Due September 9- 200 word minimum and reference
The Mental Health of Federal Offenders: A Summative Review of the Prevalence Literature*
Philip R. Magaletta,1 Pamela M. Diamond,2,5 Erik Dietz,3 and Stephen Jahnke4
To date, only a small number of government and peer-reviewed studies have examined the mental health of federal offenders. Although these studies provide isolated bits of information they have yet to be organized into a coherent body of knowledge from which clinicians, administrators and policy makers can inform their work. As a first step in constructing this knowledge and understanding the possible mental health needs of this population (currently America’s largest correctional population), this paper delineates the available government and peer-reviewed studies on federal offenders, highlights their convergent findings, and suggests opportunities for growth in research, administration and policy.
KEY WORDS: offenders; federal prisons; service utilization; diagnoses.
There is an increasing demand for effective, empirically informed, prison-based mental health services in America. It is a demand driven by the needs of the offender population, the clinicians who serve them, and the public’s expectation of accountability. It is the product of multiple factors: courts mandating that mentally ill persons receive treatment while in custody; national mental health screening and treatment standards being rigorously applied; and increasingly porous boundaries be- tween the mental health and criminal justice systems
(Fisher et al., 2002; Jemelka, Trupin, & Chiles, 1989). Furthermore, growth in the offender popu- lation has remained mostly constant (Harrison & Beck, 2005) and little debate remains that the prevalence of mental illness in prison populations is higher than that of the general population (Dia- mond, Wang, Holzer, Thomas, & Cruser, 2001; Jemelka et al., 1989). Finally, among community mental health providers there is an increasing rec- ognition that many patients have histories of crimi- nality, incarceration, and prison-based mental health treatment (Jemelka et al., 1989; Manderschied, Gravesande, & Goldstrom, 2004; Morgan, Beer, Fitzgerald, & Mandracchia, in press).
Far beyond the application of mental health principles to those who ‘‘simply’’ happen to be incarcerated, the provision of mental health services in corrections remains a complex enterprise. It re- quires strong clinicians, administrators who have a keen and sensitive understanding of the multiple systems comprising the correctional environment, and policy makers who can draw upon an empirical understanding of the population’s needs. To inform the effective deployment of mental health resources to this growing population it is imperative that this
*The views expressed in this paper are those of the authors (Philip R. Magaletta and Erik Dietz) only and do not necessarily rep- resent the policy or opinions of the Federal Bureau of Prisons, the Department of Justice, or their academic affiliates.
1 Psychology Services Branch, Federal Bureau of Prisons, 320 First
Street, NW, Washington, DC 20534, USA. 2Center for Health Promotion, University of Texas at Houston
Health Sciences Center, Houston, TX, USA. 3 Psychology Services Branch and Office of Research and Evalu-
ation, Federal Bureau of Prisons, 320 First Street, NW,
Washington, DC 20534, USA. 4 Marymount University, Arlington, VA, USA.
5 Correspondence should be directed to Pamela M. Diamond,
Center for Health Promotion, University of Texas at Houston
Health Sciences Center, 7000 Fannin, STE 2614, Houston, TX,
77030, USA; e-mail: pdiamond@sph.uth.tmc.edu.
253
0894-587X/06/0500-0253/0 � 2006 Springer Science+Business Media, Inc.
Administration and Policy in Mental Health and Mental Health Services Research, Vol. 33, No. 3, May 2006 (� 2006) DOI: 10.1007/s10488-005-0022-2
empirical body of knowledge be developed and provided sustained attention.
To date, the majority of published studies that have shaped our understanding on the mental health of offenders have sampled from state prisons (e.g., California—Asnis, Kaplan, Hundorfean & Saeed, 1997; Massachusetts—DiCataldo, Greer, & Profitt, 1995; Utah—Eyestone & Howell, 1994; North Caro- lina—Jordan, Schlenger, Fairbanks, & Caddell, 1996; New York—Steadman, Fabisiak, Dvoskin, & Ho- lohean, 1987; Steadman, Holohean, & Dvoskin, 1991). Although a few isolated studies have sampled from the Federal Bureau of Prisons (BOP) very little systematic knowledge on the prevalence of mental disorders among federal offenders exists. Further- more, given the distinctions in population character- istics, the generalizability of findings from state studies to federal offenders remains questionable. For example, the 10% of non-US citizens estimated to be in state custody (Harrison & Beck, 2005) is far less than one would find in federal custody (27.3%; Bu- reau of Prisons, 2005). Although the national average for drug crimes leading to state custody is estimated to be 20%, a total of 53% of federal offenders are incarcerated for such crimes. Further, growth in the federal population has far outstripped the states. From 1995 to 2004 the state population increased 14%, while the federal population grew 66% (Harri- son & Beck, 2005). Presently, the BOP is America’s largest correctional agency (Harrison & Beck, 2004).
It is not only population characteristics that may differentiate state and federal jurisdictions. For example, the infrastructure of the BOP is unique in being a coast-to-coast correctional system consisting of over 105 prisons, several administrative and training centers and over two-dozen community corrections offices. There is no centralized intake or reception center where all offenders receive mental health screening, assessment and treatment—this is a responsibility shared by and provided at each of the federal prisons. In most of these facilities doc- toral level psychologists provide these front-line mental health services (Magaletta & Boothby, 2003). Also of note, there is no parallel federal mental health/hygiene system as found in some states (Manderschied et al., 2004). The five geo- graphically dispersed Federal Medical Centers where the most severely mentally ill are cared for operate as part of the prison system itself. Within the context of this geographic dispersion, the cost of transporting offenders to these medical centers re- mains an administrative challenge and the difficulty
in sustaining family contact and social support is obvious.
