Introduction

Mental disorders place a great challenge on the health care system. They are highly prevalent not only in adults [1] but also in adolescents and children [2], [3], [4], [5], [6], [7], with enormous implications for health-related quality of life [8] and health economy [9]. The urgent need to integrate mental health into all aspects of health research and health-care delivery has been realized by major health institutions and scientific journals, as documented for example in The Lancet series on Global Mental Health and strategic research goals and health action plans by the Grand Challenges in Global Mental Health initiative and the American Academy of Pediatrics [10], [11], [12].

Many children and adolescents with mental disorders do not receive adequate care [13], [14], even though effective treatment approaches for mental disorders exist [15]. The median interval between the first onset of a mental disorder and first contact with the treatment sector is nearly a decade [16], although early intervention offers the hope to prevent long-term mental health problems and increase wellbeing and productivity [17].

The school sector has been reported to be a major contact point for children and adolescents with emotional or behavioral problems [13], [18], [19], [20], and the further development of school-based mental health promotion programs for children and adolescents ranks among the top ten challenges identified by the Grand Challenges in Global Mental Health initiative [10]. Results of the International School Psychology Survey reveal that the number of students per school psychologist is alarmingly high, counting from 470 students per school psychologist in Italy to more than 26000 students per school psychologist in parts of Germany [21]. Besides this low ratio of school psychologists to students, the function of school psychologists is rather complex and involves a wide range of tasks among which only one is intervention supply [22], suggesting that sound and comprehensive treatment of children and adolescents with mental disorders remains hardly feasible in the school setting to date. Instead, the school sector has unrivaled access to children and adolescents and, with this unique feature, is exceptionally situated to link children and adolescents with mental disorders to out-of-school services addressing mental health problems and to high-quality care at early disorder stages. However, the role of the school sector as a guide to such out-of-school services remains poorly understood [18], [23]. To improve our understanding of common trajectories of service use in children and adolescents, it has been claimed that it is important to clarify the sequence of health care sector utilization and to move beyond studying single sectors to the study of the potentially more relevant patterns of sector use [13], [24].

The main objective of this study was to estimate in a nationally representative United States cohort, focusing on adolescents with a lifetime mental disorder, the role of school mental health services as guide to care in out-of-school service sectors for the treatment of mental disorders.

Materials and Methods

Statistical analyses

All statistical analyses are based on post-stratification weighted data, which allowed for the correction of minor discrepancies in distributions of school and/or sociodemographic characteristics between the sample and the population [25]. We performed the descriptive analyses of service utilization in different sectors and of the sample's sociodemograhic characteristics by calculating frequencies and percentages of these discrete variables. We computed separate discrete-time proportional hazard models with a non-parametric baseline hazard function, using complementary log-log regression to estimate temporal relationships between school and out-of-school service use for mental disorders, with school service use (versus no school service use) defined as time-varying predictor and mental health specialty service use, medical specialty service use, other service use, and any out-of-school service use defined as outcomes [34]. Analyses were performed in the subsamples of children and adolescents with any i) affective disorder, ii) anxiety disorder, iii) behavior disorder, iv) substance use disorder, v) eating disorder, and vi) mental disorder. We present hazard ratios (HR) and their 95% confidence intervals (CI). To account for the complex structure of the survey data, we used the Taylor series linearization method.

We re-evaluated the results after adjusting p values of the 24 main tests for multiple comparisons using the Holm formula [35], [36], [37], thereby accounting for the number of tests.

To validate our results, we conducted the following secondary analyses: To control for potential confounding, we repeated the analyses after adjusting for several sociodemographic factors, as sociodemographic factors have previously been shown to influence mental health service utilization in children and adolescents [13], [38], [39], [40]. We used those sociodemographic factors that have been included in previous epidemiological studies of child and adolescent mental health as predictors of mental disorder prevalence [7], and simultaneously included them in the analyses, with the categories indicated in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Information on mental health service utilization in different sectors and on sociodemographic characteristics of the study sample (N = 3656^*).

https://doi.org/10.1371/journal.pone.0099675.t001

Moreover, we repeated the analyses three times, controlling for either i) number of categories in which a mental disorder was present (1/2/> = 3), ii) age of onset of the type of mental disorder characterizing the subsample in which the analysis was performed (affective disorders: 1–10 years/11–12 years/13–14 years/15–18 years; anxiety disorders: 1–4 years/5–6 years/7–10 years/11–18 years; behavior disorders: 1–4 years/5–7 years/8–12 years/13–18 years; substance use disorders: 1–13 years/14 years/15 years/16–18 years; eating disorders: 1–12 years/13 years/14 years/15–18 years; any mental disorder: 1–4 years/5–6 years/7–11 years/12–18 years; limits between the categories were defined by cutoffs as close as possible to the quartiles of the respective age distributions), or iii) presence of further types of mental disorders not characterizing the subsample in which the analysis was performed (yes/no).

