Mental Health Diagnoses and Recidivism in Paroled Offenders

Abstract

Although the issue of mental illness among offender populations has received attention in the last number of years, there are a number of issues related to mental illness among such groups that require more study. One such topic relates to the association between mental illness, actuarially assessed risk of recidivism, and observed rates of reoffending. In the present investigation, file information was reviewed to determine the presence of a variety of mental health conditions. Actuarially based risk assessment data were also collected for participants as well as information regarding suspension, new charges, and convictions. A sample of 136 offenders housed in a halfway house operated by Correctional Service of Canada was included in the present investigation. Results indicated very high rates of serious mental illness in this high-risk population. Offenders with borderline personality disorder and attention deficit hyperactivity disorder were significantly more likely to recidivate or be suspended. Suspensions refer to administrative decisions to place an offender in jail due to problematic behaviour (typically involving a breach of his release conditions or new charges/convictions). Offenders with a diagnosis of paraphilic disorder were significantly less likely to recidivate or be suspended. Results are discussed in light of the available literature.

Keywords

mental illness offenders recidivism

The issue of mental illness within correctional environments has been receiving more attention in the literature recently. Although it has been known that the rate of mental illness in prison settings has been increasing in recent decades, the alarming changes in such rates has begun to concern those working with such populations. For example, Adams and Ferrandino (2008) noted that “by all objective measures, correctional facilities in the United States have become the primary mental health institutions in the nation” (p. 913). Lamb, Weinberger, and Gross (2004) described the issues associated with the increased presence of severely mentally ill persons within the criminal justice system as an “urgent” problem. Morrissey, Meyer, and Cuddeback (2007) went further in noting that the available evidence indicates that “jails have clearly supplanted state hospitals as the main revolving door for the most disabled people in the public mental health system” (p.532).

Given that there are increasing numbers of offenders suffering from serious mental illness in the criminal justice system, it is important to determine whether these conditions are associated with risk of recidivism. Clinicians and administrators within such environments are routinely asked to make release and other related decisions regarding offender management based on an assessment of an offender’s risk. Unfortunately, there has been relatively little research related to recidivism among mentally disordered offenders that has controlled for actuarially assessed risk. This is a critical issue as it speaks to whether mental illness (particularly more severe psychiatric presentations) should be considered a criminogenic factor (that is a factor empirically associated with risk of recidivism, see Andrews & Bonta, 2010, for a detailed discussion). If mental illness is found to add to the prediction of recidivism after accounting for actuarially assessed risk, this might be considered important evidence to the effect that mental illness represents more than a “responsivity” issue as defined by Andrews and Bonta (2003, 2010). Given the changing nature of the prison population, it is possible that earlier conclusions regarding mental illness and recidivism (especially when less serious mental health issues such as “personal distress” are included) need to be revisited.

Perhaps one of the reasons that there has been relatively little recent research related to this topic is that Andrews and Bonta (2003, 2010), in their very influential text on correctional psychology, have argued that the available evidence suggests that “personal distress” is not a good indicator of risk. These authors also argue that Axis I disorders are relatively infrequent among offender populations. However, they provide a rather limited review of the pertinent literature.

Skeem and her colleagues (Skeem, Emke-Francis, & Louden, 2006; Skeem & Louden, 2006; Skeem, Manchak, & Peterson, 2011) have argued that individuals with serious and often disabling mental disorders such as the psychoses, bipolar disorder, and major depression are grossly overrepresented in the criminal justice system. Furthermore, they note that nearly three out of four jail detainees with a serious mental illness have a co-occurring substance abuse disorder. These authors also note that such groups of offenders have frequently been found to have their community term suspended or revoked.

Starting with some of the earlier work in this area, Porporino and Motiuk (1995) examined a sample of 36 offenders with psychotic disorders and a matched sample of nonpsychotic offenders, and found that the disordered offenders were given fewer opportunities for early release on full parole. Further, when released, offenders with psychotic disorders were significantly more likely to receive suspensions or revocations without the commission of a new offence when compared with comparison subjects. Such data highlight the problems faced by correctional services throughout North America and other jurisdictions in dealing with offenders with mental health problems.

Constantine et al. (2010) conducted a large-scale evaluation of the patterns of arrest/rearrests among offenders as a function of serious mental health diagnosis over a 4-year period. These authors employed a retrospective design that identified individuals with serious mental illness who were jailed in Pinellas County, Florida, between July 2003 and June 2004. A total of 3,769 participants representing 10% of arrestees within the jurisdiction were included in the study. The pattern of arrests for these offenders was investigated for the period beginning in July 2002 and ending in June 2006.

