Abstract
The last decade has seen a general increase in interest in the level of psychiatric disorder within prisons internationally. A review by Monahan and Steadman [1] found only six studies in the literature relating to true prevalence studies in USA prison populations prior to 1982. Subsequent years have seen an increase in studies focusing on the prevalence rates of mental disorders in prison and also on the degree of functional impairment of mentally ill prisoners. Dvoskin and Steadman [2], in particular, reviewed 9.4% of the 36 144 inmates in New York State looking at the degree to which mentally ill inmates were impaired, finding 5% being severely psychiatrically disabled and a further 10% significantly disabled. These figures rose to 8% and 16% respectively when measures of functional disability were also considered which were felt likely to reflect the existence of mental illness, mental retardation or other developmental disabilities. In 1990, Teplin [3] published studies showing the prevalence rate of schizophrenia within a jail population in the USA of 3%, mania 1% and current major depression at 4%. This was felt to be similar to the prevalence rates for such disorders found in the community. Similar studies were performed in Canada by Bland et al. [4] and Hodgins and Coté [5]. Bland et al. [4] surveyed 180 randomly selected male inmates using the Diagnostic Interview Schedule [6], comparing their results with those of a randomly selected non-inmate population. They reported that prison inmates were twice as likely to have a lifetime psychiatric disorder compared with a matched community sample and that 6-month prevalence rates for most psychiatric disorders were significantly raised in the prison population compared with the community sample. Similarly, Hodgins and Coté [5], in reviewing 495 inmates using the Structured Clinical Interview in a Quebec penitentiary sample, found rates of schizophrenia seven times, major depression twice and bipolar disorder four times that of the community sample.
In the UK Taylor and Gunn [7] had demonstrated higher prevalence rates for psychiatric disorder in remand male prison populations. However, the increased rates of psychiatric disorder within the sentenced inmate population of England and Wales was shown definitively by Gunn et al. [8] in their study of the treatment needs of a 5% cohort of all sentenced male inmates. They reviewed 1151 men using a semistructured diagnostic interview in conjunction with a clinical interview, demonstrating current prevalence rates of 1% for schizophrenia and 0.4% for affective psychosis. Current diagnosis of substance abuse occurred in 23% and severe personality disorder in 10%. Overall it was the authors’ view that the prevalence rate for psychiatric disorders in the prison system was sufficiently high to necessitate urgent attention. Gunn et al. 's group [8, 9] have continued their work in this area studying the treatment needs of both sentenced and remand prisoners. In these studies prisoners were allocated to one of four treatment options: prisoner health services, motivational interviewing for substance misuse, hospital transfer and assessment for therapeutic community. Need for hospital transfer was defined as a clinical judgement reserved for those prisoners who were severely mentally unwell and unable to be safely managed in prison on the basis of their risks to themselves or others. Hospital admission was recommended for 88.2% of the psychotic prisoners in the sentenced group and in 80.6% of psychotic prisoners in the remand group. Their estimates were that prisons in England and Wales contained more than 1000 men suffering from psychosis and nearly 2000 male prisoners overall who needed treatment in psychiatric hospitals. Birmingham [10] in his editorial ‘Between prison and the community: “the revolving door psychiatric patient” of the nineties’, cites the 1998 Office for National Statistics report on the prevalence of psychiatric morbidity among prisoners in England and Wales. This paints an even bleaker picture than previous studies, reporting 7% of male sentenced and 10% of male remand prisoners suffering from functional psychotic disorders. When these figures are generalized to the overall prison population, it is suggested that there may be in excess of 4500 men with psychotic disorders in prison at the present time in the UK.
Australasian studies
There have been few studies looking at the level of mental disorder within prisons in Australia or New Zealand. Glaser [11] studied 50 consecutive admissions to a psychiatric unit in Pentridge prison in Melbourne, Australia, comparing his results with other unpublished Australian studies. He found 42% suffering from schizophrenia and 16% affective disorders. Schizophrenia and affective disorders were the most frequent diagnoses. Glaser's population comprised inmates admitted into a psychiatric unit resulting in the very high proportions of severe psychiatric disorder in his group compared with prevalence studies from general prison populations. The author stressed the need for a large scale epidemiological survey in Australian prisons, to establish not only the type and extent of psychiatric disturbance, but also what the prison services were likely to be able to do about it. Hurley and Dunne [12] found high levels of psychological distress among women inmates in the Brisbane Women's Prison. They estimated the prevalence rate of severe disturbance in inmates was approximately 150% greater than in women living in the community. They noted the very high rate of psychoactive substanceuse disorders in their population. Herrman et al. [13] reported prevalence of severe psychiatric disorder in a group of sentenced male inmates in a Melbourne prison. They used the structured clinical interview for DSMIII-R, interviewing 158 men and 38 women. They found 3% with current psychotic disorders and 12% with current mood disorder (mainly major depression). Many of these were untreated. Comparable with international studies, 69% of inmates received a lifetime diagnosis of dependence or abuse of alcohol and other psychoactive substances.