Given these unique features and the absence of systematic empirical knowledge on the topic, the need to codify and synthesize the existing preva- lence literature on federal offenders is apparent. For the approximately 186,000 offenders, 35,000 employees and the multitude of community mental health providers who will be treating these offenders once they release from custody1 the significance of this need cannot be overstated. The purpose of this paper is to meet this need. We provide a summative review of all available studies on the mental health of federal offenders; highlight the points of conver- gence among their collective findings; and suggest opportunities for future growth in the areas of re- search, administration and policy.
GATHERING AND ORGANIZING THE STUDIES
Two methods were used to gather the studies used in this review. Since some of the government studies were unpublished and existed only as internal documents, presentations at national conferences, or reports to congress, the first search strategy was to review the BOP archives. The second method included a comprehensive search of MedLine, PsychLit and ProQuest medical, psychological and criminal justice databases for a thirty-year-publishing period. Studies selected from that search were germane to mental health prevalence in American, federal offenders. Together, this dual archives/data- base strategy resulted in the identification of 10 studies. As a final measure to assure the completeness of our approacheach of these 10 studies were entered into a citation index. We wanted to explore whether any additional studies referencing our 10 contained any additional, salient information. Although several studies included reference to our 10, they did not lend any additionally relevant information.
An initial examination of the 10 studies revealed that mental health prevalence was mea- sured from multiple perspectives. Some tapped mental health service utilization, others chose men- tal health diagnoses, and some provided their own hybrid definition. Adding further variance across studies, each prevalence definition was embedded in
1Approximately 60,000 federal offenders are released from prison
yearly.
254
its own timeframe. For example, mental health ser- vice utilization prior to and during incarceration, currently and over the lifetime were all represented.
To assist in the first level of organizing these studies and their multiple measures and findings, the government produced, non-peer-reviewed studies were reviewed separately from the peer-reviewed publications from scholarly journals. Within that level of organization each study was examined in terms of its core research elements: sampling design, measures, and results (see Tables 1 & 2). The nar- rative text that follows will expand the information found in the tables by highlighting the strengths, weaknesses and unique features of each study.
HIGHLIGHTS OF GOVERNMENT STUDIES
In 1991 the General Accounting Office (GAO) assessed the number of federal offenders with an Axis I diagnosis and the extent to which they were participating in treatment. To accomplish this, sur- veys were given to administrators at all BOP facili- ties (N=65; July, 1990). They were asked to have their mental health staff complete the survey based upon ‘‘what their records showed’’ or if untenable in the time frame provided, their ‘‘best estimate’’ (p. 4). Findings indicate that in July 1990, 5.5% of the population, or 3131 offenders, had received an Axis I diagnosis other than a substance use disorder. Nearly a quarter of those diagnosed (650 cases) were receiving inpatient treatment at one of the BOP’s medical referral centers. Collectively, non-medical mental health services were being utilized by nearly two-thirds of this Axis I population (3.7% of the entire BOP census that year), and it is suggested that the remainder of the Axis I group might have been on medication or under the care of a psychiatrist.
To contextualize the findings of this report, the authors suggest that by sampling a diagnosed popula- tion, results were likely to under-estimate the true number of mentally ill offenders. For example, offenders who were mentally ill but did not receive a diagnosis, those who refused an evaluation on which to build a diagnosis, or those who did not yet manifest the necessary severity or constellation of symptoms to warrant a conclusive diagnosis were not included in the sample. Consistent with this observation, more than half the facilities estimated that they had at least some offenders who were mentally ill but not diagnosed.
Despite this selection and possibly investigator bias (BOP clinicians produced the data) the scope of
this study cannot go unnoticed. The GAO researchers received census data, from all BOP facilities operating in 1990. Male and female offenders; minimum, low, medium and high security levels; and the medical centers which house the sickest mentally ill offenders were all represented. Unfortunately, what was gained in terms of the scope of this census was lost in terms of being able to provide diagnostic specificity for the group.
In 1992 Scheckenbach and Pape produced an internal, summative report from BOP mental health records (the Psychology Data System, PDS). Aggregating archived psychology service intake screenings conducted in 1992, the authors provided information on mental health service utilization among federal offenders. During that year, doctoral level psychologists conducted and recorded 41,383 intake-screening interviews. Of those offenders interviewed, 7% reported a lifetime history of inpatient mental health services and 12% reported a lifetime history of outpatient mental health services. Fourteen percent of those interviewed during the admissions process were assessed as being in current need of mental health services.
The strength of the Scheckennbach & Pape study again results from the census data that it in- cluded. Similar to the GAO report, it resulted from information collected at nearly all BOP facilities (98%) that year and includes male and female offenders across security levels. Furthermore, these data were gleaned through face-to-face interviews conducted by doctoral level psychologists and pro- vide an additional level of validity beyond offender self-report. These strengths not withstanding, it should be noted that the census was limited to offenders who met criteria for an admissions inter- view in 1992—namely new commitments, and those already in custody who transferred to another BOP facility that year. These findings did not reflect the service need or histories of offenders who were in residence at the same facility during 1992 including the most seriously mentally ill housed at the medical referral centers. In addition, the authors did not delineate whether services were utilized prior to or during the present incarceration and again, diag- nostic specificity is lacking in this study.
Further exploring the perspective of lifetime mental health service utilization, a 1998 interim re- port on an outcome study for the BOP Residential Drug Abuse Treatment Program (Pelissier et al., 1998) included relevant mental health information. The study sampled 1966 male and female drug
255
T a
b le
1 .
