenue codes.
To measure overall health care utilization, we created 4 other groups of services. Rehabilitation therapy included rehabilitation services identified by the provider and service claims for physical, occupational, and speech therapy. Primary care services were defined by provider and service claims for visits to a pediatrician or family practice clinic. ED visits were grouped together, regardless of type of provider seen. All other provider visits, including visits for pain management, neurosurgery, neurology, orthopedics, and other pediatric subspecialties were categorized as other.
Due to the nature of the database, a single health service can be recorded more than once under provider and service claims. Therefore, we initially identified all claims (provider and service claims) and subsequently only 1 claim (provider or service) per type of service per day was counted. In that way, we avoided counting the same service more than once but maximized our ability to identify services.
Receipt of services was ascertained before and after the mTBI index date to allow for pre- and postinjury comparisons of health care utilization.
Definition of Covariates
Demographic variables included sex, age, race, and ethnicity. Age, in years, was used as a categorical variable (0–7, 8–14, and 15–20 years). Race/ethnicity was categorized by Marketscan as non-Hispanic white, non-Hispanic black, Hispanic, and other. Type of Medicaid plan was dichotomized as fee-for-service and capitated. Capitated plans include health maintenance and preferred maintenance organizations that are fully and partially capitated, respectively.
The longitudinal cohort included all claims that occurred in 12 months before the index mTBI date and in the subsequent 12 months. Some subjects had nonspecific non-TBI, head and face injury diagnoses (ICD-9 code 959.0), a week before the index mTBI. Considering that the nonspecific diagnosis might indicate that their mTBI happened earlier but diagnosis was delayed, we considered the first nonspecific injury code to be the index mTBI. The number of services was summed by month by type of service and an indicator variable for each service type for each month was also created. We examined demographic characteristics by the mental health disorder status pre-mTBI. We estimated monthly age- and sex-adjusted service types (mental health, rehabilitation, ED visits, primary care visits, and other types of services) by pre-mTBI mental health status using generalized estimating equations regression with a Poisson distribution and robust variance estimator. Marginal estimates for each month were graphed to visually explore service usage trends. Generalized estimating equations with a Poisson distribution and robust variance were also used to evaluate any mental health service usage post-TBI by mental health disorder status pre-TBI in a multivariable model adjusting for demographics and head AIS.
We identified 106 388 patients with an outpatient visit record for TBI. We excluded 2949 patients with previous psychotic conditions and intellectual disabilities, 44 233 patients who were not continuously enrolled in Medicaid during the year before injury, 26 254 who were not continuously enrolled in the year after the injury, and 1680 who had head AIS scores of ≥3.
We included 31 272 children younger than 21 years with a diagnosis of a mTBI, between 2008 and 2011, of whom 8577 (27.4%) had a diagnosis of a mental health disorder before their mTBI and 22 695 did not. Groups were similar in their distribution by sex, with most children being boys. Children with a previous diagnosis of a mental health disorder were older, more likely to be non-Hispanic white, and more likely to be covered by Medicaid capitated plans when compared with those without previous mental disorders (Table 1). The most common diagnoses among those with previous mental health disorders were attention-deficit/hyperactivity disorder (44%) and mood disorders (33%) (Table 1).
During the 12 months before the index mTBI, receipt of primary care, ED, and other specialty health services was similar among children with and without previous mental disorders. Utilization of rehabilitation therapy (speech, occupational, or physical therapy) and mental health services was significantly higher pre-mTBI (at baseline) among children with a previous mental health disorder compared with those without one. Children with a previous mental health disorder received on average 0.91 mental health services per month per child and 0.44 rehabilitation therapy services per month per child compared with 0.006 mental health and 0.05 rehabilitation services per month for children without a mental health disorder before the TBI (Fig 1). Psychological/mental health visits for children with no pre-mTBI mental health disorders were both for individual and family therapies. The diagnoses associated with these services were varied and in some instances nonspecific. These included counseling for parent-child problems (ICD-9 code V6120), counseling for child-abuse (ICD-9 code V6121), mental health services for family (ICD-9 code V619), and educational (V623) circumstances, as well as “other” psychological stress and circumstances (V6281, V6282, and V6289). Because these visits do not reflect a specific mental health disorder, we did not exclude these patients from the analyses.