We interpreted with caution all results that were statistically significant in the main analyses but not stable after Holm correction or secondary analyses.

There were a low percentage of subjects with missing information on service use and we dealt with missing data by restricting each analysis to subjects with complete data (see Table 2). We evaluated the results using two-sided tests of significance set at a level of 0.05. Statistical analyses were performed using STATA/MP 13 (Stata Corporation, College Station, Texas).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Discrete-time proportional hazard models for school mental health service utilization (time-varying) predicting out-of-school service use in different sectors.

https://doi.org/10.1371/journal.pone.0099675.t002

Results

Discussion

This study, conducted with 6483 adolescents of a nationally representative United States cohort, provides as yet missing association estimates of school and subsequent out-of-school service utilization for mental problems in children and adolescents with a lifetime mental disorder. Results indicate that school services may serve as guide to certain out-of-school service sectors addressing mental disorders, especially the medical specialty sector, but not the mental health specialty sector.

Our results complement previous descriptive evidence regarding service use patterns in adolescents with mental disorders [13], [18], [23] by elucidating the role of the school service sector in the trajectory of child and adolescent service utilization for mental disorders, and suggest that the school sector plays a central role as guide to the medical specialty sector.

The roles of the medical and the mental health specialty sectors in mental health care have been addressed previously. Several studies have confirmed the prominent position of the medical specialty sector in providing services for mental disorders [41], even though the mental health specialty sector has been reported to be the major service sector for children and adolescents with mental disorders [13] and the most common subsequent service sector in adolescents entering the service system through the school mental health sector [18]. While mental health care in the medical specialty sector has improved over the past decade through a better understanding of mental disorders, the availability of practicable screening tools, provision of psychotherapy in these settings, and further development and promotion of psychotropic medications that are mostly prescribed by primary care physicians [24], [42], [43], [44], most primary care pediatricians and child and adolescent psychiatrists believe that pediatricians should identify and refer, but not treat, their patients' mental health problems [45]. Indeed, the appropriateness of treatment of mental disorders in the medical specialty sector as compared to the mental specialty sector remains a matter of debate [46], [47], [48], [49], [50], which should be kept in mind when evaluating the referral patterns observed in the present study.

There is considerable evidence that sociodemographic characteristics of children and adolescents with mental disorders influence whether or not services for mental disorders are utilized; such sociodemographic characteristics include, amongst others, ethnicity, family income, different patient, parental and familial factors, health insurance coverage, and service availability [13], [38], [39], [40], [51], [52], [53]. Much less is known about what influences, in those making use of mental health care, the sequence of sector utilization on their pathway through the service system; for example what influences whether children/adolescents with mental disorders are seeking care in the medical sector or the mental health sector. Therefore, it has been claimed that it is important to clarify the role of potential factors, including referral patterns by professionals, service availability, child and family preferences, professional competencies, and financial considerations, in order to better understand what determines the service paths of children and adolescents seeking help for mental disorders [13].

Our study has several strengths, including the large nationally representative sample [25], a fully structured diagnostic interview covering a wide range of adolescent mental disorders, with good quality criteria [25], [29], and the integration of adolescent and parent information on adolescent mental disorders, as previously suggested [32]. Given the good response rate and the small amount of missing data, it is unlikely that loss of subjects has introduced relevant selection bias.

The study also has several limitations to be considered, including amongst others the cross-sectional design of the study, as previously discussed [28], [54].

Moreover, first, service utilization has been assessed by self-report, but the validity and reliability of such data has been challenged, especially in terms of non-response and recall bias, which is no issue in large health care utilization databases [55], [56]. However, the NCS-A has not been linked with health care utilization databases, but instead provides self-reported information on service utilization for mental disorders based on a well-established instrument [13], [33]. In adults, concordance between self-reported health care and mental health care utilization and registration data has been shown to be fair [57], [58], [59], and the rather low magnitude of bias in estimates of health care utilization due to non-response in health surveys has previously encouraged the continued use of interview health surveys [60]. A study on the reliability of children's and adolescents' responses on the Child and Adolescent Services Assessment, the groundwork of the Service Assessment for Children and Adolescents [33], revealed that reports of lifetime service use were as reliable as were reports of service use in the preceding three months, even though it should be acknowledged that respondents reported inpatient, other overnight and juvenile justice services more reliably than outpatient and school services [61]. It should also be noted that relying on health care utilization databases faces other limitations, including the exclusion of services provided outside the health care system [56], and the advantages of checklists over routine data sources have been previously documented [62].

