Mental disorder in federal offenders: A Canadian prevalence study

https://doi.org/10.1016/S0160-2527(01)00071-1Get rights and content

Introduction

Although the “criminalisation of the mentally ill” hypothesis offers a partial explanation for the increase in mental illness among inmates, there is no consensus as to the cause for this disturbing trend. Mentally ill persons may be contained in, or seek access to, the criminal justice system for a variety of reasons, but some researchers hold that this applies primarily to those who commit minor crimes (e.g., Abramson, 1972, Kagan, 1990). The high rates reported from prevalence studies in remanded and jail populations (Bland et al., 1990, Gingell, 1991, Gunn et al., 1978, Guy et al., 1985, Powell et al., 1997, Roesch, 1995; Teplin, 1994) are, therefore, not unexpected, and jails have become, as Teplin (1990) remarked, “a repository of the severely mentally ill.”

The “criminalisation of the mentally ill” hypothesis proceeds from the basic premise that mentally ill persons access health care and social support services through the criminal justice system and, thus, are expected to be overrepresented in transient, short-term remand and jail settings. Such access is facilitated by the celerity with which relatively minor offences, or disturbed public behaviour, result in arrest and incarceration. Higher than community rates of mental affliction are, however, also reported from prison studies (Chiles et al., 1990, Dvoskin & Steadman, 1989, Herman et al., 1991, Hodgins & Cote, 1992, Motiuk & Porporino, 1991; Neighbours et al., 1987). These findings suggest that inadequate social policy and arrest bias alone fail to explain the high prevalence rates of mental affliction in prison populations.

The notion that mental disorder, in general, is related to criminal behaviour is accepted in law, has prompted the development of innovative treatment programs for mentally ill offenders, and underpins the design of specialised risk assessment instruments Webster et al., 1997, Webster et al., 1994. As Monahan and Steadman (1983) succinctly pointed out, “contemporary public policy, no less than legal doctrine, is premised on the assumption that a population exists in which mental disorder and criminal behaviour converge.” The notion of a convergence between mental disorder and criminal behaviour that is nontrivial in nature and inadequately explained in terms of social policy or demographic variables alone, thus, appears to be supported by most research studies. The conceptual cartography of the complex nature and degree of this convergence is incomplete and requires further study.

Powell et al. (1997) pointed out that prevalence studies tend to examine jail and prison populations in isolation, and that researchers have neglected “the ways dynamic variables might interact within the criminal justice and mental health systems, contributing to possible relationships affecting these rates. Factors related to the movement of mentally ill inmates into jails, and thence either back into the community or into the prison system, have not been systematically explored.” The authors postulated that lower rates would be found in rural as compared to urban jail and prison samples, based on lower population rates noted in the Epidemiologic Catchment Area (ECA) study (Robins et al., 1984). Their results confirmed earlier findings involving prison inmates, but were inconsistent with studies of urban jail inmates. The rural prison population had higher rates across diagnostic categories. However, the rural jail sample revealed only slight increases over rates reported in the rural ECA community sample, and significantly lower than the rates reported from other urban jail studies (Bland et al., 1990, Roesch, 1995; Teplin, 1994).

Arboleda-Florez, 1994, Arboleda-Florez, 1998 has commented eloquently on the conceptual difficulties that attend attempts at a clear and unambiguous delineation of the relationship between crime and mental disorder. He also alerted us to the potential for tautologous reasoning in attempts to examine the links between violent behaviour and mental disorder. As an example, he points out the circular reasoning whereby rendering a diagnosis (at least in the DSM-III) of antisocial personality disorder, paraphilia, or kleptomania, ascribes to the individual a set of behaviours that are by definition criminal, thereby increasing the prevalence of criminality among the mentally disordered, in tautologous fashion. Although matters are perhaps not as clear as Arboleda-Florez suggests, his comments raise important questions about what constitutes a “case” in forensic settings.

The plight of mentally ill offenders has received increased attention over the past several decades, with inmate and advocacy groups demanding expanded mental health services in forensic facilities. In Canada, offenders who receive sentences of 2 years or more fall under the jurisdiction of the federal Correctional Services of Canada (CSC), while provincial prisons accommodate offenders serving sentences of less than 2 years. CSC, as part of its mandate, has a strong commitment to the rehabilitation agenda and provides offenders with a variety of correctional programs to address their criminogenic needs, such as mental illness, substance abuse, cognitive distortions, anger, impulsiveness, as well as programs to address crimes of a violent and sexual nature. CSC also provides mental health services through a number of specialised forensic psychiatric hospitals, and although all newly sentenced federal offenders participate in a detailed assessment protocol to identify their specific set of criminogenic factors, no clear diagnostic protocol exists at the reception level for the identification and referral of mentally disordered offenders to psychiatric services.

