International Journal of Law and Psychiatry 32 (2009) 115–119

Contents lists available at ScienceDirect

International Journal of Law and Psychiatry

Psychiatric disorders and personality characteristics of prisoners at regular

prison wards
Erik Bulten a,⁎, Henk Nijman b, Cees van der Staak b
a
Pompe Foundation, Forensic Psychiatric Centre, Nijmegen, The Netherlands
b
Radboud University of Nijmegen, The Netherlands

a r t i c l e i n f o a b s t r a c t

Keywords: Background: Dutch correctional officers are trained to observe prisoners with severe mental disorder. This
Psychiatric disorder ‘behavioural’ approach is assumed to detect psychiatrically disordered prisoners with striking symptoms. On
Personality the basis of this screening procedure about 10% of the Dutch prison population is classified as needing special
Prison care or control. In the current study, what psychopathology can still be found among the remaining 90%
Screening
prisoners residing at regular wards is investigated and which personality traits characterize them. When the
Prevalence
prevalence of major mental illness would still turn out to be high in this group, the question arises whether
the current, rather unstandardized, way of screening prisoners is sufficient.
Aims: To assess the prevalence of psychiatric disorders among adult Dutch prisoners on regular wards, and to
determine the concurrent and convergent validity of self-report measures and their screening characteristics
in such a sample.
Method: 191 randomly selected prisoners admitted to the general wards were administered a number of tests
(SCL-90, NEO-PI-R, MINI) during the first weeks of their incarceration.
Results: Including substance abuse, 57% of the participants suffered from one or more Axis I disorders. About
seven out of ten detainees with psychopathology as assessed with the MINI did not receive professional help.
The concurrent and convergent validity of the SCL-90 and the NEO-PI-R turned out to be reasonable. The
predictive validity of self-report measures in detecting prisoners with an Axis I disorders or suicide risks was
moderate.
Conclusions: The Dutch ‘behavioural approach’ seems to be quite accurate in detecting prisoners with
psychotic disorders. Most prisoners with other mental disorders on regular wards, however, did not receive
professional help. The current study suggests that self report scales such as the SCL-90 and the NEO-PI-R may
be helpful in screening detainees on important DSM-IV disorders.
© 2009 Elsevier Ltd. All rights reserved.

1. Introduction
Psychiatric disorders are prevalent among prison inmates. Fazel
and Danesh (2002) systematically reviewed 62 studies on the
prevalence of major psychiatric disorders among delinquents: 10%
suffer from major depression and about 4% have psychotic illnesses.
Apart from that about 65% have a personality disorder, which
commonly is an antisocial personality disorder (i.e., 47%, on average).
In the studies included in the review of Fazel and Danesh (2002), the
diagnoses mainly were obtained by means of structured diagnostic
interviews performed by clinicians, such as the Diagnostic Interview
Schedule (DIS), the Structured Clinical Interview for the Diagnostic
and Statistical Manual (SCID), and the Present State Examination — 10
(PSE). In an earlier study among young prisoners (age range from 18 to
23) in the Netherlands, the first author of this paper (Bulten, 1998)
⁎ Corresponding author. Radboud University, Academic Centre Social Sciences (K.1.14),
Mailbox nr. 9104, 6500 HE Nijmegen, The Netherlands.
E-mail address: e.bulten@acsw.ru.nl (E. Bulten).
used the DIS. Again, the most prevalent psychiatric diagnosis turned
out to be antisocial personality disorder (i.e., 42%). Almost one out of
ten of the young adult prisoners had a life-time diagnosis of a
psychotic disorder (schizophrenia or a schizophreniform disorder),
and about the same proportion suffered from depression. Further-
more, it was found that 9% suffered from obsessive compulsive
symptoms.