O v
e rv
ie w
o f
G o
v e
rn m
e n
t S
tu d
ie s
T it
le S a m p li n g D es ig n
M ea su re s
R es u lt s
S e
rv ic
e U
ti li
z a
ti o
n *
D ia g n o si s
G e
n e
ra l
A c c o
u n
ti n
g
O ffi
c e
(1 9
9 1
)
A ll ex is ti n g B O P fa ci li ti es
in
1 9 8 9 (N
fa ci li ti es
= 6 5 ;
N o ff en d er s =
5 6 ,9 0 0 )
S u rv ey
o f cl in ic ia n s a n d o ff en d er
re co rd
re v ie w
o r cl in ic ia n ‘‘ b es t es ti m a te ’’
3 .7 %
en ro ll ed
in se rv ic es
a t
ti m e o f su rv ey
5 .5 %
id en ti fi ed
a s h a v in g A x is I
d ia g n o si s, n o t in cl u d in g
su b st a n ce
u se
d is o rd er s
S c h
e c k
e n
b a
c h
&
P a
p e
(1 9
9 2
)
A ll n ew
co m m it m en ts
a n d
n ew
a d m is si o n s to
B O P
n o n -m
ed ic a l fa ci li ti es
(N =
4 1 ,3 8 3 )
C o m p u te ri ze d re co rd
o f P sy ch o lo g y
S er v ic es
In ta k e S cr ee n in g (P D S )
7 %
in p a ti en t (l if et im
e)
1 2 %
o u tp a ti en t (l if et im
e)
N /A
1 4 %
id en ti fi ed
a s n ee d in g m en ta l h ea lt h se rv ic es
cu rr en tl y
P e
li ss
ie r
e t
a l.
(1 9
9 8
)
8 D ru g tr ea tm
en t si te s a n d
m a tc h ed
n o n -t re a tm
en t
co m p a ri so n g ro u p ;
st ra ti fi ed
b y g en d er :
T re a tm
en t—
N =
6 6 3
(M );
N =
1 3 9 (F )
C o m p a ri so n —
N =
7 7 1 (M
); N =
1 5 1 (F )
D ia g n o st ic
In te rv ie w
S ch ed u le
(D IS ) &
C li n ic a l in te rv ie w
1 7 .6 %
(M ); 3 9 %
(F ) D ru g
tr ea tm
en t g ro u p , A n y
m en ta l h ea lt h se rv ic es
u ti li za ti o n (l if et im
e)
C o m p a ri so n g ro u p :
1 9 %
(M ); 4 1 %
(F )
1 4 .9 %
(M ); 3 0 .3 %
(F ) D ru g
tr ea tm
en t g ro u p , D ep re ss io n
(L if et im
e)
C o m p a ri so n g ro u p :
1 6 %
(M ); 3 7 .3 %
(F ).
D it
to n
(1 9
9 9
)— B
J S
1 9 9 7 S u rv ey
(N =
4 0 4 1 )
S el f- re p o rt
o n fi v e m en ta l h ea lt h it em
s
fr o m
‘‘ S u rv ey
o f In m a te s in
S ta te
o r
F ed er a l C o rr ec ti o n a l F a ci li ti es ’’
4 .7 %
in p a ti en t (l if et im
e)
1 1 %
m ed ic a ti o n
1 2 %
co u n se li n g
4 .8 %
cu rr en t m en ta l o r
em o ti o n a l co n d it io n
7 .4 %
es ti m a te d to
b e m en ta ll y il l—
h y b ri d o f cu rr en t m en ta l h ea lt h
co n d it io n o r li fe ti m e in p a ti en t h is to ry
G a
e s
& K
e n
d ig
(2 0
0 2
)
O ff en d er s re le a se d to
th e
co m m u n it y in
2 0 0 0
(N =
4 3 ,1 8 7 )
A rc h iv a l re co rd s d o cu m en ti n g m en ta l
h ea lt h d ia g n o se s (I C D
co d e)
a ss ig n ed
b y a m ed ic a l p er so n n el
d u ri n g
in ca rc er a ti o n
N /A
2 .6 3 %
— m a jo r m en ta l d is o rd er
(s ch iz o p h re n ia , b ip o la r, et c)
(l if et im
e)
V e y
se y
& B
ic h
le r-
R o
b e
rt so
n (2
0 0
2 )
1 9 9 5 P ri so n C en su s
W ei g h te d es ti m a te s o f D S M
d ia g n o si s
b a se d u p o n m en ta l h ea lt h p re v a le n ce
ra te s fr o m
N a ti o n a l C o m o rb id it y
S u rv ey
(N C S )
N /A
2 .5 %
sc h iz o p h re n ia
2 .7
% b
ip o
la r
1 5
.7 %
d e
p re
ss io
n
1 1
.6 %
d y
st h
y m
ia
2 3
% a
n x
ie ty
6 .8
% P
T S
D (l
if e
ti m
e )
* T
e x
t a
p p
e a ri
n g
a c ro
ss se
rv ic
e u
ti li
z a
ti o
n a
n d
d ia
g n
o si
s c a
te g
o ri
e s
a re
c o
n si
d e
re d
‘‘ h
y b
ri d
’’ a
n d
w e
re n
o t
e x
c lu
si v
e to
e it
h e
r c a
te g
o ry
.
