Abstract
The National Inquiry into Human Rights of People with Mental Illness [3] noted that in Australia 30–50% of children in correctional facilities could be expected to suffer from a mental health problem. This is up to seven times higher than in the normal population. Burdekin [3] suggested that behaviours arising from mental health problems were often misinterpreted as delinquency, and that detention centres themselves were frequently the cause of further mental illness. He suggested that of all criminals, delinquent young people were most in need of treatment as early intervention could prevent recidivism and promote rehabilitation.
In 1980, a group of US psychiatric researchers [4] interviewed 120 of 150 13–15 years olds consecutively admitted to a detention centre, using a structured diagnostic interview, with a specific focus on depression. They found that 23% met the DSM-III-R [5] criteria for a major affective disorder, compared with a 4% prevalence rate for depression in adults from the general population at the time.
In 1987, US mental health professionals [6] administered the Child Behaviour Checklist (CBCL) [7]; to 541 12–17-year-old offenders on admission to a detention centre. The results were compared with CBCL results obtained from 250 young people referred to a mental health clinic. The CBCL does not provide diagnoses; rather, it classifies young people as suffering from externalizing (e.g. hyperactivity, aggression, delinquency) or internalizing (e.g. depression, obsessions, somatic complaints, schizoid) problems. The young offenders had the same level of internalizing, or emotional, problems as the mental health clinic referrals, which was about double that of the general community.
One of the few Australian studies of young offenders’ mental health problems was conducted by Kosky et al. [8] at the Adelaide Youth Remand Centre. A two-part study commenced by having 78 of 82 adolescents consecutively admitted to the centre complete the Youth Self Report (YSR) [9]. Of these, 35% said they were usually unhappy/sad/depressed (compared to 7% in the community) and 17% said they were often anxious (compared to 4% in the community). Some 22% said they suffered from frequent suicidal thoughts, while only 3% of the community sample did. Again, young offenders’ internalizing scores were equivalent to those of young people referred to the mental health clinic, and nearly twice those of young people in the community. Thus, even if delinquency and antisocial behaviours were disregarded, offenders still had double the incidence of mental health problems of the community sample. Kosky and Sawyer [10] obtained similar results in the second study, where 32 participants of the original sample were followed up after their release.
More recent investigations have focused on trying to obtain prevalence rates of mental health diagnoses from the DSM-IV, using semistructured and structured clinical interviews and the Millon Adolescent Clinical Inventory (MACI) [11]. Diagnostic categories are often easier for clinicians to conceptualize than dimensional scales, such as those derived from the CBCL, and there is also a substantial body of clinical research in this area that uses the DSM classificatory system. The only Australian study yielding diagnostic categories was carried out by Richards [12], using a semistructured clinical interview of 100 detained adolescents consecutively referred to a psychiatric outreach service. Of these, 25% of the young offenders were diagnosed as suffering from a mood disorder and 4% from PTSD, a smaller proportion than in other studies, as these participants had already been preselected through referral rather than being screened on arrival at the centre. A study in a Canadian youth detention centre [13] compared 49 current residents who had agreed to participate with 49 adolescents from the community using structured clinical interviews. In this sample, 69% of offenders had at least one internalizing, or emotional disorder, such as depression or anxiety. In addition, 25% suffered from PTSD and 31% from clinical depression, compared to 0% and 4.1%, respectively, for the community sample. Three recent larger scale studies have been carried out in the US; one involving diagnostic interviews with 1500 youths entering a detention centre [14], one using the MACI with 800 youths in detention centres in Massachusetts [15], and one using the MACI with 300 young offenders in Ohio [16]. The overall result of these studies was that the prevalence rates for mood disorders were 3–4 times higher for young offenders than for young people in the general population, but there was no difference for anxiety disorders.
Over the past few years there has been an increasing amount of information available about the rates of mental health problems in young people in Australia and New Zealand. The West Australian Child Health Survey examined the rates of mental health problems in young people aged 4–16 years [17, 18]. The study used the CBCL to explore eight common child mental health problems and found that the overall rate of mental health problems in children aged 12–16 was 21%. In this age range the study found that 9% experienced problems with delinquency, 8% had attentional problems, 5% suffered from anxiety/depression and 6% had somatic complaints.