The authors found that rates of arrests among offenders with serious mental illness were unusually high relative to the rates reported for offenders in general. The mean number of arrests for this sample of mentally ill offenders was 4.6. The sample also had an average of 14.1 Emergency Room (ER) or inpatient contacts over the follow-up period. These data highlight the increased frequency of serious mental illness among offender populations as well as their frequent contact with the mental health and criminal justice systems.

Långström, Sjöstedt, and Grann (2004) studied the presence of psychiatric disorders and their relationship to recidivism in a sample of sexual offenders released from Swedish prisons (N = 1,215). Alcohol use disorder, drug use disorder, personality disorder, psychosis, and any inpatient care were all found to predict sexual offence recidivism.

Butler et al. (2006) investigated the presence of mental disorders in Australian prisoners. The authors noted that the 12-month prevalence rate for any psychiatric illness was 80% in prisoners and 31% in a community sample, with this significant difference remaining even after accounting for demographic and other differences. Interestingly, rather than there simply being a higher prevalence rate of those disorders frequently encountered in forensic populations such as substance abuse (although this diagnosis was also noteworthy in the sample), there was also a strong positive association between being a prisoner and reporting signs of psychosis or posttraumatic stress disorder (PTSD) in the previous 12 months. For example, 7% of prisoners and 0.7% of the community sample were diagnosed as psychotic. This difference corresponded to an odds ratio of 11.7 for having a psychotic disorder among prisoners.

Looking specifically at particular disorders, there is evidence to suggest that some diagnoses are more strongly associated with criminal recidivism than others. Attention deficit hyperactivity disorder (ADHD) has been associated with antisocial behaviour and impairment in other areas of functioning (e.g., social integration, family functioning) in children and adolescents (e.g., Barkley, 2006; Barkley, Murphy, & Fischer, 2008). ADHD has also been observed to be associated with substance abuse disorders (Wilens, Biederman, Abrantes, & Spencer, 1997). Given that ADHD and substance abuse disorders have been associated with negative emotionality (Hull & Slone, 2004; Wolraich et al., 2005), it is possible that the presence of ADHD may further contribute to the difficulties that offenders with substance abuse disorders have in the area of affect regulation.

There are a variety of studies documenting the frequently noted association between substance abuse and recidivism among offender populations. Andrews and Bonta (2003, 2010) listed substance abuse among their “Big 8” risk factors regarding offender risk. These authors provide a comprehensive review of research in this area. There is some preliminary research suggesting that borderline personality disorder (BPD) may be associated with interpersonal physical aggression as well (e.g., Langton, Hogue, Daffern, Mannion, & Howells, 2011). At present, however, these findings should be viewed as tentative requiring replication before any conclusions can be reached.

Although the findings in relation to BPD and recidivism should be viewed as preliminary given the available evidence, there is now a more mature body of research available with reference to the association between psychotic disorders and criminal behaviour. A comprehensive meta-analysis has recently been published (Douglas, Guy, & Hart, 2009) which has demonstrated that psychosis is, in fact, significantly associated with recidivism. The authors observed that psychosis was associated with a 49% to 68% increase in the odds of violence. Perhaps of equal or greater importance was the fact that the effect size associated with psychosis (in relation to the prediction of violence) was similar to the effect sizes more commonly associated with variables associated with criminal behaviour (e.g., use of a weapon, history of nonviolent criminal behaviour).

Our team has produced several studies which have a bearing on paraphilic diagnoses and the association of these disorders and recidivism. Abracen and Looman (2006), examining a high-risk sample of sex offenders treated within the Ontario Region of Correctional Service of Canada (CSC), observed that neither a paraphilic diagnosis nor a diagnosis of a personality disorder alone significantly increased the risk of recidivism among this sample of sex offenders. However, those offenders with a personality disorder and a diagnosis of a paraphilia were at significantly increased risk of recidivism.

In a more recent investigation, Looman and Abracen (2012) again investigated recidivism among a group of high-risk sex offenders assessed/treated at the Regional Treatment Centre (Ontario) Sex Offender Treatment Program (RTCSOTP). In this more comprehensive study, it was found that, after statistically accounting for actuarially assessed risk, only having had a history of psychiatric impairment was found to add significantly to the prediction of recidivism. A variety of psychometric instruments, including the Multiphasic Sex Inventory (MSI), a measure designed to assess a variety of domains related to sexual offending, failed to predict recidivism after accounting for actuarially assessed risk using Cox regression analysis. Official criminal records compiled by the Royal Canadian Mounted Police (RCMP) finger print service were used as the dependent measure in this investigation and the study reported by Abracen and Looman (2006).