What to do with these high prevalences? Due to their legal or
professional standards jails and prisons have the obligation to provide
adequate mental healthcare for their inmates. Early detection is
important, but are prisoners with a severe mental disorder indeed
detected at an early stage of their incarceration? Whereas in other
countries screening tools like the Brief Jail Mental Health Screen
(Steadman et al., 2005) are used, the psychiatric screening procedures
are not standardized in the Netherlands. On the other hand, Dutch
correctional officers are trained to observe and detect prisoners with
severe mental disorder. This ‘behavioural’ approach is assumed to detect
psychiatrically disordered prisoners with striking symptoms. Symptoms
which attract attention are odd, strange, or overtly depressive behaviour,

0160-2527/$ – see front matter © 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijlp.2009.01.007
116 E. Bulten et al. / International Journal of Law and Psychiatry 32 (2009) 115–119
as well as disruptive and aggressive behaviour. As a result of this
“behavioural” approach about 10% of the Dutch prison population is
classified as needing special and intensive care or control, and for this
reason they are admitted to specialized prison departments. But what
about the remaining 90% of detainees who are incarcerated at regular
prison wards? During this study the Dutch prison system contained
about 13,500 detainees, and therefore more than 12,000 inmates
resided at regular units in the Netherlands. What is the prevalence of
mental disorders in this group of prisoners? When the prevalence of
major mental illness would still turn out to be high in this group, the
question on how to screen prisoners becomes important.
The use of self-report measures to screen delinquents on
psychiatric symptoms may have many (practical) advantages (e.g.,
time-efficiency), compared to the more time-consuming administra-
tion of structured interviews by professional caregivers. Routine
health screening on admission to the prison by means of self-reports
would, if valid, present “(…) a unique opportunity to identify health
needs at an early stage” (Morrison & Gilchrist, 2001). How valid are
self-reports of psychiatric complaints by criminal subjects at regular
prison wards? What is the self-reported psychopathology among
these ‘ordinary’ prisoners and what personality traits characterize
them?
In the present study we sought to gain more insight in the
prevalence of psychiatric disorders among the ±90% of adult Dutch
prisoners who are detained at regular prison wards. Apart from that,
we tried to investigate the concurrent and convergent validity of self-
report measures (in comparison to structured diagnostic interviews)
and their screening characteristics in such a sample.
2. Methods
2.1. Participants
A cohort of consecutively admitted prisoners to the general wards
of the correctional institution in Vught, the Netherlands, was
considered for participation in the current study. In the correctional
institution in Vught both convicted prisoners, as well as accused
persons waiting for their case to come on trial reside. As only
detainees of regular, non-specialized wards were asked to participate,
prison inmates with obvious psychopathology and behavioural
problems are likely to have been excluded automatically from the
study. That is to say, prisoners suffering from acute psychopathology
(e.g., florid psychotic symptoms) or prisoners who display severe
disruptive behaviour are generally admitted to one of the specialized
wards of the Dutch correctional system.
Out of a total of 309 randomly selected prisoners admitted to the
general wards, 41 were judged to be unable to participate, mostly due
to language problems. Some of those 41 were transferred to one of the
special units or wards after all and could therefore not be included in
the sample. The remaining 268 subjects were asked for their
cooperation in the study. Fifty of these 268 prisoners refused (19%),
leaving 218 subjects who provided written informed consent. Of the
218 prisoners that consented, however, another 27 did not complete
(all) tests in the end (10%), mainly because they were released shortly
after having given approval to the study. In other words, 71% of the
268 detainees who were initially asked for informed consent, finally
were able to give full cooperation to the present study, leaving 191
prisoners in the sample.
2.2. Materials and methods
The 191 prisoners were administered a number of tests during the
first weeks of their incarceration in the correctional institution in
Vught. To begin with, the MINI International Neuropsychiatric Inter-
view was administered (MINI; see Lecrubier, Sheenan, Weiler,
Amorium, Bonora et al., 1997; Sheenan, Lecrubier, Harnett Sheenan,
Janavs, Weiller, et al., 1997). This instrument assesses the presence of
17 DSM-III-R psychiatric disorders, among which antisocial person-
ality disorder. Other Axis II diagnoses, however, are not assessed with
the MINI. In the current study, the MINI was administered by trained
psychologists.