Reported prevalence rates of mental disorder vary considerably across studies in remand centres, jails, and prisons (Arboleda-Florez, 1994). For example, the reported rates for schizophrenia range from 1% to 8% in prisoners, and from 2% to 75% in remanded populations. Similar variance is notable with respect to affective disorder, organic mental disorders, alcoholism, drug dependency, and antisocial personality disorder. Despite this high variability, researchers have consistently reported that the prevalence rate of major mental disorders among inmates exceeds that of the general population. Researchers Arboleda-Florez, 1994, Coid, 1984 have commented on the effect of divergent and disparate research methodologies on reported prevalence rates, which have ranged widely from 7% (Coid, 1984) to 90% (Bland et al., 1990) (see Table 1). Some of the reasons that could explain this variance include the use of different psychiatric classification systems (DSM-II, III, III-R, IV, ICD, Feighner's research criteria), differential sampling and selection methods, inconsistency in assessor number, qualification, training or interrater reliability, and the utilisation of divergent assessment instruments (file review, unstructured or semistructured interview, DIS, SCID). Methodological differences and uneven reporting of assessor selection and interrater reliability present serious obstacles to a meaningful meta-analysis of the research findings. Arboleda-Florez et al. (1995) point out that some studies have used large numbers of raters from various professional backgrounds, but fail to address or report the issue of interrater reliability. Sampling methods similarly vary across studies, with relatively few researchers using a random sampling method.

The Diagnostic Interview Schedule (DIS) (Robins, Helzer, & Croughan, 1981) has acceptable reliability (between lay-raters) and validity for the diagnosis of a small set of major mental disorders, validity being assessed against diagnoses completed by experienced clinicians. Drawbacks of the DIS include poor sensitivity and a restricted range of applicability in terms of types of mental disorders diagnosable. In addition, the DIS is based on DSM-III, and not the current DSM-IV nosology, thus, limiting the relevance of DIS data in terms of clinical validity and mental health service planning. Additional methodological difficulties relate to the low sensitivity of the DIS. Wittchen, Semler, and von Zerssen (1985) pointed out that severely psychotic patients may not recall or tend to deny psychotic experiences, which may lead to an underestimation of the true lifetime prevalence rate of schizophrenia. On the other hand, previous psychiatric treatment may result in an overestimation of mental illness because memorable symptoms are more readily recalled. On specific disorders, cases that are near the threshold of standard DSM diagnostic definition lead to most disagreement (Helzer et al., 1985), and the DIS seems to agree best with clinical diagnosis in cases where no diagnosis is made (Folstein et al., 1985).

In special settings like prisons and forensic psychiatric hospitals, a semistructured interview such as the Structured Clinical Interview for DSM-IV (SCID) is the best available diagnostic tool for the assessment of major Axis I disorders, (Arboleda-Florez, 1994).

The SCID is considered the optimal structured diagnostic interview and it is intended solely for use by qualified clinicians. It allows the clinician to challenge any response inconsistency, and to paraphrase questions to fit the subject's understanding and ask additional questions based on clinical judgement (Spitzer et al., 1992). Its use in correctional settings has shown less bias than other instruments in specific and overall prevalence for all mental disorders (Arboleda-Florez, 1994). The SCID has been tested for reliability and validity, yielding interrater coefficients of .61 for current and .68 for lifetime diagnoses in-patient samples (Williams et al., 1982). The kappa values are considerably higher for the SCID compared to other instruments used to diagnose mental disorders.

Few studies have documented the prevalence of mental disorders in Canadian prisoners. Allodi, Kedward, and Robertson (1977) collected data from the annual reports of the Ministries of Health and Correctional Services and from files of over 3000 offenders that underwent psychiatric assessment between 1969 and 1973 in Ontario correctional facilities. Their findings supported the hypothesis that the decrease in the number of beds in mental hospitals was associated with an increased incarceration of psychiatric patients. The highest proportion of mentally disordered offenders was identified with psychotic disorder (47.6%), and the most common nonpsychotic disorder was alcohol abuse/dependence (16.3%). About two-thirds of all inmates had a history of psychiatric treatment. The data source for this study is, however, unreliable due to the diagnostic protocols used to identify mental disorders.

