Prescriptions of Psychotropic Medications by Providers Treating Children of Military Service Members

Introduction: There are approximately 1.5 million U.S. military-dependent children. However, little is known about mental health referrals for these youths. This study sought to examine the type of mental health treatment referrals made by primary care providers for child military-dependent beneficiaries receiving care in the direct (within Military Treatment Facilities) and private care (civilian-fee-for service facilities) sectors of the Military Health System. Materials and Methods: A between-subjects, cross-sectional study was performed on children aged 5–18 years old in fiscal years 2011–2015 and enrolled in TRICARE Prime. Study analyses examined specialty (“talk therapy”) mental health care and psychotropic medication referrals from TRICARE Prime (the Defense Health Agency-managed health care program) providers for beneficiary children diagnosed with attention-, mood-, anxiety-, or behavior-related disorders in direct versus private sector care. Results: Of 1,533,630 children enrolled in TRICARE Prime (50.03% female), 8.6% ( n = 131,393) were diagnosed with a psychological disorder during FY 2011–2015. Most were attention-related (5.2%, n = 79,770), followed by mood (1.7%, n = 25,314), anxiety (1.1%, n = 16,155), and conduct-related diagnoses (0.7%, n = 10,154). Adjusting for age, sex, and sponsor rank, children within direct care diagnosed with attention-related disorders were 1.7 times more likely to receive a prescription for psychotropic medication than those in private sector care, odds ratio (OR) = 1.72, 95% confidence interval (CI): [1.66, 1.77]. Children diagnosed with mood-related disorders in direct care were 2.1 times more likely to receive a prescription for psychotropic medication than those in private sector care, OR = 2.08, 95% CI: [1.96, 2.21]. Across disorders, children who received private sector care were more likely to have a referral specialty mental health (“talk therapy”) follow-up ( ps <0.0001). Conclusions: For attention- and mood-related disorders, but not anxiety- or conduct-related disorders, direct care providers were more likely than private sector care providers to prescribe psychotropic medications. Inconsistencies of provider referrals within and outside of the Military Health System should be elucidated to determine the impact on outcomes.


INTRODUCTION
Over 42% of military service members have at least 1 child, with an average of 2 children per active duty household. 1 Children of service members have protective factors many civilian children do not, such as parental employment, low poverty levels, and access to health care. 2 Yet, they may also be at higher risk for psychological problems than their civilian counterparts. Military children face unique stressors that civilian children do not encounter, including, but not limited to, multiple and frequent relocations, parental deployments, and fears about parental safety. 3 The culture of the military family may result in an increased need for psychological services and potentially higher rates of referrals for psychopharmacological prescriptions for military-dependent children. Family dynamics and unique career concerns due to a parent's military status may contribute to a military child's mental health treatment. [4][5][6] However, there is a lack of information regarding the type of treatment mental health referrals provided for these children.
In children, the most common psychological diagnoses are generally attention, mood, anxiety, and behavior/conduct disorders. 7,8 While military children are socio-demographically diverse and representative of the nation's children, [9][10][11] the accumulation of these aforementioned stressors may put military children at higher risk for psychological distress than civilian children. 12,13 In general, there has been an increase in behavioral health-related visits of military children. 6 For psychological disorders, treatments can typically take the form of behavioral or "talk" therapy, medication, or some combination of these 2 methods. 14 The American Academy of Child and Adolescent Psychiatry and the Society of Clinical Child and Adolescent Psychology recommend that many psychological diagnoses be treated first and primarily with evidence-based behavioral or "talk" therapy modalities. 15 Yet, the prescription of psychotropic medications for children and adolescents with psychiatric disorders has become increasingly common over the last 2 decades. Of concern, medications that were developed for adults are being delivered to children without thorough understanding of the long-term impact they may have on physical, cognitive, and emotional development. 16 TRICARE is a Defense Health Agency-managed health care program for uniformed service members, retirees, and their families. This study focuses on children enrolled in TRICARE Prime, which is the managed care option for beneficiaries. 17 TRICARE Prime coverage includes talk therapies and psychopharmacotherapy. 17 In TRICARE Prime, there is no out-of-pocket cost for pharmaceuticals filled at a Military Treatment Facility (MTF) and is referred to as "direct care." By contrast, individuals in TRICARE Prime can also receive care in a civilian setting ("private sector care"), which does involve additional fees. The Military Healthcare System itself factors into the treatment of the military child. Given that many physicians who treat this population are also service members, they may hold beliefs about the treatment of mental illness stemming from their service or their academic background. 18,19 For example, stigma regarding mental health, which has been well-established in the military culture, could also play a role in treatment dissemination and utilization, 1,20 potentially prompting a greater likelihood of prescribing a medication over talk therapy for military children.
We, therefore, examined within the Military Health System whether primary care providers who practice in MTFs (direct care), compared to those who practice in civilian treatment facilities (private sector care), treating attention, mood, anxiety, and behavior/conduct disorders, are more likely to use medication as a first-line treatment for military children rather than refer to nonpharmaceutical or talk therapy. We hypothesized that, compared to those practicing in civilian facilities, primary care providers who practice in MTFs would more often prescribe medication as a first-line treatment for the aforementioned psychological disorders. In addition, we expected MTF providers would be less likely to provide a specialty "talk therapy" mental health care follow-up referral.