Secondly, patients completed the Dutch version of the Symptom
Checklist — 90 (SCL-90; authorized Dutch version of Arrindell &
Ettema, 1986). This 90-item checklist measures to what extent the
respondent suffered from psychiatric complaints during the week
prior to completion of the scale. To be more precise, the extent is
assessed to which the respondent suffered from “Anxiety” (10 items),
“Agoraphobia” (7 items), “Depression” (16 items), “Somatic com-
plaints” (12 items), “Suspicion and interpersonal sensitivity” (18 items),
“Insufficient thinking and behaviour” (9 items), “Sleeping problems”
(3 items) and “Anger-hostility” (6 items).
Apart from the MINI and the SCL-90, subjects were asked to
complete the NEO-PI-R (Costa & McCrae, 1992) which is a 240 item
self-report instrument aiming at measuring five important dimen-
sions of personality, namely Neuroticism, Extraversion, Openness,
Agreeableness and Conscientiousness. The five domains (factors)
measured by the NEO-PI-R provide a general description of person-
ality, while the facet subscales allow more detailed analyses.
2.3. Statistics
By means of descriptive statistics, the prevalence rates of psy-
chiatric disorders, as assessed with the structured MINI interview,
were investigated. The focus of this descriptive investigation was on
antisocial personality disorder, depression and suicide risks, ADHD,
and anxiety disorders. Although the prevalence was expected to be
rather low in this selected prison sample, it was also investigated
whether any subjects with major Axis-I disorders such as psychosis or
bipolar disorder could still be detected.
Following the analyses of the MINI results, we studied the extent in
which the diagnoses reached by clinicians on the basis of the
structured interviews, were corroborated by the self-report measures.
For this, the associations between the MINI results and the self-
reported psychopathology (i.e., the scores on the SCL-90) and
personality characteristics (i.e., the scores on the NEO-PI-R), were
explored by means of t-tests. The screening potential of some self-
report measures were explored by means of ROC-curves, their
sensitivity and specificity, as well as their positive predictive power
(PPP) and negative predictive power (NPP).
3. Results
3.1. Age and criminal histories
The mean age of the 191 male participants was 30.4 years (s.d. = 8.9),
with a minimum age of 18 and a maximum of 59 years. Crimes against
property were the most common reasons for incarceration (i.e., 37%),
followed by violent crimes (23%), homicide offences (15%), drugs related
crimes (14%), and sexual offences (6%). The majority of the 191 prisoners
(i.e., 68%) had already served time earlier for other crimes. The average
number of earlier imprisonments periods was 2.6 (s.d.= 3.3) for the
entire sample.
No significant differences in age [t(266) = 0.62, p = 0.54] or type of
offence [χ2 (5) = 3.5, p = 0.63] were found between the subjects who
gave full participation (n = 191) and subjects who refused participation
or dropped out during the test period (n = 77).
3.2. Prevalence of psychiatric disorders
Childhood diagnoses of ADHD (38%), antisocial personality
disorders (ASP; 37%), and drug and alcohol dependence (30% and
28%, respectively), were the most prevalent disorders in the sample. A
 E. Bulten et al. / International Journal of Law and Psychiatry 32 (2009) 115–119 117

current diagnosis of psychosis was only found in one of the 191 For patients suffering from anxiety symptoms (i.e., phobias, panic
subjects, and no diagnoses of mania or bipolar disorder were detected. disorders, obsessive compulsive disorders, and PTSD) the anticipated
Suicide ideation (18%) and major depression (9%), however, were differences were also found. In other words, patients with an anxiety
rather prevalent, and in 7 cases the risk of suicide (4%) was judged to disorder scored higher on the “Anxiety” [t(183) = 3.3, p b 0.01] and
be high. Furthermore, anxiety disorders, such as phobias, panic “Agoraphobia” [t(183) = 2.5, p b 0.01] subscales of the SCL-90 as
disorders, obsessive compulsive disorders, and Post Traumatic Stress compared to prisoners without such a diagnosis.
