Prevalence of mental disorder in remand prisoners: consecutive case study

BMJ 1996; 313 doi: http://dx.doi.org/10.1136/bmj.313.7071.1521 (Published 14 December 1996)
Cite this as: BMJ 1996;313:1521
  1. Luke Birmingham, research associatea,
  2. Debbie Mason, research associatea,
  3. Don Grubin, senior lecturera
  1. a Department of Forensic Psychiatry, University of Newcastle upon Tyne, St Nicholas's Hospital, Gosforth NE3 3XT
  1. Correspondence to: Luke Birmingham.
  • Accepted 24 September 1996

Abstract

Objective: To define the prevalence of mental disorder and need for psychiatric treatment in new remand prisoners and to determine to what extent these are recognised and addressed in prison.

Design: Study of consecutive male remand prisoners at reception using a semistructured psychiatric interview.

Setting: Large remand prison for men (HMP Durham).

Subjects: 569 men aged 21 years and over on remand, awaiting trial. Main outcome measures—Prevalence of mental disorder at reception, prisoners need for psychiatric treatment, identification of mental disorder by prison reception screening, and numbers placed appropriately in the prison hospital.

Results: 148 (26%) men had one or more current mental disorders (excluding substance misuse) including 24 who were acutely psychotic. The prison reception screening identified 34 of the men with mental disorder and six of those with acute psychosis. 168 men required psychiatric treatment, 50 of whom required urgent intervention; 16 required immediate transfer to psychiatric hospital. Of these 50, 17 were placed on the hospital wing because of mental disorder recognised at prison screening.

Conclusion: Not only is the prevalence of mental disorder, in particular severe mental illness, high in this population, but the numbers identified at reception are low and subsequent management in prison is poor.

Key messages

  • In this study in Durham prison 26% had a serious mental disorder

  • Only about a quarter of the mentally disordered patients were recognised by the prison doctor and hospital officer

  • Only two fifths of patients who required urgent psychiatric treatment were put into the prison hospital

  • Improved diagnosis of psychiatric illness in prison will greatly increase the pressure on psychiatric beds

Introduction

The problems of mentally disordered people in prisons were highlighted in a series of articles in 1984.1 2 Bluglass subsequently drew attention to the lack of improvement despite numerous inquiries into the management of mentally disordered people in prison.3 He pointed out the inadequacy of Home Office and Department of Health and Social Security reports, which relied heavily on the limited information available from censuses of mentally disturbed prisoners carried out by prison medical officers.

The effectiveness of health screening by prison medical staff has also been questioned.4 5 The conditions and time constraints militate against the detection of clinically important information and the screening questionnaires used are of doubtful validity. In addition, some prisoners are difficult to deal with and allegations have been made of poor standards and lack of training in prison medical staff.

A recent national study of convicted prisoners using data collected by psychiatrists reported that 37% of sentenced prisoners have mental disorders.6 The rates in remand prisoners are probably higher (S Dell, personal communication), partly because mentally disordered people are often remanded in custody for psychiatric reports. Evidence from North America suggests that mentally disordered people are more likely to be arrested than those who are not mentally disordered in similar circumstances.7 Factors such as homelessness8 and petty offences that are associated with mental disorder make remand more likely.

British research in the 1980s reported high rates of psychiatric morbidity in remand prisons,9 10 but these studies may have underestimated the problem as they were retrospective case note studies relying on diagnoses by prison medical staff. More recent studies of remand prisoners have confirmed the high rate of mental disorder in this population.11 12 However, it is still unclear how much of the mental disorder is recognised by prison healthcare services.

We conducted this study to determine the prevalence of mental disorder in remand prisoners at the time of their reception into prison and the efficacy of the prison reception screen in detecting this. It is the first part of a longitudinal study of a cohort of new remand prisoners who will be followed throughout the course of their remand.