In New Zealand, little information regarding the prevalence rates of psychiatric disorder in the prison population was available prior to the Christchurch Prisons Psychiatric Epidemiology Study [14]. Bushnell and Bakker [15] published a study of prevalence rates for drug and alcohol abuse and dependence in Canterbury prisons, but no other published studies have focused on prevalence rates for major mental disorder in New Zealand prisons.
The Christchurch Prisons Epidemiology Study
The Christchurch Prisons Epidemiology Study [14] reviewed a total of 225 inmates, representing a census of female inmates and remand males and a randomly selected cohort of 125 sentenced males. Inmates were assessed by administering the Composite International Diagnostic Interview –Automated (CIDI-A) [16], a structured interview generating both DSM-III-R and ICD-10 diagnosis. Three different measures of personality were also used to assess the level of personality disorder within this population, those being questions from the Antisocial Personality Section of the Structured Clinical Interview for DSM-III-R Personality Disorders (SCID 2) [17], the Four ‘A's [18], and the Temperament and Character Inventory −238 item version (TCI) [19]. Results revealed high lifetime and current prevalence rates for drug and alcohol abuse and dependence, as well as high current and lifetime prevalence rates for affective disorders. Schizophrenia had a lifetime prevalence rate of 5% within the sentenced male population. When comparison was made with community epidemiological figures for psychiatric disorder [20], elevated lifetime and current prevalence rates among prison inmates were found for all major psychiatric diagnoses except generalized anxiety disorder. Personality disorder assessment showed high levels of antisocial personality disorder (71% of sentenced males, 71% of remand males and 39% of sentenced females) [21]. Other types of personality assessment (Four ‘A's and TCI) also revealed significant numbers of other types of personality disorder, most noticeably large numbers of inmates who were avoidant and dependant, socially indifferent and asocial. It was felt that these inmates would pose particular management difficulties within the prison system given their tendency to operate on the very periphery of the prison ‘culture’ and choosing to isolate themselves both from a social and therapeutic perspective.
The Christchurch Prisons Epidemiology Study acted as a pilot study for the national study that is the subject of this paper.
Aim
New Zealand, as an international signatory to many documents from the United Nations Commissioner for Human Rights, is required to maintain an adequate standard of mental health care in prisons. The authors were aware from their own clinical experience working in the prisons that many claims were being made by both prison staff and inmates that inmate mental health needs were not being met. The aim of this study was to measure the level of mental illness in New Zealand prisons and estimate how well treatment needs were being currently met.
Method
Subjects
Ethical approval was obtained for the study from the relevant regional ethics committees in New Zealand.
We approached all female prisoners, all male remand prisoners and a 18% sample of sentenced male prisoners from every prison in New Zealand. The 18% of male sentenced prisoners were selected by random number tables from the muster list and stratified by ‘security rating’. Security rating refers to the rating of each sentenced prisoner according to a protocol developed by the Department of Corrections. It allocates the prisoner to maximum, medium or minimum security level according to a mixture of offence, sentence and behavioural criteria. Having a mental illness contributes to a higher security rating. The stratification ensured that 15% of prisoners from each security rating were included, to ensure adequate sampling of those at higher security levels where it was anticipated there may be more mentally disordered prisoners. Sentenced prisoners who were selected but were unavailable were replaced by the next name on the muster list.
All female prisoners (remand and sentenced) and all male remand prisoners were approached. Interviewing occurred from October to December 1997, and February to June 1998, with a break during the Christmas period when prisoners may be granted early release or home leave and the Courts are generally in recess.
Assessment tools
The interview process was comprised of five sections: informed consent, demographic data collection, diagnostic interview for mental illness (CIDI-A), screening diagnostic interview for relevant personality disorders (PDQ 4+) and specific information about suicidal thoughts or actions. Demographic data collection included a variety of different demographic variables including information regarding past psychiatric history, past psychiatric treatment and history of previous psychiatric hospitalization. These results and those regarding suicidal thoughts will be published separately.