256
T a b
le 2
. O
v e
rv ie
w o
f P
e e
r R
e v
ie w
e d
S tu
d ie
s
T it
le S a m p li n g D es ig n
M ea su re s
R es u lt s
S e
rv ic
e U
ti li
z a
ti o
n *
D ia g n o si s
R o
th &
E rv
in (1
9 7 1
) A ll m a le
in m a te s a t F C I
L ew
is b u rg , P A
o n Ja n u a ry
1 ,
1 9 6 9 (N
= 1 1 5 4 re co rd s
co d ed — 9 3 %
o f to ta l
a v a il a b le )
A rc h iv a l re co rd
re v ie w
o f p re -
se n te n ci n g in te rv ie w s
co n d u ct ed
b y F ed er a l
P ro b a ti o n O ffi ce rs
1 2 %
o u tp a ti en t se rv ic es
(l if et im
e)
8 %
in p a ti en t m en ta l h ea lt h
(l if et im
e)
1 8 %
h a d a t le a st
o n e
p sy ch ia tr ic
co n ta ct
p ri o r to
in ca rc er a ti o n
8 %
p sy ch o si s
6 %
n e
u ro
si s
3 1
% P
e rs
o n
a li
ty D
is o
rd e
r
(l if
e ti
m e )
W a
lt e
rs e
t a
l. (1
9 8 8
) M a le
In m a te s a t U S P
L ea v en w o rt h w h o ei th er
ex h ib it ed
p sy ch ia tr ic
sy m p to m s a t in ta k e,
w er e
cu rr en tl y in
tr ea tm
en t, o r
re fe rr ed
fo r ev a lu a ti o n
(N =
5 2 )
S tr u ct u re d cl in ic a l in te rv ie w
(P sy ch ia tr ic
D ia g n o st ic
In te rv ie w )
7 5 %
o u tp a ti en t (l if et im
e)
7 3 %
in p a ti en t (l if et im
e)
6 0 %
sc h iz o p h re n ia
2 7
% m
a n
ia
5 4
% d
e p
re ss
io n
4 %
sc h
iz o
p h
re n
if o
rm *
*
(c u
rr e
n t/
p a
st y
e a
r)
9 %
o f
2 1
8 c o
n se
cu ti
v e
ly in
te rv
ie w
e d
h ig
h -s
e c u
ri ty
in m
a te
s c o
n fi
rm e
d
to h
a v e
m a
jo r
m e
n ta
l h
e a
lt h
d is
o rd
e r
K o
e n
ig (1
9 9 5
) M a le In m a te s a g e 5 0 o r o ld er
a t
F C C
B u tn er , N C
(N =
9 6 )
S tr u ct u re d cl in ic a l in te rv ie w
(C E S -D
)
3 7 %
h a d h is to ry
o f p sy ch ia tr ic
p ro b le m s re q u ir in g tr ea tm
en t
(l if et im
e)
2 4 %
D ep re ss iv e sy m p to m s in
cl in ic a l ra n g e (c u rr en t)
K o
e n
ig e
t a
l. (1
9 9 5
) M a le In m a te s a g e 5 0 o r o ld er
a t
F C C
B u tn er , N C
(N =
9 5 )
S tr u ct u re d cl in ic a l in te rv ie w
(D IS ); re co rd
re v ie w
N /A
1 0 .5 %
m a jo r d ep re ss io n
(c u rr en t)
3 3
.7 %
m a
jo r
d e
p re
ss io
n
(l if
e ti
m e )
1 .1
% sc
h iz
o p
h re
n ia
(l if
e ti
m e
&
c u
rr e
n t)
4 .2
% a
n x ie
ty d
is o
rd e rs
(c u
rr e
n t)
1 6
.8 %
a n
x ie
ty d
is o
rd e
rs
(l if
e ti
m e )
257
abusers from over 20 institutions who were assigned to either a treatment or comparison group for the study. During a clinical interview with a trained re- search assistant offenders were asked, ‘‘Did you ever get mental health counseling for your emotions, nerves, or mental health?’’ The male treatment group endorsed this item at 17.6%, as did 19% of the comparison group subjects. For females the endorsement was significantly higher, 39% and 41%, respectively for the treatment and comparison groups.
Addressing the diagnostic specificity concern from the two earlier studies, in addition to lifetime mental health service utilization this study used the Diagnostic Interview Schedule (DIS; Helzer & Robins, 1988) to measure lifetime rates for a diag- nosis of Depression and/or Antisocial Personality Disorder. Among the men in the treatment group 14.9% met criteria for a lifetime diagnosis of Depression, as did 16% of the comparison group. Again, the trend for women was higher, with 30.3% in the treatment group meeting criteria for Depression, along with 37.3% from the comparison group.
Obvious strengths of this study were the use of validated, structured clinical interviews and the multi-site sampling strategy. In terms of the ob- served sample however, it must be remembered that they were purposefully selected to have confirmed substance use disorders. There remains a strong likelihood that as substance abusing offenders they were younger and more likely to have other co- morbid Axis I conditions compared to the general population of offenders.
In 1999, the Bureau of Justice Statistics released the Mental Health and Treatment of Inmates and Probationers (Ditton, 1999). The mental health data presented in the report were extracted from a larger survey interview of offender needs and relied en- tirely on offender self-report. The prime findings of relevance for the present review were the weighted estimates that revealed 4.7% of the federal offend- ers reported an overnight stay in a mental health treatment facility over their lifetime and that 4.8% reported having a current mental or emotional condition. Among the 7.4% of offenders who met either of these two conditions, and whom Dietz refers to as the mentally ill group, mental health services were used by approximately 60% of the group during their present incarceration. Specifi- cally, 24% reported an additional overnight stay in a mental hospital or treatment program; 49.1% had
taken a prescribed medication; 45.6% had received counseling or therapy.
To expand the information available from this study, the raw data and weights used by Ditton (1999) were extracted from a public web site and one of the present authors, P. Diamond, conducted additional analysis. These analyses revealed that when medication and counseling were added to the overnight stay and current condition items, 15.8% of the entire federal sample had received mental health services over their lifetime. The most frequent types of services utilized among all federal offenders sur- veyed were medication (11%) and counseling (12%). In contrast to Ditton’s mentally ill group, those who had only reported medication or outpa- tient but not an overnight stay or current condition reported using few mental health services while in prison. Only 4% of these individuals reported receiving received medication and 4% reported engaging in counseling.
Although the Ditton study did employ a fairly robust research design, including a cluster sampling method weighted to be proportional to the overall census of federal offenders at that time, several caveats should be mentioned. It remains unclear how the federal medical referral centers (where the sickest mental health offenders receive treatment) were factored into the design (if at all). In addition, it is unlikely that those who were truly, actively mentally ill at any facility would have been able to complete the full interview from which the mental health data were culled. Finally, since entry into the mentally ill group could be gained either through an event historically distant (inpatient treatment) or current (emotional condition) it remains difficult to assess the meaning of service utilization during incarceration for this group.