Population Data
We conducted a cross-sectional analysis using the Military Health System Data Repository (MDR) between fiscal years 2011 and 2015 (October 1, 2010 through September 31, 2015) to identify TRICARE Prime beneficiaries aged 5-18 years old with a dependent beneficiary (child) status. The MDR houses all health care encounters and claims for the Military Health System beneficiaries receiving care at both MTFs and civilian treatment facilities through their TRICARE benefit. 9-11 TRI-CARE Select beneficiaries were excluded, as were those in overseas and overseas remote locations. This work was found exempt by the Institutional Review Board of the Uniformed Services University of the Health Sciences.
Of those identified in our cohort, we queried all of their health care claims from fiscal years 2011-2015 for the first visit in which patients were diagnosed with attention-, mood-, anxiety-, and behavior-related disorders and were seen in a primary care clinic by one of the following providers: Nurse Practitioner, Family Medicine Physician, Internal Medicine Physician, Generalist, Physician Assistant, or Pediatrician. As data precluded matching a specific provider to a psychotropic prescription, we defined subsequent treatment as psychotropic prescriptions filled within 3 weeks after and mental health visits completed within 4 weeks after the patient's initial diagnosis (i.e., psychotropic medication vs. talk therapy). As there is no specified timeframe for the filling of a prescription, 3 weeks was chosen to try to minimize the possibility that the child had a follow-up appointment (often conducted at 30 days) and received the prescription for the medication at a subsequent, rather than the initial, visit. Specialty mental health care was identified by selecting claims associated with the selected mental and behavioral health disorders and care provided by psychologists, psychiatrists, and mental or behavioral health social workers, counselors, and specialty nurses. Data from dependent children were excluded if the child was seen in both the direct and private sector care settings. Concurrent diagnoses were not identified among the study cohort or examined in our analysis.

Analyses
Descriptive data for demographics, including age, sex, sponsor's rank, and branch of service are presented. Sponsor's rank (i.e., enlisted or officer) was used as a proxy for socioeconomic status. Independent samples t-tests were used to compare mean ages, and chi-square tests were used to make comparisons in the remaining demographic categories for all beneficiaries in either direct or private sector care settings. Multivariate logistic regressions, with 95% confidence intervals and adjusted by age, sex, and sponsor's rank, were performed within each diagnostic category (attention, mood, anxiety, and behavior/conduct) population to determine the probability of a beneficiary receiving medications and followup mental health care from primary care providers in direct versus private sector care. Any children who received care in both direct and private sector care during the study period were excluded from the regression analyses. Statistical significance was set at a priori α < .05, and all analyses were conducted using SAS, Version 9.4.