Second, even though we controlled for several potential confounders, we cannot exclude confounding by unconsidered factors, including disorder severity. However, we did not reveal a relationship between school mental health service use and subsequent utilization of the mental health specialty sector, and controlling for number of disorder categories did not change the results remarkably, both of which suggests that disorder severity did not confound the presented results in a relevant manner.

Third, the observed associations do not inform about causal but about temporal relationships, and even though the study of causality is wanted, it remains challenging in such a large representative cohort. Until such studies are available, the elucidation of the position of the school sector in the trajectory of service use for mental disorders allows for a better understanding of common service paths of children and adolescents, as has previously been claimed [13].

Fourth, the aim of our analyses was to estimate whether school mental health service use serves as a guide to the use of any out-of-school service sectors; however, we did not compare school service use with use of a specific out-of-school service sector regarding their prediction of service use in further sectors. Therefore, it may be interesting to scrutinize in future studies whether the here observed referral patterns are specific for school mental health services.

The clinical and public health relevance of the role of the school sector in the trajectory of service use for mental disorders in children and adolescents becomes evident against the background of the urgent need to reduce unmet mental disorder treatment demands and improve access to mental health care [13], [16], together with the limited ability of school mental health services to provide comprehensive treatment on-site on the one hand and their unique position in accessing children and adolescents with mental disorders and linking them to out-of-school services on the other hand [21], [63]. It has previously been criticized that the coordination of the different health care sectors is often insufficient, thus representing a barrier for the patients and impairing the supply with appropriate treatments [64]. By raising the awareness of unidirectional referral patterns of school mental health services, our results may pave the way to improve the school mental health service system and multilateral collaborations between sectors where required, which is in line with current strategic research goals and task forces [10], [63]. Notably, our findings raise the question as to whether the school mental health sector sufficiently takes advantage of its unique position in linking mentally disordered students with the out-of-school service system beyond the medical specialty sector.

The supply of mental health services in places where children and adolescents are easily accessible is only one among several action steps that have been proposed to increase access to mental health care for children and adolescents with mental disorders. Further suggestions include the improvement of access to information on available mental health care options, inclusion of children and adolescents in treatment planning, and offering assistance in finding one's way through the rather complicated service system [65]. The bottom line is that, to be most effective in increasing access to child mental health care, strategies should target not only at structural barriers to care, such as inadequate insurance coverage, but also at barriers related to the understanding of mental health problems and services; for example, public education campaigns may help to increase the awareness and knowledge of mental disorders and mental health care services [66], [67].

Whether the here presented role of school mental health services as a guide to out-of-school services for mental disorders in the United States is generalizable to other countries remains to be elucidated, as not only the school mental health system but also insurance and health care systems, including access to mental health care systems, vary considerably worldwide [64], [68].

Future studies should focus on the role of school mental health services in the trajectory of mental health service use in other countries. Moreover, future research should include prospective data and the integration of self-report of service utilization and service utilization databases. It will also be important to evaluate the appropriateness of the mental health service use trajectory emanating from the school mental health sector and to learn about modifiable factors influencing referral patterns of school mental health services, in order to be able to take action and guide or optimize such referral patterns, if necessary.

To the best of our knowledge, this is the first comprehensive study of the role of the school mental health sector as a guide to mental health care in out-of-school sectors, using data of mentally disordered adolescents of a nationally representative United States cohort. Results indicate that school mental health services may serve as a guide to certain out-of-school service sectors for children and adolescents with various mental disorders, especially the medical specialty sector, but not the mental health specialty sector. While highlighting the relevance of school mental health services in the trajectory of mental care, our findings also suggest to further investigate into a stronger collaboration between school mental health services and mental health specialty services in order to improve and guarantee adequate mental care at early life stages.

Acknowledgments

Disclaimer: Hereby, we acknowledge that the original collector of the data, ICPSR, and the relevant funding agency bear no responsibility for use of the data or for interpretations or inferences based upon such uses.