Perhaps the most significant study produced to date is a recent meta-analysis by Olver, Stockdale, and Wormith (2011). These authors presented data which demonstrated that those offenders with more serious mental illness (e.g., psychosis, borderline personality) predicted attrition from programs. Those offenders who failed to complete programming were in turn significantly more likely to recidivate. These data suggest that failure to address significant mental health concerns may be related to increased rates of recidivism either directly or through the impact that such conditions have on the client’s ability to complete programming.

The present study investigates the association between diagnoses of specific mental disorders and failure (including recidivism) among a sample of paroled offenders under the jurisdiction of CSC and living in a Community Correctional Centre (CCC) in the Ontario Region of CSC. Data from all offenders housed at the Keele CCC from the period April 2007 to April 2008 were included in the study. Information related to mental health diagnoses provided by licensed mental health professionals, risk assessment data, and officially recorded recidivism were also collected. It was hypothesized that a current diagnosis, or one made within the past 5 years, of a mental health disorder, and particularly more serious mental health conditions, would be related to recidivism even after controlling for actuarially assessed risk. Specifically, it was predicted that ADHD, BPD, antisocial personality disorder (APD), psychosis, alcohol, and drug abuse would all be significantly related to recidivism.

It was also predicted that less-significant forms of mental illness (that might be equivalent to “personal distress”) would not be associated with significant elevations of recidivism (e.g., anxiety).

Method

Subjects

A list of all offenders housed at the Keele CCC in fiscal 2007-2008 was collected by the first author (i.e., April 1, 2007-March 31, 2008). As difficulties with achieving 100% accuracy were observed by Abracen, Axford, and Gileno (2012) when such information was obtained by computer systems operated by CSC, the present study collected data on offenders who were physically housed (as indicated by room assignment sheets) at Keele CCC over the period indicated above. All offenders housed at the Keele CCC over the period noted above were included in the present investigation. The Keele CCC is a halfway house operated by CSC. It functions administratively as a minimum security institution. The Keele CCC is located within the city of Toronto, Ontario, Canada. Although there are no external barriers (e.g., fences, gates) that surround the facility, offenders living at Keele must sign in and out from the Centre. Many of the offenders living at Keele work in the community during the day and return to the facility at night. There are security personnel on-site 24 hr a day. Keele CCC, as the only CCC operated in the Toronto area by CSC, serves the needs of offenders deemed to be at higher risk/higher need. A number of psychologists also work on-site and serve the needs of offenders living at the CCC and other offenders on parole in the greater Toronto area. The Keele CCC has a mandate to house a maximum of 10 offenders at any one time who are classified as sex offenders based on their history of sexual offending. As offenders housed at the Keele CCC were expected to present with both higher rates of mental illness and as high risk, this sample represented an ideal population to investigate the association between mental illness and recidivism.

File information was reviewed regarding background and mental health data for all offenders housed at the Keele CCC in fiscal 2007-2008. The review consisted of obtaining information from computer generated systems operated by CSC as well as several paper files available regarding each offender. Information reviewed included coding reports available on the psychology file (a paper file which contains all psychology reports generated regarding the offender) and a review of all psychiatric reports. The Keele CCC operates a psychiatric clinic 1 day per week and the psychiatric documentation produced by the physician (R.D.) who operates that clinic were reviewed for the purpose of the present investigation. All subjects signed one or more consent forms depending on whether they were seen by psychology staff, psychiatry, or forms related to being housed at the Keele CCC. In all, data were collected on 136 consecutive admissions to the Keele CCC in fiscal 2007-2008.

Diagnoses

Psychiatric history was coded by graduate-level students or registered mental health professionals, each of whom had experience working with offender populations. Data were coded with reference to the presence (coded 1) or absence (coded 0) of the following behaviour or diagnoses. Other than with reference to psychotic behaviours (see below for additional information), only in cases where a specific diagnosis was made was a client coded as having that condition. For example, clients for whom alcohol use was related to their index (i.e., current) and/or past offence(s) were not coded as having an alcohol abuse disorder unless such a diagnosis appeared in the psychiatric/psychology report. As the majority of cases were known to either the first or last authors (J.A. or R.D., respectively), where contradictory information existed on file or where a report indicated that the client “likely” met diagnostic criteria for a condition, these cases were discussed and a consensus opinion was arrived at between J.A. and R.D. or other registered mental health professionals on-site who had recently assessed the client.