Disorders (PTSD) were prevalent (12% in total). Of the total of 22 However, the scores on all other SCL-90 subscales turned out also
patients with anxiety symptoms, 9 (5% of the total sample) were to be elevated in this group, except for the SCL-90 “Hostility”
diagnosed to have a Post Traumatic Stress Disorder (PTSD) (Table 1). dimension. Subjects with anxiety related disorders also scored higher
One out of five participants (22%) had a current Axis I disorder on “Neuroticism” of the NEO-PI-R [t(183) = 7.0, p b 0.01], but also lower
when substance abuse was excluded as the single diagnosis, and three on “Conscientiousness” [t(187) =3.9, pb 0.01] and “Extraversion” [t(183)=
out of five (61%) if substance abuse disorders were included. Four out 2.2, pb 0.05].
of five (82%) had some form of a life time Axis I disorder. Prisoners with a history of ADHD did not turn out to have elevated
Almost half of the sample (47%) was detained in a house of remand, “anger-hostility” and “insufficiency of thinking” scores on the SCL-90.
the others (53%) were sentenced prisoners. Accused detainees, Neither did these prisoners report lower scores on “Conscientious-
waiting for their case to come on trial showed more depressive ness” on the NEO-PI-R. On “Agreeableness” participants with a history
complaints on the SCL-90 ‘Depression’ subscale [t(186) = 3.7, p b 0.01]. of ADHD scored lower ([t(187) = 2.0, p b 0.05].
These complaints did not result in significant differences between the When, however, only the limited number of prisoners for whom
two groups as far as major depression was concerned, as assessed with the ADHD diagnosis was still present at the time of the research (n = 8)
the MINI. There appeared to be, however, a significant association with was included, several differences with the other participants were
suicide-risks: the subjects in a house of remand expressed higher found. The 8 patients still suffering from ADHD had markedly higher
suicide risks [χ2 (3) = 12.1, p b 0.01]. Interestingly, in the house of “Anger-hostility” and “Insufficiency of thinking” scores, and lower
remands 22% of the subjects had an antisocial personality disorder, scores on “Conscientiousness” ([t(187) = 4.3, p b 0.01]; [t(187) = 3.4,
compared to 50% in prison [χ2 (1) = 15.7, p b 0.01]. p b 0.01]; [t(183) = 5.4, p b 0.05], respectively). Besides these expected
outcomes, subjects with a current MINI-ADHD diagnosis showed high
scores on the NEO-PI-R “Neuroticism” subscale [t(183) = 4.4, p b 0.01],
3.3. Associations between Axis I disorders and self-reported symptomatology and the SCL-90 “Depression” [t(187) = 4.2, p b 0.01], “Anxiety” [t(187) =
3.1, p b 0.05] and “Suspicion and interpersonal sensitivity” [t(187) = 3.2,
As far as depressive disorders were concerned, anticipated differ- p b 0.05] subscales.
ences on the SCL-90 and the NEO-PI-R were found. To be more precise, Overall, in 18% of the prisoners suicide ideations in mild to severe
prisoners diagnosed as suffering from depression on the basis of the forms were found. In the subgroup diagnosed with a current
MINI clearly exhibited higher scores on “Depression” [t(183) = 5.9, depression, 39% appeared to have suicidal thoughts [χ2 (1) = 5.6,
p b 0.01] and “Anxiety” subscales [t(183) = 5.1, p b 0.01] of the SCL-90, p = 0.02]. The MINI provides four risk levels of suicide: no risk, low risk,
but also on all other SCL-90 subscales as compared to prisoners without medium risk, and high risk. In line with what could be expected, the
depression. Prisoners with a current depression expressed higher scores on the “Depression” subscale of the SCL-90 were strongly
“Neuroticism” scores on the NEO-PI-R [t(183) = 5.1, p b 0.01]. Apart from associated with the level of suicide risk [F (188) = 8.2, p b 0.01].
that lower scores on the “Conscientiousness” [t(183) = 3.4, p b 0.01] and However, the same seems to be true for other symptom dimensions,
“Extraversion” [t(183) = 2.8, p b 0.05] subscales of the NEO-PI-R were such as “Anxiety” [F (188) = 7.4, p b 0.01] and “Insufficiency of thinking
found. and behaviour” [F (188) = 7.4, p b 0.01] as measured with the SCL-90,
and the “Neuroticism” scale of the NEO-PI-R. [F (184) = 6.9, p b 0.01].