In the only study that examined mental health prevalence from the perspective of offender reentry, Gaes & Kendig (2002) provided an excellent over- view of the health care needs and skill sets associ- ated with the successful community reentry of federal offenders. Providing population data on all federal offenders released from custody and not deported or detained in other jurisdictions during 2000 (N=43,187) this study reported that 2.63% of those released received a medical diagnosis of a major mental disorder such as schizophrenia, bipolar disorder, depression and/or pre-senile dementia.
The findings only represented diagnoses received upon evaluation or treatment by a physician some time prior to release. They do not include those who
258
had a mental illness but were receiving non-medical mental health treatment, those who refused medical intervention, or those who were never properly referred for such intervention. Consistent with this selection bias, these findings are likely to under estimate all those released who actually had mental health conditions. Since Ditton (1999) reported that only half the mentally ill federal offenders in their sample had been prescribed a medication since their admission, selection based on a medical diagnosis is highly likely to under-represent the true population of mentally ill offenders.
Finally, since the report was written as a con- ceptual overview and not a research study, the sampling strategy was not fully explicated. It re- mains unclear how far back the authors searched in each offender’s record for the diagnosis of interest. If they only searched 1 year prior to the release date the seemingly low estimate could be seen as an artifact of the sampling strategy. Conversely, it re- mains entirely plausible that there is some intrinsic factor of release cohorts that co-varies with the likelihood of mental health diagnoses. For example, they may represent much healthier, shorter sentence offenders. With little research on this topic these possibilities remain to be explored.
The sixth and final government study (Veysey & Bichler-Robertson, 2002) used synthetic estimates to predict lifetime prevalence rates for psychiatric disorders among federal offenders. Methodologi- cally unique, these estimates were generated by taking nationally representative mental health data gleaned from the National Comorbidity Survey (Kessler, 1994) and weighting it to match key mental health and demographic strata found in state and federal prison populations from a 1995 census. Among federal offenders, lifetime prevalence of mental disorders were estimated to be 2.5% for schizophrenia or other psychotic disorder, 2.7% for bipolar (mania), 15.7% for major depression, 11.6% for dysthymia, 23% for anxiety, and 6.8% for post- traumatic stress disorder.
The strength of this study is that it begins to address the issue of diagnostic specificity. On the other hand, it must be acknowledged that although the findings result from a methodologically stringent design based upon state-of-the-art science in com- munity mental health prevalence, they do remain synthetic. Furthermore, it should be noted that the population parameters used to draw federal offen- der estimates accommodated ‘‘the greater number of white collar offenders from higher socioeconomic
status incarcerated in federal prisons.’’ The accuracy of this adjustment remains to be seen and may be particularly problematic since it did not account for the large number of drug abusing federal offend- ers—a parameter which was used to adjust estimates for findings reported on state offenders.
In summary, most of these six government studies measured and reported robust findings regarding lifetime estimates of mental health service utilization. The studies revealed strengths in terms of sample sizes, with two reporting population data. In terms of weaknesses however, it is readily apparent that what was gained in terms of access to population data was lost in diagnostic specific- ity—only two studies examined diagnosis.
HIGHLIGHTS OF PEER-REVIEWED STUDIES
One of the earliest peer-reviewed studies of federal offenders also remains one of the most thorough (Roth & Ervin, 1971). In January 1969 the lead author, a psychiatrist at the medium security federal correctional institution in Lewisburg, Penn- sylvania, began a retrospective records review for all offenders in the institution on January 1, 1969. Through this review, the authors tabulated the rate of lifetime mental health service utilization. Data on 93% of the offenders (N=1154) were captured over a 3-month period.
Highlights from this study include the finding that 18% of the sample had at least one psychiatric contact prior to their incarceration. More specifi- cally, 10% had histories of inpatient and 12% had histories of receiving outpatient care. It is important to note that in this study the definition for inpatient included inpatient drug treatment. When these offenders are excluded, making the inpatient mea- sure consistent with other studies, this percentage drops to 8%. Lifetime rates of diagnoses, which in- cluded diagnoses made during the current incarcer- ation were also reported in this study: 31% for Personality Disorder; 8% for Psychoses; (including schizophrenia and ‘‘other’’); and 6% for ‘‘Neuro- sis’’(anxiety disorders).
Although this study was limited to one correc- tional institution, the fact that over 1000 offenders entered the sample remains quite impressive and goes far to broaden the potential generalizability of the findings. Unfortunately, the application to other pertinent segments of the population (i.e. offenders
259
at medical centers and female offenders) remains unknown.
Also exploring the issue of diagnostic specific- ity, a study by Walters, Mann, Miller, Hemphill, & Chlumsky (1988) described the types of disorders represented among pre-identified groups of mentally ill offenders. In this study, purposeful sampling was achieved by applying several criteria to a group of offenders from a high security federal prison: receiving psychotropic medications (N=22); being referred to psychology services for an evaluation (N=11); or reporting psychological symptoms at in- take and being confirmed to have a major mental disorder (N=19). Once these participants were se- lected (N= 52), they were administered the Psychi- atric Diagnostic Interview (PDI; Feighner, et al., 1972). The following ‘‘current (past year)’’ diag- nostic prevalence rates emerged: Schizophrenia 60%; Mania 27%; Depression 54%; and Schizo- phreniform 4%. In this study, offenders could be counted in more than one group, although the de- gree of comorbidity was not reported. Follow-up testing for the sample was conducted using the Minnesota Multiphasic Personality Inventory (MMPI; Dahlstrom, Welsh, & Dahlstrom, 1972) and were consistent with the PDI findings.