Only recent mental health diagnoses (i.e., made within 5 years of the date of data collection) were included. Where multiple diagnoses were available, these diagnoses were coded separately. For example, clients with BPD were coded as having “any” personality disorder and as having BPD (see below for the categories of diagnoses included in the study). As diagnoses from the last five 5 years were included, it was impossible to produce reliability data regarding these conditions. For example, a client may have been diagnosed with clinical depression prior to release to the Keele CCC but no longer met criteria for such a condition when assessed by staff at Keele CCC. All diagnoses were made by registered mental health professionals.

Data were coded with reference to the presence of the following behaviours and diagnoses: suicidal behaviour (including self-harm and serious suicide attempts); depressive disorders; anxiety disorders; any personality disorder, APD and BPD specifically; paraphilic disorders; psychotic disorders; ADHD; alcohol abuse/dependence; and drug abuse/dependence. In the case of offenders who were given a diagnosis of either APD and/or BPD, these individuals were also coded as having a personality disorder.

Although it might be argued that suicidal behaviour should not be considered a diagnosis, it has been observed by our staff that, with very few exceptions, those individuals who have engaged in significant self-harm present with a number of serious mental health concerns. As it was not possible to code all diagnostic conditions for the purpose of this study (and a conservative definition of mental illness was used in the present investigation), it was felt that this item should be retained as a proxy for having evidence of serious mental illness.

Outcome

Recidivism data were collected based on officially recorded criminal history information. In Canada, the RCMP maintain a database which includes information on all charges and convictions registered anywhere in the country. This national database was used for the purpose of collecting all data related to charges and/or convictions. Data regarding suspensions and revocations of conditional release were obtained from databases operated by CSC. Offenders may be suspended or revoked, that is returned to an institutional setting, without having committed a new offence. For example, in cases where an offender has tested positive for an illicit substance (e.g., cocaine) or a drug that is legal but prohibited in the case of a particular offender (e.g., alcohol), he may be returned to prison.

Actuarial Risk

Information regarding actuarially assessed risk was obtained by collecting data on the Statistical Information on Recidivism–Revised 1 Scale (SIR-R1; Nafekh & Motiuk, 2002). Lower scores on this measure indicate that the offender is rated as being at a higher risk of failure on conditional release. The lowest possible score on the SIR-R1 is −30 (corresponding to the highest possible risk level).

The SIR-R1 consists of 15 items which have been empirically associated with recidivism. These items included age at admission, previous incarcerations, previous escapes, age at first adult conviction, previous convictions for violent sexual convictions, previous convictions for break and enter, and employment status at arrest. Although there are relatively few studies using the SIR-R1 as a dependent measure, a number of studies have demonstrated that the earlier version of this instrument (i.e., the SIR) is a moderate predictor of recidivism among offender populations (see Hanson & Morton-Bourgon, 2009).

In the developmental sample of 6,881 offenders, the internal consistency (standardized Cronbach’s α) was reported to be 0.75. The receiver operating characteristic (ROC) area under the curve for predicting general recidivism was reported to be .745.

Results

Criminal History

Criminal history data were available on all but two offenders. Almost all offenders had a history of general and violent offending (97.8% and 92.5%, respectively). Approximately 20% of the sample had a history of sexual offending (18.7%). The mean SIR-R1 score for the sample was −6.66 (SD = 7.63). A score of −6 on the SIR-R1 corresponds to the second-highest risk category on this measure and is associated with two out of every five offenders not committing an indictable offence after release.

Outcome

The mean follow-up time for the sample was 41.9 months (SD = 11.5). As noted above, outcome was coded in terms of charges and convictions. For general recidivism, 84 men (61.7%) received new charges during the follow-up period, with 51 (37.5%) receiving new convictions. For violent recidivism, 36 men (26.5%) received new charges and 27 (19.9%) new convictions. Finally, only 1 man received a new charge for a sexual offence (0.7%), and no one received a new conviction for a sexual offence. Overall, 60 offenders (44.1%) were convicted for a new offence during the follow-up period.

Mental Health Data

Information regarding mental health diagnoses/suicidal behaviour was collected on the 11 variables noted under the section “Diagnoses” (Table 1). A substantial percentage of offenders included in the sample presented with mental illness. Almost a third of the sample had a recent history of self-harm, over half of the sample had been diagnosed with a personality disorder, and almost 20% of the sample had been diagnosed with a psychotic disorder.