Table 1
Most prevalent psychiatric disorders in the sample of 191 prisoners, as assessed with 3.4. Associations between antisocial personality and self-reported
the MINI
symptomatology
Psychiatric disorder Assessment period Frequency %
ADHD (childhood) Life-time diagnosis 72 38% ASP turned out not to be statistically associated with higher
ADHD adult Current diagnosis 8 4% “Anger-hostility scores” and “Suspicion and interpersonal sensitivity”
Antisocial personality Current diagnosis 70 37% on the SCL-90. Possibly, this lack of association is due to the fact that
disorder (ASP)
Drug dependence Past 12 months 57 30%
the SCL-90 measures complaints during the week before completion,
Alcohol Past 12 months 53 28% and not personality traits. In line with expectations, prisoners with
Suicide ideation Past month 35 18% ASP did score lower on “Agreeableness” in comparison to the other
High risk 7 4% inmates [t(183) = 6.6, p b 0.01]. The anticipated lower score on “Con-
Moderate risk 10 5%
scientiousness” [t(183) = 3.1, p b 0.01] for prisoners with ASP was also
Low risk 18 9%
Anxiety disorders Current diagnosis 22 12% confirmed. Exploratory analyses of the facets of the NEO-PI-R
Depression Current diagnosis 18 9% dimensions “Agreeableness” and “Conscientiousness” in relation to
Psychosis Current diagnosis 1 1% ASP were performed. Persons with ASP showed lower scores on all
Psychosis Life-time diagnosis 8 4% facets of the “Agreeableness” subscale, such as “Trust”, “Straightfor-
≥1 Axis I disorder (without Current diagnosis 42 22%
wardness”, “Altruism”, “Compliance” and “Modesty” (all p's b 0.05),
substance abuse as the
single diagnosis) except for “Tender-Mindedness”. Three of the six facets of “Con-
≥1 Axis I disorder (without Life-time diagnosis 117 61% scientiousness” showed the same pattern; ASP prisoners scored lower
substance abuse as the on “Dutifulness”, “Self-discipline” and “Deliberation”.
single diagnosis)
≥1 Axis I disorder (substance Current diagnosis 108 57%
abuse included) 3.5. Predictive validity of self-report measurement
≥1 Axis I disorder (substance Life-time diagnosis 156 82%
abuse included) A Receiver Operating Characteristics (ROC) analysis for differen-
≥1 Axis I disorder (substance Current diagnosis 130 68% tiating prisoners with an Axis I disorder vs. prisoners without an Axis I
abuse included and/or ASP
disorder (excluding substance abuse) on the basis of the overall score
118 E. Bulten et al. / International Journal of Law and Psychiatry 32 (2009) 115–119

Table 2 of the study. When substance abuse was included 57% of the
Sensitivity, specificity, PPP and NPP of tests screening for Axis I disorders (excluding participants suffered from one or more Axis I disorder.
substance abuse) and suicide risk
Based on international and Dutch studies, a prevalence of
Test Cut off Sensitivity Specificity PPP NPP schizophrenia and mania between 4 and 6% could be expected
Axis I disorder (excluding substance abuse) (Bulten, 1998; van Panhuis, 1997; Fazel & Danesh, 2002). In the current
SCL-90 Psychoneuroticism 132 0.95 0.65 0.43 0.98 sample 0.5% (n = 1) was diagnosed with such a disorder. The relative
(AUC = 0.86 (95%CI: 0.79–0.93) 142 0.76 0.75 0.53 0.91
absence of these major Axis I disorders in the current sample suggests
SCL-90 Depression 26 0.88 0.65 0.42 0.95
(AUC = 0.83 (95%CI: 0.76–0.90) 30 0.83 0.75 0.51 0.94 that these psychiatric conditions are detected well by the correctional
NEO-PI-R Neuroticism 131 0.83 0.58 0.37 0.92 officers and the medical, psychiatric and psychological staff of the
(AUC = 0.82 (95%CI: 0.74–0.90) 142 0.81 0.78 0.52 0.93 prison that was studied. Although these results are likely to have been
influenced by selection bias due to drop out from the study and
Suicide risk (moderate to high)
SCL-90 Psychoneuroticism 132 0.82 0.57 0.16 0.97
refusals to participate, the ‘behavioural approach’ seems to be rather
(AUC = 0.79 (95%CI: 0.69–0.89) 142 0.76 0.65 0.18 0.97 effective in detecting major Axis I disorders.