One of the more intriguing findings from this study was that 75% of the sample reported lifetime outpatient service use and 73% reported receiving inpatient services, thus leaving a quarter of the diagnosed sample reporting no mental health ser- vices history. This echoes earlier concerns with selection bias issues in measuring service utilization as a proxy for past or current service need. Finally, although it is not highlighted within the study itself, one additional finding can be drawn out. The third group of participants (N=19) selected for inclusion in the study was screened from among 218 offenders consecutively seen for a standard intake-screening interview. These were offenders who reported psy- chological symptoms during their intake, were sub- sequently administered the PDI and who were confirmed to have a ‘‘major mental disorder’’. It can thus be said that in the mid-eighties at a high secu- rity federal prison, 9% of the offenders consecu- tively screened at intake were confirmed to have a major mental disorder.
The last two published studies focused on the religious and mental health needs of older offenders (Koenig, 1995; Koenig, Johnson, Bellard, Denker, & Fenlon, 1995). These studies examined lifetime and current rates of major psychiatric disorders among
offenders aged 50 or more at one minimum and one low security facility. The final sample included 95 offenders who completed a comprehensive battery including a base-line mental health services inter- view with review of medical, mental health and pharmacy records; the twenty item self-report Cen- ter for Epidemiologic Studies Depression Scale (CES-D); the depressive and anxiety disorders pro- tocols from the Diagnostic Interview Schedule (DIS); and a DSM-III-R checklist for other disor- ders.
The obvious strength of these two studies was the use of structured clinical interviews. Results from the CES-D (Koenig, 1995) suggested that 24% of the sample currently scored in the clinically de- pressed range. The DIS, with its more stringent one- month (current) criteria revealed a lower rate, 10.5%. Using the DIS for lifetime rate of Major Depression rates increased to 33.7% (Koenig et al., 1995). The rate for all anxiety disorders excluding simple phobias was 4.2% for one-month and 16.8% for lifetime. Schizophrenia was observed in 1% of the sample at both one-month and lifetime rates. Bipolar Disorder was reported as 1% for one month and 4% for lifetime. In addition to this diagnostic information, the lifetime rate for any service utili- zation was assed at 37%. On a final note, it should be highlighted that given the age of the offenders in the sample, their longer ‘‘time at risk’’ is likely to have inflated lifetime rates for both diagnosis and service utilization.
In summary, most of the peer-reviewed studies used validated structured clinical interviews or objective psychological testing, thus allowing for a fine degree of diagnostic precision. In terms of weaknesses however, this precision came at the cost of much smaller, non-generalizable, samples of offenders. Further, none of these published studies included samples of female offenders. They were each limited to one or two security levels and issues of race, ethnicity, citizenship and offense type were unexplored.
POINTS OF CONVERGENCE AND OPPORTUNITIES FOR FUTURE GROWTH
Taken together these studies form a nexus of information allowing us to observe points of con- vergence among their findings. It also allows us to determine future growth opportunities for research, administration and policy in this area. The largest
260
and clearest set of convergent findings concerns mental health service utilization. The three studies that measured outpatient services (including ‘‘counseling’’ in the newer Diamond analyses) all precisely converged on a 12% endorsement rate. This suggests that, in part, cohorts of federal offenders separated over time have not generally differed from one another on this aspect of mental health prevalence.
Inpatient mental health service utilization rates within a range of 5–8% were also consistently re- vealed through time. However, paralleling an ob- served decrease in inpatient bed availability (Manderscheid et al., 2004) the inpatient endorse- ment rate showed a downward trend through time (8% in 1970; 7% in 1992; and 4.7% in 1999). That being said, it should also be noted that overnight stay criteria throughout that timeframe had become increasingly stringent and this might suggest that those with more recent inpatient histories represent a group with more severe conditions. Additionally, it should be pointed out that the hundreds fold in- crease in the prison population during that same time means that despite the consistency observed in the absolute frequencies, the overall number of offenders with inpatient histories has increased dramatically (Jemelka et al., 1989; Manderscheid et al., 2004).
Interestingly, each of the studies measuring both mental health service utilization and diagnosis in their sample highlighted the problem of equating service utilization with diagnosis. Specifically these studies consistently indicated rates of diagnosis were higher than rates of service utilization. To only consider diagnosis in the study of mental health prevalence in federal offenders may lead to an over representation of those who eventually need or choose to receive services. Conversely, only mea- suring service-utilization may under-represent those who have a diagnosable and potentially treatable mental health condition. In terms of future admin- istrative planning these findings clearly suggest that to calibrate prison based mental health classification and/or staffing levels on the basis of observed past or present diagnosis may be misguided; looking only for historical markers of service utilization may fail to identify many who currently or will eventually need such services (see Mears (2004) for an enlightening consideration of this subject). At least in a federal population, it appears that current ser- vice need, previous service utilization and diagnostic prevalence will all represent important aspects but
not exclusive determinants of eventual need for prison based mental health services.
Points of convergence with community mental health literature (Dickey & Blumberg, 2004; Kess- ler, 1994) can also observed and suggest that dif- ferent demographic identifiers can lead to very different prevalence rates both in terms of service utilization and diagnosis. Specifically, findings from Pelliser et al. (1998) and Koenig (1995) suggest that men and women and older and younger federal offenders, at least in those samples, had divergent rates of previous mental health service utilization as well as diagnostic profiles. Potentially leading to differential policy implications for their custody and care, follow-up on these differences is imperative for future research and planning.
A final key consistency can be observed be- tween those studies that used hybrid measures of ‘‘mental illness’’ by combining prior mental health service utilization and/or self reported mental health problems (i.e., GAO, Ditton, and Walters). In these studies, the 6–9% range was consistently reported from those sampled. Unfortunately, these studies were unable to draw a nexus between the exact nature of the offender’s diagnoses and their current need for services. In fact, in terms of future research the issue of parsing out exact diagnostic prevalence for ‘‘mental illness’’ remains one of the prime areas for growth. To date, the studies that have examined similar or even the same diagnosis were either syn- thetic, or did so with very small, specialized offender samples making it impossible to generalize to the larger federal offender population.