Table 1.

Frequency of Mental Health Data.

Diagnosis/behaviour	Percentage yes	Number	Missing
Depression	25.2	127	9
Suicidal behaviour	31.5	127	9
Anxiety disorder	10.2	127	9
Personality disorder	56.6	129	7
APD	34.1	129	7
BPD	4.7	127	9
Paraphilia	11.1	126	10
Psychotic disorder	19.7	127	9
ADHD	18.8	128	8
Alcohol	37.5	128	8
Drugs	39.1	128	8

Note: APD = antisocial personality disorder; BPD = borderline personality disorder; ADHD = attention deficit hyperactivity disorder.

A series of chi-square analyses were conducted with these 11 variables, using any suspensions, charges, or convictions as the dependent measure. When the overall chi-square analysis was significant, an inspection of standardized residuals (equivalent to z scores) was used to determine whether the presence of the particular diagnosis was significantly associated with either an increased or a decreased risk of recidivism. To minimize the probability of a Type 1 error, only cells with standardized residuals of greater than ±2 were used for the basis of comparison. Where the overall chi-square was significant but none of the cells yielded standardized residuals of greater than ±2, these data were treated as nonsignificant. Those offenders with a diagnosis of BPD were significantly more likely to be suspended or receive new charges or convictions, χ²(1) = 4.67, p < .05, as were those offenders with a diagnosis of ADHD, χ²(1) = 7.9, p < .01. Interestingly, offenders with a diagnosis of a paraphilia were significantly less likely to be suspended or recidivate, χ²(1) = 16.01, p < .01.

For those variables that were determined to be significant in the analyses reported above (i.e., BPD, ADHD, and Paraphilia), subsequent Cox Regression analyses were performed to determine whether, after controlling for actuarial risk, a significant association still existed with reference to recidivism. As Cox Regression is a form of survival analysis, this method allows for the control of time at risk in the analyses.

Analyses were first conducted for any and then for violent recidivism (see Table 2). Note that for both outcomes, charges and convictions were counted as recidivism. For any recidivism, the SIR-R1 was entered on the first block and BPD, ADHD, and Paraphilia diagnoses were entered simultaneously in the second block. The SIR by itself was a significant predictor of recidivism, χ²(1) = 8.94, p = .003. When the diagnoses were entered on the second block, the change from the previous block was marginally significant, χ²(3) = 7.62, p = .054; for the overall model, χ²(4) = 16.96, p = .002. Results indicate that those with a diagnosis of BPD were approximately three times as likely to reoffend as those without, while those with a diagnosis of ADHD were approximately twice as likely to reoffend. Those with a paraphilia diagnosis, however, reoffended at a rate about one third that of other offenders. Results for the complete model are present at the upper part of Table 2.

Table 2.

Cox Regression Analysis for Dependent Variables in the Prediction of Recidivism.

	B	SE	Wald	df	p	Exp(B)
Any recidivism
SIR-R1 score	−0.065	0.024	7.111	1	.008	0.937
BPD	1.103	0.544	4.112	1	.043	3.013
Paraphilia	−1.077	1.029	1.095	1	.295	0.341
ADHD	0.687	0.374	3.379	1	.066	1.987
Violent recidivism
SIR-R1 score	−0.057	0.034	2.856	1	.091	0.945
BPD	1.901	0.589	10.409	1	.001	6.692
Paraphilia	−0.288	1.057	0.074	1	.785	0.750
ADHD	0.453	0.575	0.622	1	.430	1.573

Note: SIR-R1 = Statistical Information on Recidivism–Revised 1 Scale; BPD = borderline personality disorder; ADHD = attention deficit hyperactivity disorder.

For violent recidivism, once again the SIR-R1 was entered in the first block and the diagnosis variables were entered in the second block. On the first block, the SIR-R1 was a marginally significant predictor of violent recidivism, χ²(1) = 3.66, p = .056, and when the diagnosis variables were added on the second block, the change from the previous block was significant, χ²(3) = 7.85, p = .049, with the overall model, χ²(4) = 8.65, p = .001 (see lower part of Table 2 for the complete model). Results indicate that men with a diagnosis of BPD were more than six times as likely to reoffend as those without the diagnosis.

To rule out effects of multicollinearity, VIF and Tolerance were calculated for the four variables included in the above analyses. None of the tolerance values were below .80 and all of the VIF values were below 2.0, indicating that multicollinearity did not influence the results in the preceding paragraphs.