SCL-90 Depression 26 0.88 0.58 0.17 0.98 However, the detection of prisoners with a high suicide risk may be
(AUC = 0.78 (95%CI: 0.65–0.90) 30 0.76 0.66 0.18 0.97 more troublesome. Although none of the prisoners committed suicide
PPP = Positive predictive power; NPP = negative predictive power. during the study period, 4% (n = 7) were judged to have a high risk on
attempting this according to the standards of the MINI. Five of these
seven subjects did receive care from the prison mental health service,
of the SCL-90 (“Psychoneuroticism”) showed an area under the curve but the remaining two had not been detected as subjects needing help
(AUC) of 0.86 (95% confidential interval CI: 0.79–0.93; p b 0.01). For the or care. Three out of every ten prisoners with a medium suicide risk
NEO-PI-R “Neuroticism” scale the AUC value was 0.82. In Table 2, AUC (5% of the total sample) did not receive professional care.
values, sensitivity, specificity, and PPP and NPP of the SCL-90 total In their review on 62 studies concerning a total of 18,530 male
score, the SCL-90 “Depression” subscale, and the NEO-PI-R “Neuroti- prisoners (mean age: 29 years), Fazel and Danesh (2002) found a
cism” scale are presented. As far as the prediction of moderate to high prevalence of 10% (CI 9.0–11.0) of major depression, which is highly
suicide risk is concerned, both the SCL-90 “Psychoneuroticism” and similar to the prevalence in this study. Their review did not provide
“Depression” scales showed moderate AUC values as an overall statistics about anxiety disorders, which was 12% in this sample.
measure of predictive validity (i.e., AUC values of 0.79 and 0.78, Differences were found with regard to ASP (37% vs. 47%) and psychotic
respectively). illness (1% vs. 4%). These differences probably can be explained by the
For screening Axis I disorders in prison the SCL-90 “Psychoneur- fact that prisoners in special care units and special security wards
oticism” scale, the “Depression” subscale, and the NEO-PI-R “Neuroti- (about 10% of the Dutch prison population) were excluded from the
cism” scale seem to have rather high predictive accuracy. The first two current study. Most of the detainees in the special care wards have
tests can also be used for detecting suicide risk. Using tests suitable for severe psychopathology, and the prevalence of psychotic disorders is
both purposes (screening for Axis I disorder and suicide risk) seems to particularly high among them (35 to 40%), whereas prisoners in the
be advisable. Although the model with the SCL-90 “Depression” Dutch special security wards in most cases show severe personality
subscale resembles the model with the total score on the SCL-90, the disorders, among which high prevalences of ASP are found (Bulten,
“Depression” subscale contains only 16 compared to the 90 items of 2001).
the total scale, which may make the “Depression” subscale particu- This study was conducted in a large detention centre in the south
larly easy to use in clinical practice. of the Netherlands. The sample was not randomly drawn from the
The “Neuroticism” subscale did not turn out to be accurate at total ‘regular’ prison or jail population. However, admission to one of
detecting suicide risk (AUC = 0.71; 95%CI = 0.58–0.83). On the other the regular wards of this prison (from which the subjects were
hand, the “Agreeableness” subscale, one of the subscales of the NEO- selected) takes place on formal, objective grounds, not on personality
PI-R demonstrated acceptable statistics in predicting the absence of characteristics or other selection criteria. For this reason, the results in
ASP: AUC = 0.77 (95%CI = 0.70–0.84). our opinion are likely to be illustrative for other regular wards of other
In clinical practice, the population base rate of mental disorders, Dutch jails or prisons. The question remains, however, to what extent
personality disorders and of suicide risk influences the performance of our Dutch results will be comparable to those of prison systems in
a test. The positive predictive power (PPP) and negative predictive other countries, and to what extent selection bias may have influenced
power (NPP) are related to the (expected) prevalence in a specific the results. The non participants consisted of subjects who refused to
population and are also presented in Table 2. Using higher cut off participate and subject who dropped out for formal reasons because
points increases the PPP and specificity of a test, lower cut off points they were released shortly after given approval to the study. As the
increase NPP and sensitivity. The choice for cut off points in detecting distribution of psychiatric disorders in the group of non participants is
mental disorder or suicide risk (in Table 2 some examples are shown) unknown, selection bias cannot be ruled out.