It is also notable that few diagnostic-based conceptualizations of mental health prevalence that are highly relevant to a correctional context were examined in the existing studies. For example, rates of Axis II disorders and other types of behavioral or acting-out disorders were rarely examined and re- main largely unknown. Neurological problems and traumatic brain injury, both of which are suspected to have a higher incidence in an offender population (Osview & Fenwick, 1999) remain unexamined. Further complicating this observation, the issue of functionality of the individual and the larger reality of whether or not they ever had access to mental health care was not considered in any of the studies reviewed.
Once more precise diagnostic prevalence data on larger, representative samples of offenders is established, the next ambitious step would be to disaggregate these prevalence rates by risk factors
261
such as age, gender and ethnicity and to examine the current as well as lifetime rates of the disorders (Diamond et al., 2001). As a further step in devel- oping this stream of knowledge, offenders could be followed through time and across their vari- ous interactions with elements of the criminal justice system, as well as in their communities upon release.
Given what is now known, that a large number of offenders have already received mental health services, it seems appropriate that administrators continue their requests for the increased use of prevalence data to drive mental health services and to guide the selection of targets for intervention. A major role for such administrators will be to cham- pion the development of information technology systems and tools that simultaneously meet the needs of clinicians, administrators, researchers and policy makers. In so doing, administrators and policy makers charged with planning for and evaluating the custody and care of offenders will be allowed to make more informed staffing decisions across their expansive systems. In addition, it would allow them a degree of precision in anticipating the current and future service needs of their population.
Beyond the level of prison administrators, the need for empirical knowledge in this area also needs to be embraced by public safety and public health research funding sources2 (Magaletta & Verdeyen, 2005). Over the past decade social science researchers have been providing empirical answers to the ‘‘What Works’’ question—as it applies to criminal rehabilitation programs (Farabee, 2005). It is now time for behavioral science and mental health researchers to join with their social science col- leagues and together, effect change in the identifi- cation and treatment of mentally ill offenders in prison. Given that such research requires mastery and integration of knowledge across several domains—psychopathology, economics, the process of psychological and behavioral change, statistical methodologies, biological change agents, demo- graphics, and crime and criminality—it is clear that interdisciplinary research teams will be the ones advancing us to the next level of understanding and comprehension.
CONCLUSION
Looking through the collective lens of these studies, it is clear that this area of inquiry remains in an early developmental stage. Not one of the studies referenced or cited the other. Each study was con- ducted in isolation and each group of researchers began the process from scratch. Up to this point, the end result of such a process has been multiple bits of information but no coherent body of knowledge. By weaving together these once isolated studies, we have begun a process of integrating and expanding the available research—a process that will require and is actually receiving sustained attention. Re- cently, the BOP has devolped resources to conduct a comprehensive study on the prevalence of mental health service utilization, mental health diagnosis and behavioral disorders (Magaletta, Dietz & Dia- mond, 2005). From a sample of newly committed men and women across security levels and geo- graphic locations, initial prevalence estimates drawn from both current and lifetime mental health ser- vices utilization and diagnostic prevalence will be examined. This will be followed up with longitudinal data on eventual in-prison disciplinary histories and mental health services utilization as these offenders move through the criminal justice system and back into the community. For the many clinicians working tirelessly with these offenders, the administrators who are charged with planning, arranging and eval- uating often custody and care, and the community mental health providers who work to develop seamless reception and integration of these offend- ers back into their communities, the need for such a study has never been greater.
REFERENCES
Asnis, G. M., Kaplan, M. L., Hundorfean, G., & Saeed, W. (1997). Violence and homicidal behaviors in psychiatric disorders. Psychiatric Hospitals, 20(2), 405–425.
Bureau of Prisons (2005). Quick Facts. Washington, DC: U.S. Department of Justice, October 4, 2005. Available online at http://www.bop.gov/news/quick.jsp.
Dahlstrom, W. G., Welsh, G. S., & Dahlstrom, L. E. (1972). The MMPI Handbook, Vol. 1: Clinical Interpretation. Minneap- olis, MN: University of Minnesota Press.
Diamond, P. M., Wang, E. W., Holzer, C. E. III., Thomas, C. R., & Cruser, D. A. (2001). The prevalence of mental illness in prison: Review and policy implications. Administration and Policy in Mental Health, 29, 21–40.
DiCataldo, F., Greer, A., & Profit, W. E. (1995). Screening prison inmates for mental disorder: An examination of the
2 In fact, several recent research initiatives by the National
Institute of Justice, National Institute of Drug Abuse and the
Centers for Disease Control would indicate that this movement
has already begun.
262
relationship between mental disorder and prison adjustment. Bulletin of the American Academy of Psychiatry and the Law, 23(4), 573–585.
Dickey, W. C., & Blumberg, S. J. (2004). Prevalence of mental disorders and contacts with mental health professionals among adults in the United States: National health interview survey, 1999. In: R. W. Manderscheid & M.J. Henderson (Eds.), Mental Health, United States, 2002 (pp. 92–104). DHHS Pub No. (SMA) 3938. Rockville, MD: Substance Abuse and Mental Health Services Administration.
Ditton, P. M. (1999). Mental health and treatment of inmates and probationers. Bureau of Justice Statistics Bulletin. Washing- ton, D.C.: U.S. Department of Justice.
Eyestone, L. L., & Howell, R. J. (1994). An epidemiological study of attention-deficit hyperactivity disorder and major depres- sion in a male prison population. Bulletin of the American Academy of Psychiatry and the Law, 22(2), 181–193.
Farabee, D. (2005). Rethinking Rehabilitation. Washington, DC: AEI Press.
Feighner, J. P., Robins, E., Guze, S. B., Woodruff, R. A., Winokur, G., & Munoz, R. (1972). Diagnostic criteria for use in psychi- atric research. Archives of General Psychiatry, 26, 57–63.