The National Surveys of Mental Health and Wellbeing in 1997 and 2000 assessed 13 500 and 4500 young people in the community, respectively, using the CBCL and the Diagnostic Interview Schedule for Children [19, 20]. This study found that 3% of the population had conduct disorder, 6% had ADHD, 4% had a mood disorder, and 3% had an anxiety disorder (excluding PTSD).
The Christchurch Health and Development Study [21] is a longitudinal study which has tracked the rates of mental health problems in a cohort of 1265 children over 21 years since 1977. At age 15 for the whole cohort the study identified that 22% of the young people had experienced at least one serious mental health problem, 13% experienced an anxiety disorder, 6% had a mood disorder, and 5% had a conduct disorder. At age 18 for the whole cohort 42% of the young people had at least one mental health problem, 17% had an anxiety disorder, 18% had a mood disorder, and 5% had conduct disorder.
The Australian and New Zealand population research that is currently available differs significantly in methodology and in choice of instruments used to operationalize the concepts of ‘mental health problem’ and ‘diagnosis’. Nonetheless, there seems to be a generally robust finding that there is about a 20% rate of significant mental health problems in the general population of young people in their adolescent years. The review of studies conducted on adolescents in custody suggests that between 23% and 40% would suffer from a mood disorder. Findings concerning the prevalence of anxiety disorders give a much broader range and vary from 3% to 40%.
The aim of this study was to obtain prevalence rates of clinical disorders for detained young offenders in Australia, in a rural setting. The scale chosen was the Adolescent Psychopathology Scale (APS) [2] partly because it yielded a measure of ADHD that is frequently found in young offenders. The APS additionally provides scales that are based on the DSM-IV criteria for 19 other disorders. This study is the first to yield standardized data on the prevalence of diagnostic categories among young offenders in Australia, and is also the first using the APS in a forensic setting.
Method
Measures
The Adolescent Psychopathology Scale
The APS is a 346 item self-report tool for adolescents aged 12–18, providing scores on 4 validity scales and 20 clinical disorders from the DSM-IV, listed in Table 1.
Adolescent Psychopathology Scale clinical disorder scales
The APS was developed in the US from an item pool of DSM-IV symptom specifications that were administered to adolescents in school and clinical settings. Item selection was based on assessment of content validity and statistical item analysis. The initial standardization sample was 1827 school based adolescents and 506 adolescents from mental health clinics.
The internal consistency of the APS clinical disorder scales was high for both the standardization and clinical samples with median alpha coefficients of 0.85 and 0.87, respectively. Item-total correlations on the clinical disorder scales were moderately high for the school based sample and the adolescents from mental health clinics with median item-total correlations ranging between 0.41 and 0.61, and 0.40 and 0.65. The intercorrelations between the clinical disorder scales ranged between 0.30 and 0.69 which is to be expected when disorders have overlapping phenomenologies or common comorbidities.
Criterion related validity coefficients with the Minnesota Multiphasic Personality Inventory [22] were moderately strong overall and high (r = 0.77–0.82) and also with the Reynolds Adolescent Depression Scale [23] and the Beck Depression Inventory [24]. Highly significant differences were also found in a comparison between the clinical group and the non-clinical group on both internalizing and externalizing scales.
Consideration was given to using structured interview schedules including the Diagnostic Interview for Children and Adolescents, Revised (DICA-R) [25], and the Diagnostic Interview Schedule for Children, Version 2.3 (DISC-2.3) [19], which was used in the National Survey [20]. However, these schedules can take up to 2.5 hours to administer, particularly with clients with complex presentations such as the detained young offenders used in this study. One of the key features of conduct disordered youths and young offenders is reported to be a short attention span and impatience [26] and it was thought that a lengthy diagnostic interview would put many of the clients off from participating in the study. The APS has adequate reliability and validity and, though it is a newer measure, it was judged to have a better return for cost than the interview schedules.