depends on the specific purpose of the test within the specific Further it should be noted that the MINI diagnosis of childhood
circumstances in the prison, like for instance procedures after the first ADHD could only be based on what prisoners themselves said about
screening and the mental health policy in the prison. their behaviour during childhood. As collateral information from
parents and school teachers was lacking, it is possible that in some
4. Discussion cases the symptoms reported by the prisoners who were judged to
have suffered ADHD in childhood, reflected other problems like
Even though prisoners residing at specialized correctional wards conduct disorder. Nevertheless, the current findings again stress a
were excluded, still rather high prevalences of psychiatric disorders high degree of early ‘childhood’ vulnerability in many of the prisoners
were found among the Dutch prisoners at regular wards. Current at regular units.
anxiety disorders, depression, ADHD and substance abuse showed The concurrent and convergent validity of the SCL-90 and the NEO-
prevalences of respectively 12%, 9%, 4% and 46%. ASP was diagnosed in PI-R with the MINI as a measure of mental disorder seemed to be
37% of the sample, with a striking difference between participants in a reasonable. The SCL-90 focuses on actual psychiatric complaints and
house of remand (22%) or prison (50%). Well over a third of the several subscales differentiate between prisoners with and without an
inmates (i.e., 72 prisoners or 38%) appeared to have had ADHD in Axis I disorder. Although the NEO-PI-R is a self-report test aiming at
childhood, and in 8 prisoners (4%) ADHD was still present at the time measuring dimensions of personality, the NEO-PI-R “Neuroticism”
 E. Bulten et al. / International Journal of Law and Psychiatry 32 (2009) 115–119
dimension in particular was found to be strongly linked with mental
disorder. Several NEO-PI-R dimensions and facets were connected
with the Axis II diagnosis antisocial personality disorder. Prisoners
with an antisocial personality disorder, for instance, showed lower
scores on “Agreeableness”. The scores on several facets illustrate that
participants with this personality disorder are less trustful, straight-
forward, altruistic, compliant, modest, dutiful, self-disciplined and
deliberate.
The participants were detained in regular prison wards. This
however does not exclude them from individual mental health care.
Almost one of seven subjects (15%), received some kind of professional
psychological or psychiatric care. Of the 22% (n = 42) of prisoners with
an Axis I disorder (substance abuse excluded), only a third (n = 13 or
31%) received mental health care. In other words: 69% of the detainees
with psychopathology as assessed with the MINI did not receive
professional help. Detection of these, in most cases vulnerable,
prisoners with a mental disorder in an early stage is an important
issue. Less supervision by staff at regular units (and therefore a
decreasing quality of the so called “behavioural approach”), and more
emphasis on self supportiveness, may strongly influence prisoners
who lack sufficient coping skills due to mental disorder. In some cases
psychiatric support may not be necessary, in other cases transmission
to a more specialized ward is needed after all. Valid but easy to use
screening methods seem indispensable.
Because of the consistency between the diagnoses reached by
psychologists on the basis of the MINI and the self-reports of the
prisoners (i.e. SCL-90 and NEO-PI-R), it seems that conditions such as
depression and anxiety seem to be detectable rather well by means of
self-report inventories. The predictive validity of several subscales
proved to be reasonable. The predictive validity of the SCL-90 total
score and the “Depression” subscale in detecting prisoners with a high
suicide risk was moderate.
Detection of Axis II disorders such as ASP also seems to be possible
to a certain extent on the basis of self-reports. As anticipated, subjects
with ASP did score differently on the domains “Agreeableness” and
“Conscientiousness” and a number of underlying facets of these NEO-
PI.R domains. These lower scores on “Conscientiousness” and
“Agreeableness” by prisoners with ASP are in line with earlier
research findings (Trull, 2001; Saulsman & Page, 2004) obtained in a
non prison population.
