Abstract
When children with special educational needs are excluded from school, it should raise the concern that these children are not receiving adequate help and support. This systematic review aims to identify the prevalence of psychiatric disorder or impairing psychopathology among children who are excluded from school compared to children who are not excluded from school. The review follows the guidelines set out by the Centre of Research and Dissemination (Centre for Reviews and Dissemination, 2009). We used broad search terms (across 16 databases) in order to avoid missing papers, our initial screening stage involved 3712 titles and abstracts. Five studies finally met our inclusion criteria. The results from these studies suggest that the odds of children who have been excluded from school having some kind of impairing psychological or behavioural impairment is increased in comparison to children who have not been excluded (odds ratios ranging from 1.13–31.9). However, the paucity of primary research in this area, combined with methodological weaknesses in the identified studies makes it difficult to draw firm conclusions. The need for education and mental health professionals to work together to provide the appropriate support for all children at risk of exclusion is of great and immediate concern.
Keywords
In the United States of America, during 2006, approximately 7% of students were suspended (fixed term exclusion) from school at least once during the year, which was an increase from 3.1 million in 2002. The number of expulsions (permanent exclusions) also increased from 89,100 in 2002 to 102,100 in 2006 (Planty et al., 2009). In Australia, the number of suspensions from school is reported to range from 2.9% in higher socio-economic status (SES) children to 8.7% in lower SES children (Hemphill et al., 2010). While in Europe, Government statistics from England, UK, suggest that although the total number of children excluded from school decreased between 1998–2008 (to approximately 5% of the total school population), the proportion of excluded children with special educational needs (SEN) increased (from 61% in 2001/02 to 73% in 2011/12) (Department for Education, 2013). These figures are also supported by a recent inquiry in to school exclusion in England, conducted by the Children’s Commissioner in the UK (Children’s Commissioner, 2012). One of the main reasons for SEN designation is social, emotional and behavioral difficulties, and the exclusion of children with SEN should raise the possibility that despite the identification of additional needs, the support provided may be inadequate. This may be particularly salient for children of minority groups, and those from socio-economically and socially disadvantaged backgrounds, who may lack effective advocates for their needs (Shirley & Cornell, 2011).
There is already evidence of the difficulties that children with special educational needs in school have around the world (Lebeer et al., 2012; Ruttledge & Petrides, 2012) and in particular children with socio-emotional needs (Candelaria, Fedewa, & Ahn, 2012; Hutchings et al., 2012). The National Centre for Education Statistics at the Institute of Education Sciences (National Centre for Education Statistics, 2002) states that in the USA students with mental health problems that are disturbing their participation in school are more likely to be moved to an ‘alternative school’ or a ‘programme for at risk students’ than other students. Students who are suspended or expelled are not eligible for such a transfer, sometimes referred to as a ‘managed move’, which may reduce the prevalence of psychiatric disorder or impairing psychopathology reported among children excluded from school in the USA. Despite this, a report by the National Longitudinal Transition study in the USA suggests that children aged 11- to 17-years-old with emotional disturbances or ADHD (Attention Deficit Hyperactivity Disorder) are still more likely to have been suspended or expelled in comparison to children with any other impairments or disabilities (Gonzalez, 2006).
The relationship between childhood psychiatric disorder or impairing psychology and suspension/expulsion from school is an important issue. Research suggests that mental health problems in children intensify over time and that being excluded from school worsens educational (and other) prospects for the child (Parsons, Godfrey, Howlett, Hayden, & Martin, 2001). Research also suggests that mental health problems in children persist into and/or recur in adulthood (Costello, Egger, & Angold, 2005; Kim-Cohen et al., 2003) further adding to the global burden of mental health problems (Dockrell, Peacey, & Lunt, 2002; WHO, 2001).