Subjects and methods

Durham prison is a local male remand and short term sentence prison with a capacity of about 640. It receives nearly all men aged 21 years and over who are remanded from courts in Tyneside, Northumberland, Cumbria, and County Durham. The geography and population in this large catchment area varies considerably, although most remands come from courts covering areas of urban deprivation and high unemployment. The prison population, in common with other remand prisons, is very fluid, with large numbers of prisoners being received and discharged every day.

Current national prison policy is for all new prisoners to be screened by a hospital officer on the day of their arrival and subsequently by a prison medical officer. The hospital officer uses the F2169 first reception health screen, a standard prison questionnaire used to screen for physical and mental health problems as well as substance misuse. The prison doctor usually sees the prisoner the next working day as part of the prison induction process. We integrated our study into the induction process and so interviewed nearly all subjects within one working day of their reception into prison, usually immediately after the prison doctor's health screen.

SUBJECTS

One of us (LB or DM) approached all unconvicted men remanded to Durham prison from 1 October 1995 to 30 April 1996 and asked them to take part in the study. Prisoners were assured about confidentiality, and we obtained written consent before the interview.

We used a semistructured interview designed specifically for the study. This incorporated well validated psychiatric instruments used by Gunn et al6 11 to allow a direct comparison of results. We used the schedule for affective disorders and schizophrenia (lifetime version) to detect and classify current and lifetime mental disorders,13 the CAGE questionnaire to assess problem drinking,14 and the severity of dependency questionnaire to quantify levels of drug abuse and dependence.15 Self reported levels of alcohol and drug consumption were also recorded. We measured intelligence quotients (IQ) with the quick test.16

If personality disorder was suspected, we asked about more specific areas of functioning, and, if appropriate, made a diagnosis using DSM-IV criteria (Diagnostic and Statistics Manual of the American Psychiatric Association). The diagnosis was recorded together with the ICD-10 equivalent. If dysfunctional personality traits were present but DSM criteria for personality disorder were not met this was recorded as “vulnerability.” Interviews lasted between 20 minutes and one hour. In a few cases where serious mental disorder was suspected but inadequate information was obtained at interview, we sought information from other sources within the prison.

The interviewer recorded the diagnosis and appropriate treatment, and a random sample of the cases was reviewed by a steering committee of senior academic psychiatrists. Inter-rater reliability was monitored during a pilot and regularly throughout the study. A total of 116 prisoners were interviewed by one researcher in the presence of the other. The researchers independently recorded lifetime diagnoses and psychiatric management required, and the agreement was measured by calculating a (kappa) coefficient.17 After each interview the prisoner's medical records were examined and the prison health screen inspected.

Results

During the study 606 men were remanded to Durham prison (27 more than once). Thirty seven men returned to court the morning after admission and did not return to the prison. Of the remaining 569 men, 549 (96%) consented to be interviewed, 19 refused, and one was unfit for interview. In all, 528 (96%) interviews were fully completed; 21 were only partly completed because of language barriers, mental state disturbance, and situational constraints. We recorded IQ scores for 441 men. Poor concentration, agitation, language difficulties, or other adverse factors made the testing unreliable in the remainder.

In the 116 jointly rated interviews, 51 lifetime diagnoses of mental disorder and 184 separate substance misuse diagnoses were recorded by either one or both raters. Diagnostic agreement occurred in 216 of these ((kappa) = 0.902). Most disagreements were over diagnoses of personality disorder ((kappa) = 0.761) and adjustment disorder ((kappa) = 0.645). The two researchers agreed on all 15 lifetime diagnoses of psychosis made during inter-rater interviews ((kappa) = 1.0).

Table 1 summarises the characteristics of the population. Most of the men (378, 66%) were white and aged 30 years or under. Almost 80% of the population were unemployed or on sickness benefit. Of the 441 tested, 389 (88%) had an IQ score below the general population mean, and 57 (13%) scored 70 or less.