Discussion

In keeping with the results of the more recent literature related to mental illness and recidivism, the findings of the present study suggest that at least some forms of serious mental illness are related to recidivism among a correctional sample. Several aspects of the present study are noteworthy. First, the absolute rates of mental illness were very high in the present sample. It should be emphasized that this sample was pre-selected to be at high risk of recidivism given that the population was housed at a CCC. That being said, the sample was not pre-selected to present with serious mental illness. The Keele CCC is not a psychiatric facility; it is considered a minimum security institution that operates in a community environment. Although classified as a minimum security institution, the Keele CCC represents the highest security halfway house available in the community as it is operated by CSC rather than one of the community partners with whom CSC works. A variety of security features are present that may not be available in other halfway houses. Many of the samples that have been reported in the literature have used offenders housed in psychiatric facilities or who were included because of such diagnoses being present on file (e.g., Ashford, Wong, & Sternbach, 2008; Dinn, Gansler, Moczynski, & Fulwiler, 2009), and so they do not directly assess the rates of mental illness among groups of high-risk offenders per se. Such data are critical, however, to forensic managers as they provide a more comprehensive picture of the needs of such high-risk samples, which also tend to be very high profile offenders. The findings of the present study suggest that there may be a need for increased levels of mental health resources in forensic environments that cater to the needs of high-risk offenders. The rates of mental illness observed in the present population are very high relative to community samples and suggest that standard correctional programs may be insufficient to meet the needs of such clients.

A number of the findings for specific diagnoses suggest that studies that simply investigate overall rate of mental illness and the relationship with recidivism in correctional samples may be obscuring the results that are present if data related to the presence or absence of specific diagnoses are used.

The fact that paraphilic diagnoses were related to lower rates of recidivism may be accounted for by any number of factors. First, given the high profile nature of some of the sex offenders housed at the Keele CCC, they may be given restricted access to the community, at least for the first few months of their release. As well, all of the sex offenders housed at Keele were mandated to receive psychological treatment directed toward their sex offender treatment needs. This treatment and/or the sex drive reducing medications that these clients may have been prescribed would likely have reduced the clients’ risk of recidivism somewhat. That being said, a number of the nonsexual offenders were also mandated to receive treatment either of a psychiatric and/or of a psychological nature. It is also possible that the poor reliability/validity reported in the literature with reference to paraphilic diagnoses may have accounted for this finding (e.g., Kingston, Firestone, Moulden, & Bradford, 2007; Marshall, Kennedy, & Yates, 2002; Wollert, 2007). That is, if a diagnosis is associated with poor/unknown reliability, it is likely that some investigations will observe a positive association with an outcome of interest, some will observe a negative association, and some will find no association whatsoever. Although it is premature to come to any definitive conclusions regarding this matter, the available literature supports the contention above regarding reliability with a number of studies questioning either the reliability and/or the validity of the paraphilic diagnoses. Furthermore, our team has conducted research demonstrating that, after controlling for actuarially assessed risk, a diagnosis of pedophilia was not related to increased rates of recidivism among a group of convicted sex offenders (Wilson, Abracen, Looman, Picheca, & Ferguson, 2011).

The fact that BPD and ADHD were significantly associated with outcome in the chi-square analyses is in keeping with the literature reported above. The fact that psychotic disorder was not significantly associated with recidivism goes counter to prediction as this disorder is increasingly found to be related to recidivism as noted in the introduction.

Overall, the findings of the present investigation suggest that mental health concerns represent a significant issue among higher risk offender populations released to the community. Of particular note, the majority of the offenders in the present investigation likely met diagnostic criteria for more than one mental disorder. As well, given that diagnoses were only available for those clients who had been referred for a psychiatric assessment or where they were discussed in psychological reports, the observed rates of mental illness were most certainly underestimates of the true rates of mental illness in the current sample. As well, many clinicians do not report diagnoses for substance abuse conditions given that the vast majority of offenders meet diagnostic criteria for one or more substance abuse conditions. For example, Boland, Henderson, and Baker (1998) reported that upward of two thirds of offenders present with at least some difficulties in the area of substance abuse. This was particularly the case for substance abuse disorders where anecdotal evidence suggested that the vast majority of offenders housed at Keele have demonstrated at least moderate difficulties in the area of alcohol and drug abuse. The fact that a recent history of neither alcohol nor drug abuse was found to be related to recidivism is likely a result of the conservative strategy used to determine the presence of mental illness in the present sample. High-risk clients who present with comorbid diagnoses present unique challenges to those working in correctional environments.