Over the last ten years initiatives have been introduced to support schools to meet the needs of children with mental health problems more effectively. In the UK these include SEAL (Social and Emotional Aspects of Learning), Connexions, Every Child Matters, Education Improvement Partnerships, and the concept of ‘managed moves’ (Vincent, Harris, Thomson, & Toalster, 2007). In the USA, Early Child Mental Health Consultations have been introduced in association with Head Start to improve the mental health readiness of children preparing to attend school (Perry, Holland, Darling-kuria, & Nadiv, 2011). In New Zealand a ‘Positive Behaviour for Learning’ programme has been introduced to support the use of evidence-based interventions to help children identified with severe and persistent conduct problems (Beauchamp, 2012). In many countries there are also psychological services provided within or at least linked to the school such as counselors, psychologists, and special educational needs officers or coordinators. These practitioners have a number of roles that range from direct work with young people, support and advice to staff and parents, and the coordination of input from external services. Despite this array of interventions, only a minority (one-sixth to one-third) of children with an impairing psychological problem are in contact with mental health services whether provided from inside the school or from external services (Costello et al., 2005; Ford, Hamilton, Meltzer, & Goodman, 2007; Weist, Rubin, Moore, Adelsheim, & Wrobel, 2007). Typically, being excluded (or expelled or suspended) from school compounds a child’s difficulties and in turn reduces the likelihood that they will access mental health support or specialist educational services (Hallam & Castle, 2000; Vincent et al., 2007).
New approaches to understanding the mental health of children who have been excluded/suspended are needed but are presently overlooked. The current study set out systematically to review studies of psychiatric disorder and/or impairing psychopathology among children who are and are not excluded from school. Such children might benefit from intervention and support from education and mental health services and the implications for these services are discussed.
Methods
Search strategy
The search strategy was developed in consultation with researchers, clinicians and education professionals. The master strategy used a combination of database thesaurus headings and free-text terms to capture the three concepts of school children, poor mental health, and exclusion from school. Appendix A (see supplemental materials) provides a detailed description of the master search strategy.
Inclusion and exclusion criteria
This review aimed to identify comparative studies of school exclusion, suspension or expulsion and psychiatric disorder or impairing psychopathology in children aged 4- to 18-years-old in mainstream education. For inclusion, studies had to use a validated measure (a measure tested for its reliability and content validity for the outcome of interest) to assess psychiatric disorder or impairing psychopathology. Studies whose primary focus was the evaluation or implementation of an intervention, where children had comorbid conditions or disabilities or where only children at ‘high risk’ of being excluded, suspended or expelled, or where not enough information was provided to be able to allow quality appraisal of the study (for example conference abstracts) were excluded.
Data extraction
Titles, abstracts and full texts were screened for relevance using the inclusion/exclusion criteria independently by two reviewers (RM and RW), with arbitration of disagreements by a third reviewer (TF). Information on the content and quality of the studies was gained through a systematic data extraction process using a standardized, piloted data extraction form (see Appendix B in the supplemental materials) based on guidelines from the Centre for Reviews and Dissemination (Centre for Reviews and Dissemination, 2009). These included information on study design, data obtained, type of analysis, tools used to assess psychopathology/emotional and behavioral difficulties, and the presence of baseline details.
Analysis
Data obtained from the included studies were used to calculate the odds of having mental health problems in children excluded, suspended or expelled from school relative to those who are not (odds ratio, OR). Authors were contacted to supply additional data, if the raw data were not available in the article. Heterogeneity in study population and mental health outcomes meant that pooling of estimates across studies was not appropriate.
Results
The database searches retrieved a total of 5,043 potentially relevant articles, and after de-duplication 3,712 abstracts and titles were screened. A total of 78 full texts were retrieved for further study to determine inclusion or exclusion in the review (see Figure 1 in the Supplemental Materials). Five studies met the inclusion criteria (Bauermeister et al., 2007; Ezpeleta, 2000; Meyer et al., 1993; Ripley & Yuill, 2005; Rushton, Forcier, & Schectman, 2002). The remaining studies were excluded for reasons of study design (no comparative data, N = 24), setting inappropriateness (not a mainstream school, N = 1), population (not looking at children excluded from school, N = 28), outcomes measured (not reporting psychiatric disorder or impairing psychopathology, N = 19), or lack of data (N = 1) (see Appendix C in supplemental materials for list of excluded studies).
Characteristics of the five included studies are summarized in Table 1 (see supplemental materials). There were some differences between the studies in terms of the sample sizes, reported study design, the age of the sample population (ranging from 4- to 17-years-old), and reported outcomes of interest (ADHD, depression and other functional impairments). There was also disparity in study setting which may reflect differences in educational systems and policies. The studies include a total of 16,294 children aged 4- to 17-years-old (range 38–13,568 children) and one study included only boys (Ripley & Yuill, 2005). Most studies reported suspensions (fixed term exclusions); one study reported on permanent exclusion (expulsions) (Ripley & Yuill, 2005).