View this table:
Table 1

Demographic details of 569 unconvicted remand prisoners. Values are numbers (percentages) of prisoners unless stated otherwise

PREVALENCE OF MENTAL DISORDER

Mental disorder was present in 148 (26%, 95% confidence interval 22% to 30%) of the 569 men at the time of reception into prison (table 22); a further 22 men had a history of mental disorder but no current symptoms. Lifetime rates were 7% (6% to 8%) for psychosis and for non-psychotic mood disorders. If diagnoses of substance abuse or dependency were included the number of men with current mental disorders rose to 354 (62%, 60% to 64%) and with lifetime disorder to 404 (71%, 69% to 73%). In addition to those with a diagnosis of personality disorder, 68 men (12%, 11% to 13%) were judged to have significant personality vulnerabilities.

View this table:
Table 2

Prevalence of current mental disorder in 569 unconvicted remand prisoners

In all, 168 men required some form of psychiatric input (table 3), and 50 needed urgent attention. Of these 50, 16 required immediate transfer to an outside psychiatric hospital (14 psychotic, one severely depressed, and one mentally retarded), five needed further assessment in the prison hospital (most of whom would probably require hospital transfer), and 29 needed prison hospital placement.

View this table:
Table 1

Psychiatric management required in 569 unconvicted remand prisoners

PRISON RECEPTION SCREENING

Screening by the hospital officers and prison medical officers identified current mental disorder in 52 (9%) men. Forty eight (8%) were initially placed on the hospital wing, although in 21 cases this was for reasons other than mental health.

Of the 148 men we identified with current mental disorder (excluding substance abuse or dependence), 34 (23%) were also identified by the prison medical screen (difference = 0.169, 95% confidence interval 0.132 to 0.206). Only six of the 24 men who were acutely psychotic had any abnormality of mental state identified by the prison screening. Mental disorder was said to be present in a further 18 men in whom we found no evidence of this.

Of the 50 men we judged to require urgent intervention, 17 were picked up by the screen and placed in the prison hospital; three others were placed there for non-psychiatric reasons. The remaining 30 men (including 16 who were acutely psychotic) were placed in ordinary cells.

Discussion

Our findings are based on inmates of one prison with a large catchment area, but there is no reason to believe that Durham prison differs significantly from other remand prisons. The reception screen is used nationally, and the demands on prison health staff are similar. We believe, therefore, that the results of this study can be generalised throughout the remand population.

Twenty six per cent of new prisoners had a current mental disorder, nearly one third of whom had a serious disorder. Most of these disorders were undetected and so untreated. Although these men may have had their mental disorder identified and treated later on in their remand, our impression from our initial follow up is that this is unlikely. In a busy remand prison abnormal behaviour is often tolerated or perceived as a discipline problem and dealt with punitively, while the “quietly mad” are ignored.

Failure to identify psychiatric illness in men entering remand prisons means that the opportunity for treatment is lost. This is important since many of these men might not otherwise come to the attention of healthcare services and the severely psychotic inmates may be at high risk of suicide.

The reception screening undertaken by the prison service is neither sensitive nor specific for detecting mental disorder. Reasons for this may include lack of appropriately experienced or trained staff to conduct the screening, a questionnaire of doubtful validity, time constraints, a highly mobile population, and a perception among prisoners that prison healthcare staff do not have their best interests as a priority.

Health screening on reception into prison should provide an important opportunity to detect mental disorder and provide prompt and appropriate treatment that, particularly in the case of remand prisoners, could be continued in the community. However, to achieve this screening will need to become more effective. If more psychiatric morbidity is identified, the availability of psychiatric bed may become a problem. Beds need to be available at all levels of security to take patients rapidly from prison or as part of a court diversion scheme. Psychiatric beds in the Northern region, as in the rest of the country, are in short supply and often impossible to find at short notice. Our results provide some indication of the extent of the shortfall.

Footnotes

  • Funding Prison Service and Northern and Yorkshire Regional Health Authority.

  • Conflict or interest None.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.

THIS WEEK'S POLL