RESULTS
We identified a total of 1,533,630 dependent children aged 5-18 years old enrolled in TRICARE Prime from fiscal years 2011 to 2015. Among these youths, primary care providers diagnosed 5.2% with an attention-related, 1.7% with a moodrelated, 1.1% with an anxiety-related, and 0.7% with a behavior/conduct-related disorder.   Care setting was also significantly related to children receiving specialty mental health follow-up in all of the diagnostic categories of interest. Specifically, in all diagnostic categories, children in direct care were less likely than those in private sector care to receive specialty ("talk therapy") mental health follow-up (Table III). Out of all of the diagnostic categories, children diagnosed and treated in direct care for mood-related disorders had the lowest odds of receiving mental health follow-up (0.25 OR, 0.24-0.27 95% CI, P < .0001; 0.26 AOR, 0.24-0.27 95% CI, P < .0001).

DISCUSSION
This study examined the provider treatment referrals for child TRICARE beneficiaries diagnosed with common mental health diagnoses by providers in direct care, or MTFs, to providers in private sector care, or civilian facilities. Children with attention-and mood-related disorders who received care in the direct care system were significantly more likely to receive psychotropic medication prescriptions than children with the same diagnoses who received care in the private sector care (or civilian) system, even after adjusting for other important factors that might contribute to medical practices. However, there were no differences in prescription practices between care systems in children with anxiety-and conductrelated disorders. Moreover, for all disorders, children in private sector care were more likely to receive follow-up "talk therapy" or nonpharmacologic mental health care.
Providers working in MTFs were more likely than providers in civilian facilities to prescribe medication for attention-and mood-related disorders. While this study focused on healthcare-level processes, previous studies suggest financial and convenience considerations on the parents' part may play a role for these findings. Direct care beneficiaries receive medications at no out-of-pocket cost at an MTF and can fill their prescriptions in the same facility in which they receive their medical care. 21 By contrast, as of 2019, the cost of filling a prescription at a network retail pharmacy was $11 for generics, $28 for brand names, and $53 for nonformulary medications. 21 In a review of qualitative studies on parents of nonmilitary children with attention deficit disorder, one of the factors found to impact treatment choices these parents made was the cost of stimulant medications. 22 In a review of parental willingness to use psychotropic medications for a child's mental illness, cultural factors including race and ethnicity, language, and biomedical views about illness mediated willingness to medicate. 23 If culture can influence the choices one makes regarding medical treatments, and the military has its own culture, which includes shared attitudes and beliefs, military culture could impact military members' decisions about their child's treatment. At least one parent in the population of interest is a uniformed service member. Furthermore, talk therapy can require a large time commitment. Service members themselves may have duty commitments that prevent them from missing work to take their child to therapy appointments. 24 The most frequently cited barrier to care for service members for their own mental health care was the necessity of missing work to attend appointments. 25 Further, talk therapies, particularly for younger children, often include parental participation. [26][27][28][29] As military members underutilize mental health services in general, 24 they may opt for psychotropic treatments rather than talk therapy which may involve themselves as well as their children. These and other parental factors should be explored in future studies.
Provider logistical considerations may account for our findings. A provider in a primary care setting may only have limited time to assess and treat each patient. Additionally, due to many providers in the MTFs serving in the military themselves, they likely have additional duties, outside of their primary job, for which they are responsible (e.g., Voting Assistance Officer, Command Fitness Leader, Combined Federal Campaign Coordinator). By contrast, providers in civilian facilities may solely be focused on providing patient care. In studies of general practitioners and internal medicine physicians, visits of shorter duration have been associated with higher rates of prescriptions. 30,31 Another possible contributing factor to higher prescription rates among providers working in MTFs are legal precedents that give military medical providers protections from lawsuits brought by service members. The Federal Tort Claims Act provides medical malpractice liability coverage for medical providers working in military healthcare centers. 32 According to this Act, a provider in these institutions cannot be sued directly for medical malpractice. A claimant must instead file suit with the U.S. Government, which accepts responsibility for damages. Uniformed providers may then be less reticent to prescribe some medications, for example, SSRIs, than their civilian counterparts, who may be more restrained in prescribing "black-box" SSRI medications to children; this may partially explain much higher rates of prescriptions for children with mood disorders than for those in other diagnostic categories. Finally, the stigma of mental illness among the medical and military communities is another important consideration. Mental illness is often seen as a character flaw by physicians, 19 and fear of stigma and documentation is a barrier to care for medical students. 18 It is possible that providers in MTFs may view psychotropic treatment as less stigmatizing than talk therapy and thus may partially account for our finding that, across all disorders, private sector care was associated with more specialty mental health care referrals.
There were no differences in psychotropic prescriptions for behavioral-or anxiety-related disorders. Because conduct disorder is defined as a specific pattern of behavior, 33 it may be considered as less biologically based and more psychologically based. Supporting this notion, there are few approved medications for conduct-related disorders, 34 and the American Academy of Child and Adolescent Psychiatry does not recommend using medication to treat conduct disorders. 35 There are more medications for anxiety-related disorders; however, benzodiazepines are highly addictive, and children and adolescents may abuse these medications or suffer withdrawal syndrome upon discontinuation. 36 As a result, providers from both direct and private sector care settings may avoid prescribing these medications. Moreover, cognitive-behavior therapy has been widely tested for anxiety in pediatric samples, and its effectiveness is well-established and accepted in military medicine. 28,37,38 Although offering explanations for the differences between diagnostic categories, significant differences in practices between the 2 care settings cannot be understood by these reasons alone. We would expect the prescription rates would be the same regardless of care setting if this was solely an effect of an overall increase in psychotropic use among children with attention and mood disorders.