Secure Care Psychosocial Screening Tool
The Secure Care Psychosocial Screening Tool (SECAPS) [27] is a multigate assessment tool developed in South Australia specifically as a screening instrument for young people on admission to juvenile detention facilities. It has been designed to allow for a rapid assessment of basic literacy and numeracy. It also allows for an estimation of intelligence using the Coloured Ravens Progressive Matrices [28]. In addition the SECAPS screens for drug and alcohol use, social dysfunction, school attendance, and provides some limited information on emotional problems.
Procedure
All new admissions to the Ashley Youth Detention Centre, Tasmania, Australia, who were to remain at the centre for at least 2 weeks, were asked to participate. Clients were told that participation and follow-up was voluntary and confidential. Questionnaires were administered in a structured interview format, usually over several sessions, because young people in detention often have literacy problems, short attention spans and limited patience. Assessments were followed up with a brief report on results and recommendations. Copies were given to the participants and to other stakeholders, with the young person's consent. Weekly clinical meetings were held with relevant Ashley workers to discuss recommendations, referrals arising from them, and other case management options.
Results
Table 2 shows selected ability and demographic data for the 111 Ashley residents screened by the SECAPS between May 2000 and June 2001.
Demographic data of the detained adolescent population (n = 111)
The data indicate that these young people were more often than not subject to considerable disadvantage in their home lives. Only 30% had parents who were living together, 42% had a family member with a drug or alcohol problem, and 20% lived either by themselves or were homeless. A large proportion (30%) of these young people entered the criminal justice system from a very early age. Many had reading problems with an average chronological age of 15.7 and an average reading age of 10.5. Nearly 22% had a significant degree of intellectual impairment and a further 26% had borderline intellectual functioning.
During the period of this study, May 2000 to June 2001, 53 (62% of all admissions who were residents for at least 2 weeks) agreed to participate. Of the 53 assessments carried out 3 were excluded from the analyses because they returned invalid profiles. Eighty-six per cent (n = 43) of the participants were male, and 14% (n = 7) were female. The young offenders were aged between 12 and 18 years, with an average age of 15.7 years.
The APS results were recorded as either positive or negative for meeting the criteria for specific clinical disorders. Several of the possible disorders were omitted from the analyses because too few of the young people achieved the criteria to make their inclusion meaningful (see Table 1). In addition, separation anxiety disorder was omitted because respondents had recently been separated from their familiar environment and placed in a detention centre and distress connected with this experience was not considered to be ‘inappropriate or excessive’ as required by the DSM-IV [1]. Anxiety disorder without PTSD includes panic disorder, obsessive–compulsive disorder, generalized anxiety disorder, and social phobia in the absence of a history of trauma. The discriminant classification of disorders was carried out according to the diagnostic guidelines in the DSM-IV. For instance, a diagnosis of conduct disorder precludes the possibility of a diagnosis of oppositional defiant disorder and takes precedence over that diagnosis. Therefore, if a participant scored positive for both conduct disorder and oppositional defiant disorder the latter classification was omitted.
The only disorder that was classified using other information sources additional to the APS was conduct disorder. Because many adolescents wished to use the outcome of their assessments in court, they consistently underrepresented their criminal behaviour. Additional information was sought from youth justice and court reports, and a classification made on the basis of combined information. Young offenders are given consistent and repeated support to prevent them from re-offending, and only severe, repeated offenders are remanded in custody. Thus, it is not surprising that all participants, with one exception, could be classified as having conduct disorder.
Results of the APS assessments show that 36% were classified with PTSD, 32% with an anxiety disorder, excluding PTSD, and 30% with major depressive disorder [see Tables 1, 2 and 3]. A further 46% scored positive for ADHD, 22% for adjustment disorder, 22% for somatization disorder, and 16% for dysthymia.
It was common that the young people in this study suffered from more than one mental health problem. Therefore, the sample was further categorized as having only one problem (no comorbidity), two problems (1 comorbidity), three problems (2 comorbidities), or four and more problems (3 or more comorbidities). Only 14% of the sample had conduct disorder only, 16% had conduct disorder and one other mental health problem, 18% had two additional mental health problems and 52% had three or more mental health problems in addition to conduct disorder.