Taken with the extant literature, these data suggest that perhaps the views of Andrews and Bonta (2010, for example) may need to be revised to include significant psychiatric conditions/comorbid conditions as potentially criminogenic. Several recent studies underscore the importance of mental illness in relation to recidivism in offender populations. Cloyes, Wong, Latimer, and Abarca (2010), noting that recidivism studies do not generally focus specifically on persons with serious mental illness, observed that 23% of their sample of more than 9,000 offenders in Utah State met criteria for serious mental illness. In that sample, survival analyses indicated significant differences between offenders with and without serious mental illness. The median time for all offenders with serious mental illness to return to prison was 385 days versus 743 days for offenders without serious mental illness. In their meta-analysis, Olver et al. (2011) found that treatment noncompleters were higher risk as assessed by actuarial assessment and were more likely to recidivate. Most important for the present discussion, those conditions indicative of more serious pathology (e.g., psychosis, personality disorder, borderline) predicted treatment noncompletion. These data highlight the potential direct and indirect routes through which mental illness might be related to recidivism in forensic populations, underscoring the importance of developing programs designed specifically for such groups of offenders.

Limitations

A limitation of the present study is the small sample size. This limits the extent to which our findings can be generalized to other samples without replication. However, as noted above, the findings are consistent with other large-scale studies that have found increased rates of recidivism among offenders suffering with particular types of mental illness. There may also be some concern related to using Cox Regression with a small sample; however, the Cox regression analysis employed only four variables; thus, our sample is sufficiently large.

Summary

The present investigation demonstrates that very high rates of mental illness may be present in high-risk offender populations. Further, at least some forms of mental illness appear to be related to increased rates of recidivism. The available evidence suggests that the nature of offender populations may be changing and that there is a need to adapt our current assessment and treatment strategies to meet the needs of this changing offender population.

Footnotes

Authors’ Note

The views expressed are those of the authors and do not necessarily reflect the views of Correctional Service of Canada.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Abracen

Axford

Gileno

(2012). Changes in the profile of offender populations residing in community facilities: 1998 and 2008 (No. R-256). Ottawa, Ontario: Public Safety Canada.

Abracen

Looman

(2006). Evaluation of civil commitment criteria in a high risk sample of sexual offenders. Journal of Sexual Offender Civil Commitment: Science and the Law, 1, 124-140.

Adams

Ferrandino

(2008). Managing mentally ill inmates in prisons. Criminal Justice and Behavior, 35, 913-927.

Andrews

D. A.

Bonta

(2003). The psychology of criminal conduct (3rd ed.). Cincinnati, OH: Anderson.

Andrews

D. A.

Bonta

(2010). The psychology of criminal conduct (5th ed.). New Providence, NJ: Mathew Bender.

Ashford

J. B.

Wong

K. W.

Sternbach

K. O.

(2008). Generic correctional programming for mentally ill offenders: A pilot study. Criminal Justice and Behavior, 35, 457-473.

Barkley

R. A.

(2006). Attention-deficit hyperactivity disorder: A handbook for diagnosis and treatment (3rd ed.). New York, NY: Guilford.

Barkley

R. A.

Murphy

K. R.

Fischer

(2008). ADHD in adults: What the science says. New York, NY: Guilford.

Boland

Henderson

Baker

(1998). Case need domain: Substance abuse assessment review. Forum on Corrections Research, 10, 32-34.

10.

Butler

Andrews

Allnutt

Sakashita

Smith

N. E.

Basson

(2006). Mental disorders in Australian prisoners: A comparison with a community sample. Australian and New Zealand Journal of Psychiatry, 40, 272-276.

11.

Cloyes

K. G.

Wong

Latimer

Abarca

(2010). Time to prison return for offenders with serious mental illness released from prison: A survival analysis. Criminal Justice and Behavior, 37, 175-187.

12.

Constantine

R. J.

Petrila

Andel

Givens

E. M.

Becker

Robst

Howe

(2010). Arrest trajectories of adult offenders with a serious mental illness. Psychology, Public Policy, and Law, 16, 319-339.

13.

Dinn

W. M.

Gansler

D. A.

Moczynski

Fulwiler

(2009). Brain dysfunction and community violence in patients with major mental illness. Criminal Justice and Behavior, 36, 117-136.

14.

Douglas

K. S.

Guy

L. S.

Hart

S. D.

(2009). Psychosis as a risk factor for violence to others: A meta-analysis. Psychological Bulletin, 135, 679-706.