After the index visit for the mTBI, regardless of the child’s previous mental health status, health care utilization for primary care and ED visits increased compared with baseline, with a peak in service utilization in the month of injury. Among children with previous mental health disorders, utilization of all services returned to their baseline by the second month after injury. For children without previous mental health disorders, receipt of primary care and ED services returned to baseline the second month after injury, whereas receipt of mental health services increased with a peak of 0.1 services per month per child at 12 months after injury. Rehabilitation therapy services also increased after injury with a peak of 0.1 services per month per child at 12 months (Fig 1).
When comparing total number of mental health services received before and after the mTBI, we found that children without previous mental health disorders had a significant increase in mental health services utilization after injury adjusted risk ratio (aRR) 9.03 (95% confidence interval [CI] 6.16–13.24), whereas mental health services utilization remained unchanged for the group of children with previous mental health disorders (aRR 0.99, 95% CI 0.95–1.04). However, children with previous mental disorders received the most mental health services after injury (Fig 2). Children with mental health disorders received 98% of all mental health services before injury and 85% after injury, although accounting for only one-quarter of all children in the sample.
Having a mental health diagnosis pre-mTBI was the most important risk factor for receiving any mental health services post-TBI (aRR 7.93, 95% CI 7.40–8.50). Other factors associated with a higher likelihood of mental health services post-mTBI included age, race, and ethnicity and type of Medicaid plan. Older children and children insured by Medicaid fee-for-service plans were more likely to receive any mental health services (Table 2). On the other hand, Hispanic children were less likely to receive mental health services when compared with non-Hispanic white children (Table 2).
In this study, we examined utilization of mental health services after mild pediatric TBI. In agreement with previous reports, our study found that utilization of mental health services increased after mTBI. New to our study is our finding that most of the mental health services were provided to children with previous mental health disorders, and that there were racial and ethnic disparities in utilization of mental health services. These findings address gaps in knowledge regarding mental health service use and access after mTBI.
In this study, we also provide longitudinal information on health services utilization after an mTBI. Children with previous mental health disorders increased their use of mental health services after injury, with a peak 1 month after injury and returning to baseline by 2 months after injury. We also found that children without a mental health disorder before their mTBI showed a significant increase in use of services the first month after injury, which continued over time with a peak in service utilization 12 months after injury. This slow and continuous increase in service use can be explained by a cumulative need for services among the group of children without a previous mental disorder, and is in agreement with previous research showing that onset of new post-mTBI psychiatric disorders and symptoms vary depending on the type and location of the mTBI.15Our study supports previous findings regarding the increased risk for children with mental health disorders experiencing higher mental health and behavioral problems after mTBI15 and underscores the importance of close follow-up of this patient population.
Barriers to access mental health services are multifactorial, and despite efforts to increase services to specific vulnerable populations, disparities persist. One study among children covered by the Children’s Mental Health Initiative, a federal government program aimed to increase access to mental health services for minority populations, shows that despite an increase in offered services, Hispanic, African American, and Pacific Islander children have lower use of these services than non-Hispanic white children.16 Additional barriers may include failure to perceive or diagnose a problem, implicit biases by health care providers, as well as lack of culturally and linguistically competent care.17 In this study, we provide new information that Hispanic children were less likely to receive mental health services after mTBI. Previous mTBI studies have not examined this association.18,19 This finding in mild TBI is similar to previous studies in moderate and severe TBI in which Hispanic children have poorer outcomes compared with non-Hispanic white children and are also less likely to access post-TBI services.20,21 This study is also consistent with non-TBI studies documenting lower receipt of mental health services among Hispanic and “other” race children22 and lower receipt of antipsychotic prescriptions among Hispanic, African American, and Asian children covered by Medicaid.23 Based on our findings and previous findings from other studies, we recommend clinicians caring for minority patients, especially those with language and cultural barriers, to increase efforts to educate these families about possible mental and behavioral problems after mild TBI, emphasizing the importance of consulting if these symptoms are apparent. Close follow-up of these children might be warranted to diagnose new mental or behavioral disorders that might be missed by parents or that parents might be reluctant to address.