Strengths and Limitations
Study strengths include the size of the population. The database used includes all children who receive care in the entire Military Health System, including MTF facilities and the civilian purchased care sector throughout the continental United States. It included child dependents of parents of all ranks and socioeconomic statuses. Thus, these children are geographically diverse and socio-demographically representative of the nation's children. [9][10][11] However, despite the extremely large sample size, clinical significance of differences must be interpreted cautiously given the small effect size. Insurance claims data cannot provide information about the drivers of the treatment the beneficiary received. It also was necessary to manually match codes (e.g., provider types) used in the 2 claims systems. Further, methods to obtain rates of prescription and follow-up were derivative of TRI-CARE policies as well as experiences in practice and were based primarily on assumptions. For example, the assumption is that the mental health service appointment following a primary care appointment was a result of a referral by the PCM rather than a pre-existing appointment or one that was made against the advice of the PCM. It would be beneficial to readdress this study's aims with a focused geographic population to answer more specific questions about differences in practices between provider types (e.g., Family Medicine vs. Pediatrician). Other limitations include a lack of available data on race and ethnicity and that analyses on more recent data are needed. Indeed, research in civilian samples indicates that under-represented minorities tend to be less likely to opt for psychotropic treatment. 39 Additionally, claims data cannot provide information about the drivers of the treatment the beneficiary received, including the role of parental decisionmaking in a child's treatment. However, this study is the first of its kind on military children, and findings should be considered a first step for generating hypotheses for future research. As this study was cross-sectional, directionality and causality cannot be determined. Importantly, it is unclear as to whether treatment practices in either direct care or private sector care translate to child psychological outcome.

Conclusion
In conclusion, the findings from this study provide an initial understanding the types of mental health treatment referrals for child military dependents. As the effects of prolonged wars, high operational tempos, and long separations continue to be a concern for military healthcare, the inconsistencies found in how the children of service members are treated in different settings warrant further exploration. Particularly in light of the increase in mental health concerns among youth and the changing landscape of mental health care since the start of the COVID-19 pandemic, 40 future research should elucidate the discrepancies between recommended treatments in civilian and military facilities and the relevance for psychological outcomes of child military dependents.