The ‘predictive’ validity of the subscales found in the current study,
however, is based on a cross-sectional design. The predictive validity
of these screening tools for the mental condition in the longer run can
only be established in a longitudinal study. Nevertheless, screening by
observation, as well as more standardized screening by for instance
self-report measures, in our opinion, should be part of the total mental
health policy within prisons. Although self report screening tools can
be time-efficient, positive ‘hits’ require follow-up assessments and
appropriate treatment when necessary. The effects of systematic
psychological screening highly depend on the possibilities to inter-
119
vene in cases psychiatric problems are detected. Both the screening
procedure and intervening appropriately will have rather substantial
financial implications but may be worthwhile if this would improve
the mental health of prisoners and reduce recidivism risks after
imprisonment. For this reason, more research on the associations
between providing mental health care to prisoners and recidivism is
needed.
In conclusion, the Dutch ‘behavioural approach’ seems to be quite
accurate in detecting prisoners with severe psychiatric disorders such
as psychotic disorders and mania. The current study further suggests
that self-report scales such as the SCL-90 and the NEO-PI-R, can be
helpful in screening for some other important DSM-IV disorders.
However, more research on the sensitivity and specificity of the
instruments for the prison system is needed.
Acknowledgments
The authors are grateful to Mieke Schoenmakers, Ester Beekmans
and Wendy van Bergen for their assistance in the current study. The
research upon which this publication is based was supported by
funding from the Ministry of Justice (Dienst Justitiële Inrichtingen).
The views expressed in this publication are those of the authors and
not necessarily those of the Ministry of Justice.
References
Arrindell, W. A., & Ettema, J. H. M. (1986). SCL-90: Handleiding bij een multidimensionele
psychopathologie-indicator. (SCL-90: Manual for a multidimensional indicator of
psychopathology) Lisse, The Netherlands: Swets & Zeitlinger.
Bulten, B.H. (1998). Gevangen tussen straf en zorg [Captured between punishment and
care]. Ph. D. thesis, Vrije Universiteit, Amsterdam.
Bulten, B. H. (2001). Too mad to handle. Sancties, 2, 69−83.
Costa, P. T., & McCrae, R. R. (1992). Revised NEO Personality Inventory (NEO-PI-R) and the
Five Factor Inventory (NEO-FFI): Professional manual, Odessa, Florida: Psychological
Assessment Resources Inc.
Fazel, S., & Danesh, J. (2002). Serious mental disorder in 23000 prisoners: A systematic
review of 62 surveys. The Lancet, 359, 545−550.
Lecrubier, Y., Sheenan, D., Weiler, E., Amorium, P., Bonora, I., Sheenan, K., Janvan, J., &
Dunbar, G. (1997). The Mini International Neuropsychiatric Interview (MINI). A
short diagnostic structured interview: Reliability and according to the CIDI. Euro-
pean Psychiatry, 12, 224−231.
Morrison, D. S., & Gilchrist, G. (2001). Prison admission health screening as a measure of
health needs. Health Bull (Edinb), 59, 114−119.
Saulsman, l. M., & Page, A. (2004). The five-factor model and personality disorder
empirical literature: A meta-analytic review. Clinical Psychology Review, 23,
1005−1085.
Sheenan, D. V., Lecrubier, Y., Harnett Sheenan, K., Janavs, J., Weiller, E., Keskiner, A.,
Schinka, J., Knapp, E., Sheenan, M. F., & Dunbar, G. C. (1997). The validity of the Mini
International Neuropsychiatric Interview (MINI) according to the SCID-P and its
reliability. European Psychiatry, 12, 232−241.
Steadman, H. J., Scott, J. E., Osher, F., Agnese, T. K., & Robbins, P. C. (2005). Validation of
the Brief Jail Mental Health Screen. Psychiatric Services, 56, 816−822.
Trull, T. J. (2001). DSM-III-R personality disorder and the five-factor model of personality:
An empirical comparison. Journal of Abnormal Psychology, 101, 553−560.
van Panhuis, P. J. A. (1997). De psychotische patiënt in de TBS van kwaad tot erger, Een
forensisch psychiatrische studie. Gouda Quint Deventer.