The CIDI-A was selected for this study. It can be administered by lay interviewers, is widely accepted internationally, fully computerized and had been successfully employed in the Christchurch Prison Study [14].
The CIDI-A (version 2.1) data were scored for DSM-IV diagnosis using the WHO SPSS scoring algorithms (Release 2.1, May 1998). The version of the CIDI-A was chosen which comprised sections examining anxiety disorders, depression, mania, psychosis, eating disorders, alcohol, drugs, obsessive–compulsive disorder, posttraumatic stress disorder, and the mini-mental state examination. These sections were chosen as they identified the most common and clinically significant disorders in the prison setting. The PDQ 4+ [22] was selected to screen for relevant personality disorders. Several studies attest to the high sensitivity and only modest specificity of the PDQ. The PDQ questionnaire seems particularly overinclusive with high rates of false positives [23–26]. To limit interviewing time we screened for the five most relevant personality disorders which DSM-IV refers to as the cluster A and B disorders, specifically borderline, antisocial, narcissistic, histrionic and paranoid personality disorders. The PDQ 4+ was computerized and followed the CIDI-A interview.
Five extra questions were added at the end of the PDQ regarding suicidal thinking, feelings, plans or past actions while in prison. These were to complement the questions about suicidal thoughts that are embedded in the depression subsection of the CIDI-A.
Data were collated into Microsoft Excel spreadsheets, then Access tables. Matching between demographics and computerized files for the CIDI-A and PDQ 4+ resulted in master data files that were convertible to the Statistics Package for the Social Sciences (SPSS, Chicago, IL, USA) [27] for analysis. Results of the PDQ4 + will be published separately.
Procedure
The interviews were performed by four regional teams. Each team had a team leader, an experienced clinician, who supervised eight to 14 interviewers and their data collections from 220 to 350 prisoners. Liaison meetings were held with prison management and at times with groups of prisoners prior to the interviews commencing. All interviewers were trained for 2 days in the use of all instruments in the prison environment. Interviewers had a range of skills from mental health knowledge held by clinical psychology students and psychiatric nurses to interviewing skills held by lay people. All were instructed to raise any concerns about the wellbeing of any prisoner they interviewed with the team leader or the clinical coordinator of the team. Referral to the regional forensic psychiatric services was offered to prisoners where appropriate.
Each potential subject was approached by the team leader, who explained the study, gave the prisoner a copy of the subject information sheet and obtained a signed consent form. An 18-point demographic form taking approximately 10 min, was completed with the prisoner, after which the team leader scheduled the diagnostic interview within the next few days. If the prisoner could not speak English, an interpreter was arranged.
Interviews took approximately 1 h and 15 min (range 35 min to over 3 h). As the PDQ 4+ interview was performed last, the completion rate for the PDQ 4+ was lower than for the CIDI-A.
Results
Ethnicity
At the time of the study there were 4447 sentenced males, 540 male remands and 200 remand and sentenced women in the New Zealand prison population. By far the largest ethnic group within the prison population is Maori, constituting 52.4% of the women, 49.8% of remand men and 46.2% of sentenced men. When compared with the general population of New Zealand, in which Maori make up 12% of the population, it can be seen that Maori are grossly overrepresented in all inmate groups. (Results from Maori are to be reported in a separate paper).
Acceptance and completion rates
At the time of the study there were 4447 sentenced males, 540 remand males and 200 remand and sentenced women in the New Zealand prison population. Acceptance and completion rates for this study are shown in Table 1. All female and remand inmates were approached to take part in the study as were 18% of the sentenced male inmates. Of these, 170 women, 452 male remands and 660 male sentenced inmates consented to enter the study and completed the consent and demographic forms. Thirty-four inmates either later withdrew consent or were not available to complete the CIDI-A [16], resulting in 1253 completed demographic and CIDI-A interviews. A further 94 inmates did not complete the PDQ 4+ or suicide questions, resulting in 1159 completed interviews for those sections.
Acceptance and completion rates
There were some problems encountered when inmates were approached to take part in the study. First, a majority of inmates in the Christchurch Women's Prison had previously been approached to take part in the Christchurch Prisons Epidemiology Study. This prison is somewhat unique in having a number of women serving long sentences who did not wish to be re-interviewed. It was also notable that a policy initiative to reduce drug use in prison was launched during our study period, which some inmates associated with the research project resulting in some declining to participate. Such effects were deemed to be largely local and did not involve large numbers of inmates.