The quality assessment of the five studies is summarized in Table 2 (see Supplemental Materials). All studies reported eligibility criteria poorly and only two studies (Ezpeleta, 2000; Rushton et al., 2002) ensured that the outcome assessors were blinded to whether or not the children were excluded, suspended or expelled. Although valid and reliable measures were used to assess the relevant outcomes in all the studies, it is notable that the study by Meyer and colleagues (Meyer et al., 1993) used DSM III (Diagnostic and Statistical Manual of Mental Disorders version 3) when DSM III-R, which used more stringent definitions of disorder, had been available since 1987. The lack of methodological detail limits the interpretation and extrapolation of results from these studies.
In two longitudinal studies (Meyer et al., 1993; Rushton et al., 2002) the methods suggest that the population studied was a group of children with a psychiatric disorder or impaired psychopathology with their exclusion status as the outcome. However, the results were reported in the reverse format so that the population described were children who had been excluded, suspended or expelled and their psychiatric disorder or impaired psychology was the reported outcome; the current review was focused on the latter approach. The reporting of these longitudinal studies was particularly challenging to understand whilst the cross-sectional surveys (Bauermeister et al., 2007; Ezpeleta, 2000) could legitimately (and easily) portray their results to examine either question.
The results for all the studies can be seen in Table 3 (see supplemental materials), which reports the odds ratios (OR) for each study where possible. The psychiatric disorders and impairing psychology identified in the included studies were depression, ADHD and functional impairment or social, emotional, behavioral and mental health functioning as measured by the Children’s Global Assessment Scale (CGAS; (Shaffer et al., 1983). The relative odds (OR) of having some kind of psychiatric disorder or psychopathological impairment for children excluded, suspended, or expelled from school ranged from 1.13 (95% CI: 0.55, 2.33) to 31.9 (5.1, 199) compared to children who had not been excluded, suspended or expelled from school.
Two studies (Meyer et al., 1993; Rushton et al., 2002) compared the odds of reporting severe depression symptoms in children aged 11- to 17-years-old who had and those who had not been suspended from school. Both studies report statistically significant associations between being suspended from school and persistent severe depressive symptoms (OR = 4.92, 95% CI: 2.11, 11.5) (Meyer et al., 1993) and (OR = 1.9, 95% CI: 1.3, 2.7) (Rushton et al., 2002) respectively.
Bauermeister et al. (2007) conducted a cross-sectional survey which reported the OR for presenting with ADHD between children aged 4- to 17-years-old who had and those who had not been suspended/expelled from school. They reported a crude OR of 2.46 (95% CI: 1.31, 4.63), but when this was adjusted for confounders that are also linked to the prevalence of ADHD this decreased to 1.13 (95% CI: 0.55, 2.33) with little evidence of a true association.
The final two studies (Ezpeleta, 2000; Ripley & Yuill, 2005) reported the relationships between psychological impairment symptoms in children aged 7- to 17-years-old who had and those who had not been suspended or excluded. These studies were different in their design and in their sample size and population. The study by Ezpeleta (2000) was a cross-sectional study of 172 children and used the CGAS (Shaffer et al., 1983) to measure functional impairment. There was little evidence of a difference in functioning between those who had been suspended once or twice and those who had never been suspended. Those who had been suspended three or more times, however, had a mean functioning score that was 4.9 (95% CI: 0.7 to 9.2) units lower than their counterparts without suspensions.
Ripley and Yuill’s study (Ripley & Yuill, 2005) was a much smaller case-control study, which measured boys’ behavioral and social impairments using the Strengths and Difficulties Questionnaire (SDQ) (Goodman, 1997). This was the only study that included children who had been permanently excluded from school, and also the only one that included only boys. The boys had been permanently excluded from school for verbal and physical aggression, failure to follow rules, in some cases possession of a weapon and in one case for absconding from school. They reported that boys who had been permanently excluded from school were more likely to show symptoms of emotional, or conduct, or hyperactivity problems, or to exhibit poorer prosocial skills or more difficulties with peer relationships than those boys who had not been permanently excluded (ORs ranging from 8–32). Only the association between being permanently excluded from school and presenting with symptoms of emotional disorders was not statistically significant at the 5% level, although the very small sample means the estimates are not very precise as illustrated by the wide confidence intervals.
Discussion
The evidence contained within the few studies identified in this review suggests that there is likely to be an association between being excluded, suspended, or expelled from school and having ADHD or severe depression, and to some extent clinically relevant difficulties with behavior, peer relationships, and prosocial skills. No inference can be drawn about causality from these studies as they all collected information about suspension/exclusion retrospectively and only reported psychological disorder or impairment at the time of the study. The majority were cross-sectional surveys and consequently, we cannot assume that being excluded, suspended, or expelled led to any impairment or that there was an impairment previously (identified or unidentified) that influenced (at least in part) an exclusion, suspension or expulsion.