We found that age was a factor associated with receipt of mental health services, and that older children received more mental health services than younger children, findings that are in agreement with previous research.18,19 We also found that children insured under capitated plans are less likely to receive services, echoing results from an earlier study in which children covered under capitated Medicaid plans report higher unmet mental health care needs.24 However, not all studies report similar findings. One study among Medicaid-insured children in the District of Columbia report that children covered under Medicaid capitated plans experienced fewer barriers to access services compared with children covered by fee-for-service plans.25 Medicaid plans and coverage vary from one state to another. Our sample included children from 14 different states and therefore provides a broader picture. Nonetheless, due to confidentiality agreements of Marketscan, we could not assess differences between states.
Our study has some limitations. Data from states participating in Marketscan are protected by confidentiality agreements that preclude identifying individual states limiting our ability to analyze regional variations. Data are restricted to administrative claims and therefore we could determine only receipt of billed services; services provided by the school system or community organizations are not included. Identification of an mTBI was done by using ICD-9 codes used during outpatient visits, a method that has not been validated. Nonetheless, we excluded patients based on hospitalizations after their injury, and patients with high injury severity scores to ensure that our cohort included exclusively patients with mTBI not requiring inpatient care. The definition of mental disorders was limited to ICD-9 diagnosis associated with billed services before the mTBI. It is possible that some of the children in the non–mental health group would actually have undiagnosed mental disorders. If that was the case, our results would have underestimated the effect of having a previous mental disorder. Our study cannot estimate the actual need for services. This is an important consideration because children of low socioeconomic status are at higher risk of experiencing mental health pathology and are also more likely to report unmet mental health needs. We cannot determine if post-TBI mental health services were needed as a result of a new mental health diagnosis, an exacerbation of a previous one, or a new diagnosis of a previous disorder prompted by the TBI visit. Last, our study population is restricted to children covered by Medicaid, and therefore baseline characteristics may be different from the general population.
mTBI was associated with an increase in utilization of mental health services, but most of these services were received by children with previous mental health disorders. Our documentation of disparities in the receipt of mental health services for Hispanic children reemphasizes the importance of providing individualized, culturally and linguistically competent care, to improve outcomes after mTBI for all children.
We want to thank Janessa Graves, PhD, MPH, for her methodological discussions on this study.
- Accepted November 22, 2016.
- Address correspondence to Nathalia Jimenez, MD, MPH, Department of Anesthesiology and Pain Medicine, University of Washington, Box 9824, 4800 Sand Point Way NE, Seattle, WA 98105. E-mail:
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: This work was supported by the National Institute of Child Health and Human Development (1K23HD078453-02 to Dr Jimenez), and the Centers for Disease Control and Prevention (1U01CE002196 to Dr Jaffe). 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.
- Graham R,
- Rivara FP,
- Ford MA,
- Spicer CM
- Centers for Disease Control and Prevention
- Bryan MA,
- Rowhani-Rahbar A,
- Comstock RD,
- Rivara F; Seattle Sports Concussion Research Collaborative
- McCrory P,
- Meeuwisse WH,
- Aubry M, et al
- Kurowski BG,
- Hugentobler J,
- Quatman-Yates C, et al
- Hacker KA,
- Penfold RB,
- Arsenault LN,
- Zhang F,
- Murphy M,
- Wissow LS
- Jimenez N,
- Symons RG,
- Wang J, et al
- Clark DE,
- Turner M
- Cataife G,
- Weinberg DA
- Mitchell JM,
- Gaskin DJ
- Copyright © 2017 by the American Academy of Pediatrics