The prevalence rates (month prior to interview) for the most frequently diagnosed mental disorders are shown in Table 2. Thirty-six inmates were found to be suffering a psychotic disorder as a result of schizophrenia or related disorder (schizoaffective and delusional disorder). It is, however, acknowledged that these disorders tend to be lifelong and it is important therefore to note the lifetime prevalence rates of 8.5% (n = 14) for female inmates, 7.9% (n = 35) for remanded males and 6.6% (n = 43) for sentenced inmates with regard to schizophrenic disorders. We therefore identified 92 inmates overall with a lifetime diagnosis of schizophrenia or related disorder. All those inmates identified as currently suffering bipolar disorder met criteria for moderately severe or severe mania. In excess of 10% of female prisoners and male remand prisoners were suffering from a current episode of major depression. The level of major depression in the sentenced inmate population is approximately twice the rate of current depression in the general population [20]. Levels of obsessive–compulsive disorder in the prison population are similarly elevated when compared with the general population. However, we found the current prevalence rate of posttraumatic stress disorder to be grossly elevated in the prison population compared with the community [20, 28, 29].
Prevalence rates for mental disorder in last month
Results for substance abuse and dependence reflect a definition of abuse where psychoactive substance use results in some degree of psychosocial harm, and substance dependence where more serious substance use is associated with tolerance and withdrawal symptoms. Substances classified as ‘other’ included hallucinogens, sedatives, opioids, cocaine, amphetamines and inhalants.
Current prevalence rates for mental disorder reveal the male remand population to be more disturbed overall than the sentenced male inmate population. With the exception of obsessive–compulsive disorder, female inmates also displayed greater levels of mental disorder than their sentenced male counterparts, in particular suffering twice the rates of major depression and posttraumatic stress disorder when compared with sentenced males.
Inmates were also asked whether or not they were receiving treatment for any mental disorder during their current period of incarceration (see Table 3). Results revealed that while 80.8% of those with bipolar disorder were receiving treatment, a lesser percentage of those suffering major depressive disorder were in treatment (46.4%) and only 37.0% of those identified as suffering schizophrenia and related disorders were in treatment. A total of 55.3% of inmates with obsessive–compulsive disorder and 41.4% of inmates with posttraumatic stress disorder reported receiving treatment. Inmates were not asked if they had been offered treatment and refused. Their medical files were not consulted.
Self-report of treatment by those inmates with a lifetime diagnosis of mental disorder
Discussion
The purpose of the study was to identify the level of psychiatric need within the New Zealand prison population on any typical day. The only previous large-scale New Zealand psychiatric epidemiology study in the community is the Christchurch Psychiatric Epidemiology Study [20]. That study gives an indication of 1-month prevalence rates and lifetime prevalence rates for major mental disorders. We have compared our results with that study. Current prevalence rates for schizophrenia and related disorders in the community are estimated to be 0.1% with a lifetime rate of 0.3%. Our rates of schizophrenic disorders within the prison population are grossly elevated compared with those within the community. This is also the case for bipolar disorder, major depressive episode, posttraumatic stress disorder, obsessive–compulsive disorder and alcohol and drug abuse and dependence. This confirms the findings of the Christchurch Prison Study [21]. Using the CIDI-A, we have interpreted the detection of symptoms in the last month in an inmate's history as an indication of current mental status abnormality, indicating a need for consideration to be given to intervention and treatment. Gunn et al. [8] used clinical and semistructured interviews to identify the level of psychiatric disorder within the prison population in England and Wales. While we did not clinically interview the inmates in our study, it was felt that the CIDI-A was an appropriate instrument with which to identify those inmates currently suffering from mental disorder and in need of treatment. It is possible that the CIDI-A exaggerates the prevalence of non-affective psychotic illness, however, it is known to be particularly accurate in predominantly young, male populations [16], the primary characteristics of this prison population. We felt therefore that a conservative comparison with Gunn et al. 's study of treatment needs of prisoners with psychiatric disorder was appropriate [8]. At the time of our study the New Zealand prison inmate population consisted of 200 remand and sentenced women inmates, 540 remand male inmates and 4447 sentenced male inmates. By extrapolation our current prevalence rates for psychiatric disorder reveal that approximately 121 inmates nationally would suffer from schizophrenia or a related disorder currently, 54 would be suffering from moderately severe or severe mania as a result of bipolar disorder, the majority of these requiring hospital treatment. If we apply the same rates of need for hospital transfer that Gunn's group found on clinical examination, this would represent about 135 inmates currently requiring hospital care for acute psychotic illness (see Table 4).