It is useful to consider the type of impairment within this context to gain insight into the associations presented here. For example, it might be expected that a child with ADHD may be excluded, suspended, or expelled in the future but it would not be expected that such an event could precipitate ADHD. By contrast, the association between depression and exclusion, suspension or expulsion may be bi-directional. It is possible that a depressed child may exhibit behaviors that might lead to exclusion, suspension, or expulsion due to non-attendance and apathy—but similarly the consequences of being excluded, suspended, or expelled may lead in itself to the depression or may exacerbate it.
None of the children in the five included studies had been reported as being in contact with mental health services in relation to their ADHD, depression, or functional impairment and there was no information about any support they might have received in school. This may be due to a lack of data rather than lack of assessment, but research also suggests that a large proportion of children with an impairing psychiatric disorder have no contact with services (Ford et al., 2007; Inkelas, Raghavan, Larson, Kuo, & Ortega, 2007).
Studies suggest that psychiatric disorder places the highest cost burden on educational services, particularly schools, compared to the other public sector services in Britain (Snell et al., 2013; Telford et al., 2013). This notion is supported by Ford and colleagues (2007) who reported that a higher proportion of children with psychiatric disorders have contact with schools and educational services in relation to their mental health than with specialist child and adolescent mental health services. This small but emerging literature suggests that there will be many more children ‘at risk’ of exclusion from school (Parker et al., resubmitted) who are also likely to have similar social, emotional, and/or behavioral symptoms and with whom there is a greater chance to change their course of direction because they are still in school (and therefore able to access educational services albeit not always the most appropriate one). The occurrence and impact of managed moves has little coverage in the research literature, and should become a focus for future research. While schools need to respond to severe behavioral actions and escalating behavior problems, children with special educational needs, and particularly impairing psychopathology, appear to be over-represented among children suspended or excluded from school. It is therefore, important to study ways of managing the breakdown of a school placement that are least detrimental to vulnerable children and to support educational professionals to do this.
This review has a number of strengths including explicit and rigorous methods for quality assessment and data extraction, and recalculation of individual study associations on a metric that allows some degree of cautious comparison between studies. We have deliberately chosen to adopt the rigorous approach to systematic reviews proposed by evidence-based focused organizations like CRD, NICE and the Cochrane collaboration, rather than a more narrative and less systematic approach. This serves to highlight methodological issues among studies, and gaps in the literature as well as providing more robust evidence of an association. The main limitation is that, despite comprehensive and exhaustive searching, we were only able to identify five studies that met the inclusion criteria. We also are aware that there was no independent reporting of the psychiatric disorders or impaired psychopathology. This means the results may have been biased to over- or under-estimate the strength of the association between psychiatric disorders or impaired psychology and exclusion, suspension, or expulsion from school. This limits the interpretation and extrapolation of the results and our ability to draw conclusions on the causality of any association between exclusion and psychiatric disorder or impairing psychopathology.
It could be argued that children who are permanently excluded (i.e. expelled) from school are quite different from those who are suspended based on the types of offences that might normally lead to those events. However, from the description in the study by Ripley and Yuill (2005), some children were permanently excluded on the basis of persistent behavior problems, and others simply in response to single severely challenging actions such as being in possession of a weapon. It is those unresolved behavior problems (perhaps with psychiatric or psychological origins) that are of interest in this review and therefore it is fitting that they are presented together with those children who have been suspended. Furthermore, while the links between psychiatric disorder or impaired psychology and exclusion, suspension, or expulsion from school are likely to be a common concern across countries, there is some need to reflect on the differing administration systems involved in education across them.
This review forms part of an enquiry into school exclusion and mental health (the current review looks at the odds of having impairing psychology if a child is excluded from school and the other looks at the odds of a child being excluded from school when experiencing impairing psychology (Parker et al., resubmitted). We believe that these are the first systematic reviews in this area and should form the basis of further work to improve the developmental trajectory of children with psychological distress who struggle to function within the school setting.