15.

Hanson

R. K.

Morton-Bourgon

K. E.

(2009). The accuracy of recidivism risk assessments for sexual offenders: A meta-analysis of 118 prediction studies. Psychological Assessment, 21, 1-21.

16.

Hull

J. G.

Slone

L. B.

(2004). Alcohol and self-regulation. In Baumeister

R. F.

Vohs

K. D.

(Eds.), Handbook of self-regulation: Research, theory, and applications (pp. 466-491). New York, NY: Guilford.

17.

Kingston

D. A.

Firestone

Moulden

H. M.

Bradford

J. M.

(2007). The utility of the diagnosis of pedophilia: A comparison of various classification procedures. Archives of Sexual Behavior, 36, 423-436.

18.

Lamb

H. R.

Weinberger

L. E.

Gross

B. H.

(2004). Mentally ill persons in the criminal justice system: Some perspectives. Psychiatric Quarterly, 75, 107-126.

19.

Långström

Sjöstedt

Grann

(2004). Psychiatric disorders and recidivism in sexual offenders. Sexual Abuse: A Journal of Research and Treatment, 16, 139-150.

20.

Langton

C. M.

Hogue

T. E.

Daffern

Mannion

Howells

(2011). Personality traits as predictors of inpatient aggression in a high-security forensic psychiatric setting: Prospective evaluation of the PCL-R and IPDE dimension ratings. International Journal of Offender Therapy and Comparative Criminology, 55, 392-415.

21.

Looman

Abracen

(2012). The Static-99: Are there really differences between the treatment need and high risk normative groups? International Journal of Offender Therapy and Comparative Criminology. doi:10.1177/0306624X12443657.

22.

Marshall

W. L.

Kennedy

Yates

P. M.

(2002). Issues concerning the reliability and validity of the diagnosis of sexual sadism applied in prison settings. Sexual Abuse: A Journal of Research and Treatment, 14, 301-311.

23.

Morrissey

Meyer

Cuddeback

(2007). Extending assertive community treatment to criminal justice settings: Origins, current evidence, and future directions. Community Mental Health Journal, 43, 527-544.

24.

Nafekh

Motiuk

(2002). The Statistical Information on Recidivism–Revised 1 (SIR-R1) Scale: A psychometric examination (No. R-126). Ottawa, Ontario: Correctional Service of Canada.

25.

Olver

M. E.

Stockdale

K. C.

Wormith

J. S.

(2011). A meta-analysis of predictors of offender treatment attrition and its relationship to recidivism. Journal of Consulting and Clinical Psychology, 79, 6-21.

26.

Porporino

F. J.

Motiuk

L. L.

(1995). The prison careers of mentally disordered offenders. International Journal of Law and Psychiatry, 18, 29-44.

27.

Skeem

J. L.

Emke-Francis

Louden

J. E.

(2006). Probation, mental health, and mandated treatment: A national survey. Criminal Justice and Behavior, 33, 158-184.

28.

Skeem

J. L.

Louden

J. E.

(2006). Toward evidence-based practice for probationers and parolees mandated to mental health treatment. Psychiatric Services, 57, 333-342.

29.

Skeem

J. L.

Manchak

Peterson

J. K.

(2011). Correctional policy for offenders with mental illness: Creating a new paradigm for recidivism reduction. Law and Human Behavior, 35, 110-126.

30.

Wilens

T. E.

Biederman

Abrantes

A. M.

Spencer

T. J.

(1997). Clinical characteristics of psychiatrically referred adolescent outpatients with substance use disorder. Journal of the American Academy of Child & Adolescent Psychiatry, 36, 941-947.

31.

Wilson

R. J.

Abracen

Looman

Picheca

J. E.

Ferguson

(2011). Pedophilia: An evaluation of diagnostic and risk prediction methods. Sexual Abuse: A Journal of Research and Treatment, 23, 260-274.

32.

Wollert

(2007). Poor diagnostic reliability, the null-bayes logic model, and their implications for sexually violent predator evaluations. Psychology, Public Policy, and Law, 13, 167-203.

33.

Wolraich

M. L.

Wibbelsman

C. J.

Brown

T. E.

Evans

S. W.

Gotlieb

E. M.

Knight

J. R.

Wilens

(2005). Attention-deficit/hyperactivity disorder among adolescents: A review of the diagnosis, treatment, and clinical implications. Pediatrics, 115, 1734-1746.