Acute service needs of prisoners with a diagnosis of serious mental illness in the last month
We do not believe that this represents the true number requiring transfer (due to the possibility of the CIDI-A overdiagnosing some psychotic illness), but it should be remembered that these inmates are in addition to all the inmates currently being treated in hospital at the time of the study.
Approximately 10 per cent of all inmates currently suffer from an episode of major depression, representing over 500 inmates. Depending upon the level of depression and associated suicidal ideation or self-injury, a proportion of these will also require treatment in hospital. While not necessarily requiring hospital treatment, both obsessive–compulsive disorder and posttraumatic stress disorder can prove particularly debilitating and cause great subjective suffering. Inmates suffering from these disorders require psychiatric assessment and treatment where indicated. Screening for these disorders could be performed by primary medical staff, psychologists or nursing staff within the prison system with referral made as appropriate for psychiatric evaluation by psychiatrists.
Our results suggest that approximately 35% of all inmates suffering from substance-misuse disorders had received some sort of treatment during their sentence. While current prevalence rates for substance-misuse disorders are lower than those found in the Christchurch Prison Study [14], they still represent a problem for medical staff within the prison system, particularly given the frequency of other comorbid major mental disorders. Lifetime prevalence rates for substance-misuse disorders (reported separately) are grossly elevated and similar to those found in the Christchurch Prison Study.
Forensic psychiatric services in New Zealand are centred within the mental health services, each large centre of the country having a regional forensic psychiatric service operating secure inpatient beds, outpatient services, court-liaison services and psychiatric clinics within the regional prisons. Forensic psychiatric services were established following the 1988 Mason Report [30] with the purpose of establishing mental-health based facilities to care for mentally ill offenders and potential offenders. While overall agreement was achieved that the treatment of mentally ill inmates should be performed by the regional forensic psychiatry services, no estimation of need was ever performed. It is also evident that the inmate population in New Zealand has grown significantly since the establishment of regional forensic psychiatric services, with a resulting increase in demand from the prisons for psychiatric consultation. The most severely mentally ill inmates are transferred from prison into psychiatric hospital pursuant to Section 45 of the Mental Health (Compulsory Assessment and Treatment) Act 1992. Inpatient forensic services are already working to capacity with bed occupancy rates of over 100% in most areas of the country [Chaplow D: personal communication]. Despite this, we feel that the results of our study have demonstrated significant unmet need within the prison system with regard to psychiatric treatment.
Many of those inmates suffering from a major mental disorder within the prison system are not receiving treatment. This is particularly noticeable for inmates with major depression (46.4% in treatment) and schizophrenic disorders (37.0% in treatment). This may not be simply due to lack of recognition of illness. Psychiatric treatment [under the Mental Health (Compulsory Assessment and Treatment) Act, 1992] cannot be enforced in prison, and inmates, therefore, make a choice as to whether or not to accept treatment. With conditions such as schizophrenia or delusional disorder, inmates’ mental state may well be disturbed by illness to the point where they cannot make an informed choice regarding treatment. Many may refuse to comply with prescribed medications. Whatever the cause of lack of treatment (either non-diagnosis or non-compliance) only about one-third of inmates with schizophrenia and related disorders are receiving any treatment for the condition.
Using comparisons with Gunn et al. 's UK study [8] we have identified not only a significant number of inmates (suffering from acute psychotic illness) who would require inpatient psychiatric hospitalization, but also large numbers of inmates who would benefit from ongoing psychiatric treatment within the prison system during their current sentence. A review of current service provision nationally reveals that this would be quite beyond the capacity of current forensic psychiatric services, Department of Corrections Psychological Service, or prison nursing and medical officers.
Conclusions and recommendations
We believe that a significant expansion of services should be considered with attention being paid to providing a larger number of secure inpatient beds within psychiatric hospitals. Provision of increased resources is also required to the forensic psychiatric services in order to significantly improve their capacity to provide ambulatory services to inmates with mental disorder of lesser severity that nonetheless cause considerable subjective distress and functional impairment within the prison system. As the number of inmates held in New Zealand prisons continues to rise, the urgent need for such a review of service provision is increasingly evident. Our results also suggest that there is a need for forensic mental health and corrections staff to develop and implement screening procedures that would increase the level of recognition of mental disorder in new receptions into prisons as well as serving inmates.