Fisher, W. H., Packer, I. K., Banks, S. M., Smith, D., Simon, L. J., & Roy-Buinowski, K. (2002). Self-reported lifetime psychi- atric hospitalization histories of jail detainees with mental disorders: Comparison with non-incarcerated national sam- ple. The Journal of Behavioral Health Services and Research, 29(4), 458–465.
Gaes, G. G., & Kendig, N. (2002). The skill sets and health care needs of released offenders. Presented at From Prison to Home: The effect of incarceration and reentry on children, families and communities. January 30–31, 2002. U.S. Department of Health and Human Services and the Urban Institute.
General Accounting Office (1991). Mentally ill inmates: BOP plans to improve screening and care in Federal prisons and jails. GAO/GGD-92–13. Washington, DC: United States General Accounting Office November.
Harrison, P. M., & Beck, A. J. (2004). Prisoners in 2003. Bureau of Justice Statistics Bulletin. Washington, D.C.: U.S. Department of Justice.
Harrison, P. M., & Beck, A. J. (2005). Prison and jail inmates at midyear 2004. Bureau of Justice Statistics Bulletin. Washing- ton, D.C.: U.S. Department of Justice.
Helzer, J. E., & Robins, L. N. (1988). The diagnostic interview schedule: Its development, evolution and use. Social Psychi- atry and Psychiatric Epidemiology, 23, 6–16.
Jemelka, R., Trupin, E., & Chiles, J. A. (1989). The mentally ill in prisons: A review. Hospital and Community Psychiatry, 40(5), 481–491.
Jordan, B. K., Schlenger, W. E., Fairbanks, J. A., & Caddell, J. M. (1996). Prevalence of psychiatric disorders among incarcer- ated women: Convicted felons entering prison. Archives of General Psychiatry, 53, 513–519.
Kessler, R. C. (1994). The national comorbidity survey. Interna- tional Review of Psychiatry, 6, 365–376.
Koenig, H. G. (1995). Religion and older men in prison. Inter- national Journal of Geriatric Psychiatry, 10, 219–230.
Koenig, H. G., Johnson, S., Bellard, J., Denker, M., & Fenlon, R. (1995). Depression and anxiety disorder among older male
inmates at a federal correctional facility. Psychiatry Services, 45(4), 399–401.
Magaletta, P. R., Dietz, E. F., & Diamond, P. (2005). The Prev- alence of Behavioral and Psychological Disorders Among an Admissions Cohort of Federal Inmates. Bureau of Prisons, Research Review Board # 01–038. Washington, DC: Depart- ment of Justice.
Magaletta, P. R., & Boothby, J. L. (2003). Correctional mental health professionals. In T. J. Fagan & R. K. Ax (Eds.), Correctional mental health handbook (pp. 21–38). Thousand Oaks, CA: Sage.
Magaletta, P. R., & Verdeyen, V. (2005). Clinical Practice in Corrections: A conceptual framework. Professional Psy- chology, Research and Practice, 36, 37–43.
Manderscheid, R. W., Atay, J. E., Male, A., Blacklow, B., Forest, C., Ingram, L., Maedke, J., Sussman, J., & Ndikumwami, A. (2004). Highlights of organized mental health services in 2000 and major national and state trends. In: R. W. Manderscheid, & M. J. Henderson (Eds.), Mental Health, United States, 2002 (pp. 243–279). DHHS Pub No. (SMA) 3938. Rockville MD: Substance Abuse and Mental Health Services Administration.
Manderscheid, R. W., Gravesande, A., & Goldstrom, I. D. (2004). Growth of mental health services in state adult mental health correctional facilities, 1988 to 2000. Psychiatric Services, 55(8), 869–872.
Mears, D. (2004). Mental health needs and services in the criminal justice system. Houston Journal of Health Law and Policy, 4, 255–284.
Morgan, R. D., Beer, A. M., Fitzgerald, K. L., & Mandracchia, J. T. (in press). Graduate student’s experiences, interests and atti- tudes towards correctional/forensic psychology. Criminal Jus- tice and Behavior.
Osview, F., & Fenwick, B. C. (1999). Neuropsychiatry in the prison. In F. Osview (Ed.), Neuropsychiatry and mental health services (pp. 197–220). Washington, DC: American Psychi- atric Press Inc.
Pelissier, B. M., Gaes, G., Phodes, W., Camp, S., O-Neil, J., Wallace, S., & Saylor, W. (1998). TRIAD drug treatment evaluation project six-month interim report. Washington, DC: Federal Bureau of Prisons.
Roth, L. H., & Ervin, F. R. (1971). Psychiatric care in federal prisoners. American Journal of Psychiatry, 128(4), 56–62.
Scheckenback, A., & Pape, M. (1992). Psychology data system 1992 summary report. Washington, DC: Federal Bureau of Prisons.
Steadman, H. J., Fabisiak, S., Dvoskin, J., & Holohean, E. J. (1987). A survey of mental disability among state prison inmates. Hospital and Community Psychiatry, 38(10), 1086–1090.
Steadman, H. J., Holohean, E. J., & Dvoskin, J. (1991). Esti- mating mental health needs and service utilization among prison inmates. Bulletin of the American Academy of Psy- chiatry and the Law, 19(3), 297–307.
Veysey, B. M., & Bichler-Robertson, G. (2002). Prevalence esti- mates of psychiatric disorders in correctional settings. In: The Health Status of Soon-To-Be-Released Inmates: A Report to Congress, Vol. 2 (pp. 57–80). Chicago, IL: National Com- mission on Correctional Health Care.
Walters, G. D., Mann, M. F., Miller, M. P., Hemphill, L. L., & Chlumsky, M. L. (1988). Emotional disorder among offend- ers. Criminal Justice and Behavior, 15, 433–453.
263
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.