To answer the question of whether the presence or absence of a disorder was influenced by the level of participants’ non-verbal IQ, a series of χ2 tests were performed (see Table 4). Major depressive disorder was the only disorder occurring significantly more frequently when IQ levels were in the borderline range. Further interesting results were obtained from χ2 analysis investigating whether presence or absence of ADHD was related to presence or absence of other disorders (see Table 5). Posttraumatic stress disorder, major depressive disorder, and somatization disorder were present significantly more often in participants who also suffered from ADHD. Females were not more likely than males to suffer from an internalizing disorder in general, but there was a higher incidence of mood disorder and anxiety disorder excluding PTSD among females compared to males (see Table 6).
Prevalence rates of disorder based on the Adolescent Psychopathology Scale for detained adolescents
Diagnosis by IQ
Diagnosis by presence or absence of ADHD
Diagnosis by gender
Discussion
Prevalence rates of disorders for detained adolescents were found to be as high, if not higher than expected; 46% suffering from a mood disorder and 32% suffering from an anxiety disorder, excluding PTSD. It is remarkable that 36% suffered from PTSD but this may be explained with the fact that a large proportion of conduct disordered young people have histories of abuse and neglect [26].
This study is the first to use the APS for the assessment of adolescents in custody. The APS was found to be a useful tool, posing simple, straightforward questions and covering a useful spread of DSM-IV diagnostic categories. This study is also the first in Australia using a standardized, psychometric tool with detained adolescents to obtain diagnostic categories. Kosky [8] used the YSR, which does not provide DSM-IV diagnostic categories, and Richards [12] used a non-standardized semistructured clinical interview. Thus, these results provide a useful addition to the currently limited Australian data on the psychological morbidity of young offenders.
Compared with the Australian population studies, the rate of internalizing mental health problems of detained adolescents was some five times higher than these disorders in the general community. These comparisons are similar to those obtained by Kosky [8] and in other studies.
The data show quite clearly that the young people in detention were significantly disadvantaged. Not surprisingly, these young people have major deficits in educational attainment and basic literacy. It is disturbing to note that there was an average delay in reading attainment of 5 years and that 48% of the sample had intellectual levels that were borderline and below in intellectual impairment. It is difficult to know whether the educational difficulties are part of the cause of the young offenders’ behavioural problems or whether they are a natural by-product of a young person pursuing a behaviourally disturbed lifestyle. Nonetheless, the overrepresentation of young people with psychoeducational difficulties in this group suggests that early intervention efforts should focus on targeting children with learning disorders and behavioural problems early on in their educational setting. This data also suggest that young people in detention need to be given enriched and supportive opportunities to address their learning deficits since it is likely that their continued experience of inadequate skills will lock them into the cycle of delinquency and crime.
The results of comparisons involving non-verbal IQ levels and presence and absence of ADHD are interesting and would benefit from further investigation. It appears that young offenders with higher IQs are less likely to become depressed, possibly because their greater intelligence has allowed them to develop more effective coping skills. The relationship between ADHD and PTSD, major depressive disorder and somatization disorder could represent the effects of consistent underachievement, negative feedback, and deficient coping strategies.
An obvious limitation of this study is that a community control group could not be obtained. This was due to the lengthy administration time of the APS. Obtaining Australian norms for the APS would be an advantage in its use in Australian studies. A definitive judgement of diagnoses cannot be made as the APS only provides suggestions of possible diagnoses. A more in-depth study should follow-up such suggestions with a clinical interview. The relatively low participation rate (62%) requires this data to be treated with some caution. It is possible that those young people with emotional distress are more likely to seek help and this would lead to an overestimation of the prevalence of mental disorders within the total sample.
Nevertheless, this study provides evidence that Tasmania is no exception to the rule that adolescents in custody have a high rate of mental health problems. In his review of adolescents in custody, Kosky [29] concluded that all detained adolescents should be assessed for mental health problems as they present with equivalent levels of mental health problems as young people referred to mental health services in the community.
Detention centres provide circumstances where considerable opportunities for individual therapy and group programs could be exploited and there is a clear need for the provision of intensive mental health intervention to this high needs group. However, due to community and political pressure our detention centres are generally still places of imprisonment and punishment, rather than providing opportunities for treatment and rehabilitation. Nearly a decade on, we continue to make the error that Burdekin [3] identified of mistaking ‘sadness’ for ‘badness’ in many of these young people.