Two studies Hodgins & Cote, 1992, Motiuk & Porporino, 1991 have examined the prevalence of mental disorder in the Canadian federal prison population. Both studies underscored previous findings about the prevalence, nature, and severity of mental disorders in male prison populations. Both studies, however, are subject to some of the criticisms raised by Coid and Arboleda-Florez. For example, both utilised the DIS for diagnosis, both used large numbers of assessors of divergent professional background, one study (Motiuk & Porporino, 1991) used a modified random selection technique, and only Hodgins and Cote (1992) provided interrater reliability information. Apart from those studies that have concentrated on remanded populations, no other published study has, to our knowledge, examined the prevalence of mental disorder upon entry to the federal prison system. The present literature on prison populations is, therefore, unable to separate the prevalence rates of mental disorders upon admission to prison, from those that arise de novo during incarceration.

Motiuk and Porporino (1991) assessed the prevalence, nature, and severity of mental illness among a sample of 2000 randomly selected Canadian federal male offenders. Using the DIS, the results revealed that the lifetime risk of at least one psychotic episode (e.g., schizophrenia, schizophreniform psychosis) was 10.4%. The lifetime prevalence of depressive disorders (e.g., major depression, dysthymia, bipolar) was 21.5%, anxiety disorders (e.g., panic, generalised anxiety, phobia, agoraphobia, somatisation disorder), 44.1%, psychosexual disorders (e.g., sexual dysfunction, transsexualism), 56.9%, antisocial personality disorder, 40.9%, substance abuse/dependence, and 47.2%, alcohol use/dependence disorders. Interestingly, however, there were fewer mentally disordered offenders in minimum-security institutions. Moreover, they were released less frequently on full parole, and once released, were more likely to be readmitted for technical breaches of release conditions. The findings suggest that mentally disordered offenders are more likely than nonmentally ill offenders to remain incarcerated because of a lack of mental health services to meet their special needs.

Hodgins and Cote (1993) suggested that severely disordered offenders are characterised by the same risk-predictive factors as other offenders. Porporino and Motiuk (1995), following their earlier study, found that severely mentally disordered offenders (who met stringent diagnostic criteria for manic episode, schizophrenia, or schizophreniform disorder) did not differ from other offenders in terms of the number and type of previous and current criminal convictions.

The most recent Canadian prevalence study of mental illness in a forensic setting was by Arboleda-Florez (1994). Over 1000 people admitted to the Calgary Remand Centre, Alberta, were interviewed by psychiatrists using the SCID and the Hare Psychopathy Checklist-Revised to screen for 1-month and lifetime prevalence of mental illness. One principal diagnosis was made in 60.7%. Of these, 92.2% were for Axis I and 8.8% for Axis II conditions. Fifty-six percent of males and 49.5% of females were diagnosed with a mental disorder, alcohol-dependency being the most frequent in both subsamples (31.7% and 26.1%, respectively). Unexpectedly, only 1.2% of males and no females were diagnosed with schizophrenia. While drug abuse/dependency was most frequent among young people, schizophrenia and depression were more characteristic of the oldest group. Alcohol abuse/dependence was evenly spread across all age groups. Only 5.5% of males received an Axis II diagnosis, the most frequent being personality disorder not otherwise specified and antisocial personality disorder. Alcohol abuse/dependence disorder was reported as elevated above community rates, while disorders such as schizophrenia and affective disorders seemed similar to those found in general population (Arboleda-Florez, 1994). Mental illness, coexisting with the principal diagnosis (comorbidity), was present only in 7.5% within Axis I and 15.6% for Axes I and II combined. Among males, the most common combination within Axis I was psychotic disorder with substance abuse disorder and somatoform disorder with substance abuse disorder. The prevalence of mental illness was not significantly correlated with type of crime or the total number of instant charges.

The main objective of the present study is to examine the prevalence rates of mental disorders among newly sentenced, male federal prisoners admitted to the intake assessment unit of one of the federal correctional regions in Canada. This study addresses the methodological concerns raised by Coid and Arboleda-Florez and, to our knowledge, is the first prevalence study of this offender population to use random selection sampling, a small number of trained and experienced forensic professional assessors with good interrater reliability, and an optimal assessment diagnostic instrument (SCID).