New approaches to discipline and particularly exclusion, suspension and expulsion are being discussed in many countries as it is recognized that exclusion from school often goes against the principle of ensuring all children are engaged in education. In England, a Green paper ‘Support and Aspiration: A New Approach to Special Educational Needs and Disability’ (Department for Education, 2011) currently in consultation, has highlighted the need to help professionals (education, health, and social services) to identify problems in children as they arise and may consider using a ‘robust system of early checks for children’ to reform current processes. In New Zealand, early intervention programmes are being introduced to support children with particular problems as is the case in Australia and the USA (Planty et al., 2009). With this increasing focus on early intervention in health and educational policy and an increased recognition of the importance of early identification in the promotion of positive outcomes, a greater understanding of the mental health support needs of children excluded, suspended, or expelled from school (Allen, 2011) is necessary (see Supplemental Materials for Figure 2 describing a possible pathway). For example, a child with depression may be failing to complete their work or be failing to attend classes, or participate socially in school. If this leads to a child becoming at risk of exclusion, suspension, or expulsion from school, a form of mental health screening or investigation by psychologists or trained educational professionals might identify possible underlying psychiatric or psychological disorders. This system would provide more opportunities for children with problems in school to be identified and supported early on rather than disciplined and excluded from school.
Equally, it will be important for the appropriate support services to be available for every child who is identified. Certainly the potential for impairing psychopathology to go undetected even after exclusion may be resolved by a systematic assessment of the children at risk of exclusion or a managed move (Abdelnoor, 2007; O’Regan, 2009) and more general collaboration between education and mental health services might help identify what the support should look like. Voluntary mental health screening programmes in schools have already begun in some US states under the recommendation of the President’s New Freedom Commission on Mental Health ‘Achieving the Promise: Transforming Mental Health Care in America’ (2003) and are being considered in the UK (Williams, 2013).
The recent inquiry into school exclusions by the UK Children’s Commissioner (Children’s Commissioner, 2012) has suggested that where a child with SEN is at risk of being permanently excluded from school, the school should be required to make a ‘proposal’ for exclusion which should trigger a review of the child’s ‘statement’. This would allow all the professionals who are involved with the child, as well as the child themselves and their carers, to determine the best course of action. Exclusion should only be a last resort and if the appropriate strategies could be found to identify, manage and support children in their school environment there would be benefits for society as well as for schools and the children themselves. For children who find school a struggle, exclusion may unintentionally reinforce the very behaviors that are being punished by providing them an escape route from a situation that they experience as aversive. The need for education and mental health professionals to work together to provide the appropriate support for children at risk of exclusion is of great and immediate concern.
Implications related to the practice of educational/school psychology
Despite indications from national statistics and observational research of higher rates of school exclusion, suspension, and expulsion in those with special educational needs, there is little attempt to manage those children differently with much psychopathological impairment continuing to go unidentified and unsupported. The evidence in this review suggests a potential association between being excluded, suspended, or expelled from school and having psychiatric disorders (such as depression or ADHD) or clinically impairing psychopathology. Below are some key messages related to the practice of educational/school psychology:
There is a paucity of literature looking at the causal relationship between the mental health needs of children excluded from school. Interventions aimed to help this population ideally need to establish this before being implemented. The findings may indicate that there is a particular burden on teachers, education professionals, and schools who are involved with children with educational and mental health needs. Education professionals (with the support of mental health professionals) could play an important role in identifying the mental health needs of children within their school, especially given the difficulty accessing further mental health support services for these children. This may involve a mental health needs assessment, particularly in cases of ongoing behavioral/disciplinary difficulties. Early identification of children with these needs is important so that effective support can be put in place in the classroom to help teachers (and other carers and professionals) manage these pupils. This may signal the need for school psychologists to support teachers with children who are presenting with these difficulties. For example, by providing training on core features so that pupils are identified earlier, on preventative programmes for the children, or on providing additional support in the classroom with strategies to manage these difficulties. A holistic approach to supporting children with additional needs to stay in school is needed. The acceptance and involvement of parents is preferable as this could be a potential barrier for the provision of successful support/interventions. Further support for young people vulnerable to depression may also reduce school exclusion. School absence could be a useful indicator for further assessment of that child’s mental health.
Footnotes
Acknowledgements
The authors thank Brahm Norwich and Alison Bethel of the University of Exeter, for their assistance in developing the literature search strategy and for extremely helpful and insightful comments on the manuscript. We also thank the authors of papers for which we provide a review for their help in providing supplemental information that helped present the results consistently in our review.
Funding
This systematic review was funded by the National Institute of Health Research (NIHR) Collaboration for Leadership in Applied Health Research and Care (CLAHRC) for the South West UK. The views expressed in this publication are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health for the UK.
Author biographies
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