Section snippets

Subjects

During the study period (February 24, 1999–September 23, 1999), 267 male offenders, newly sentenced to a term of federal incarceration, were admitted to the Regional Reception and Assessment Centre (RRAC) in Abbotsford, British Columbia, for a standardised CSC intake assessment. The assessment focuses on the identification of criminogenic factors, such as educational deficits, alcohol and substance abuse, procriminal beliefs, and sexual deviance, deemed central to their offending behaviour.

Results

The current (1-month prevalence) and lifetime prevalence rates for any DSM-IV Axis 1 disorder are presented in Table 2. The results indicate that mental disorder is highly prevalent in this population, with 84.2% having at least one 1-month or lifetime DSM-IV Axis I diagnosis. A current mental disorder was found in 31.7%, while 15.8% had no diagnosis. When substance use disorders, which was the diagnosis most frequently endorsed, is excluded, the total prevalence rate was 43.1%

Table 3 compares

Discussion

The forensic literature on the prevalence of mental disorder suggests high rates in all diagnostic categories for incarcerated individuals compared to the general population. The results of the present study confirm these findings. The high overall prevalence rate of mental disorder (84.2%), as well as significant comorbidity rates, indicate that mental health needs constitute a significant challenge to correctional facilities, particularly in the provision of integrated rehabilitative

Acknowledgements

The authors wish to thank Drs. C. Hunter, N. Hodelet, J. Ronsley, and M. McBride, as well as Ms. D. Mawson, for their assistance in this study.

First page preview

First page preview
Click to open first page preview

References (31)

  • J.A Dvoskin et al.

    Chronically mentally ill inmates. The wrong concept for the right services

    International Journal of Law and Psychiatry

    (1989)
  • F.J Porporino et al.

    The prison careers of mentally disordered offenders

    International Journal of Law and Psychiatry

    (1995)
  • M.F Abramson

    The criminalization of mentally disordered behaviour. Possible side-effect of a new mental health law

    Hospital and Community Psychiatry

    (1972)
  • R Allodi et al.

    Insane but guilty: psychiatric patients in jail

    Canada's Mental Health

    (1977)
  • Arboleda-Florez, J. (1994). An epidemiological study of mental illness in a remanded population and the relationship...
  • J Arboleda-Florez

    Mental illness and violence: an epidemiological appraisal of the evidence

    Canadian Journal of Psychiatry

    (1998)
  • J Arboleda-Florez et al.

    An epidemiological study of mental illness in a remanded population

    International Medical Journal

    (1995)
  • R.C Bland et al.

    Prevalence of psychiatric disorders and suicide attempts in a prison population

    Canadian Journal of Psychiatry

    (1990)
  • R.C Bland et al.

    Lifetime prevalence of psychiatric disorders in Edmonton

    Acta Psychiatrica Scandinavica

    (1988)
  • J.A Chiles et al.

    Substance abuse and psychiatric disorder in prison inmates

    Hospital and Community Psychiatry

    (1990)
  • J Coid

    How many psychiatric patients in prison?

    British Journal of Psychiatry

    (1984)
  • M.F Folstein

    Brief report on the clinical reappraisal of the DIS carried out at the Johns Hopkins Site of the Epidemiological Catchment Area Program of the NIMH

    Psychological Medicine

    (1985)
  • C.R Gingell

    The criminalisation of the mentally ill: an examination of the hypothesis

    (1991)
  • J Gunn et al.

    Psychiatric aspects of imprisonment

    (1978)
  • E Guy et al.

    Mental health of prisoners in an urban jail

    Criminal Justice and Behaviour

    (1985)
  • Cited by (90)

    • The prevalence of comorbid serious mental illnesses and substance use disorders in prison populations: a systematic review and meta-analysis

      2022, The Lancet Public Health
      Citation Excerpt :

      Between-sample heterogeneity was high for estimates of current and lifetime prevalence of comorbid non-affective psychosis and substance use disorders (I2 >80%), which was not explained by any tested factor in our meta-regression after adjustment for multiple comparisons (appendix p 13). Co-occurring major depression and any type of substance use disorder was reported in 42 studies.19,23–25,27–41,43–46,48–51,55,57–60,62–71 The prevalence of current co-occurring major depression and substance use disorders was 9·1% (95% CI 5·6–13·3; n=11 133; k=27), representing 1105 (51·6%) of 2143 people with major depression and 1105 (23·3%) of 4733 with substance use disorders (figure 3; appendix p 12).

    • Mental Health: Crime and Mental Health

      2015, Encyclopedia of Forensic and Legal Medicine: Second Edition
    View all citing articles on Scopus
    View full text