A P P L I E D

F O R E N S I C S

2000 P Street, N.W., Suite 405
Washington, DC 20036

L L P
Tel: (202) 365-1387
Fax: (202) 543-3025

Keith R. Curry, Ph.D.
2000 P Street, N.W.
Suite 405
Washington, DC 20036
Tel: (202) 365-1387
Fax: (202) 543-3025

March 19, 2002
Donna Brorby, Esq.
Law Office of Donna Brorby
660 Market Street
Suite 300
San Francisco, CA 94104
Dear Ms. Brorby,
I have completed my site visits regarding Ruiz et al v. Gary Johnson et al.,
having surveyed eight prison units on a total of 15 site days. These included Robertson
on January 31st and February 1st; Allred on February 2nd and 3rd; Clements/PAMIO on
February 4th and 5th; McConnell on February 26th and 27th; Lewis/AMPP on February
28th and March 1st; Estelle High Security on March 4th; Coffield on March 5th and 6th; and
Smith on March 11th and 12th.
Before the site visits, I reviewed various documents provided by your office
relating to the provision of mental health services to inmates housed in Administrative
Segregation throughout the Texas Department of Criminal Justice (TDCJ). These
included the following:
§

Notice of Filing of Defendants’ Intervention Plan for Seriously Mentally-Ill Offenders
in Administrative Segregation, dated December 14, 2001

Re: TDCJ
March 19, 2002

§

TDCJ Health Services Division (HSD) Policy E-39.1, Health Evaluation and
Documentation – Offenders in Segregation

§

TDCJ Mental Health Services (MHS) Policy IV J, Pre-Segregation Mental Health
Evaluation

§

TDCJ MHS Policy IV J.1, Mental Health Evaluation of Offenders in Disciplinary
Segregation

§

TDCJ MHS Policy K, Segregation Rounds – Mental Health Patients

§

TDCJ MHS Policy II.A, Referral of Offenders to Psychiatric Inpatient or Crisis
Management Facilities

§

Texas Tech University Health Services Center (TTUHSC) pre-service training
curriculum relating to mental health issues.

§

TDCJ pre-training curricula used for correctional officer Mental Health Training

§

A unit summary of transfer data resulting from administrative segregation reviews
conducted by TTUHSC and the University of Texas Medical Branch (UTMB) staff
from 1999 through 2001

§

Admission and referral procedures for the Program for the Aggressive Mentally Ill
Offender (PAMIO) program operated by Texas Tech

§

A program description for the Administrative Segregation Maintenance
Psychiatric Program (AMPP) proposed by UTMB

§

Smith Unit In-Cell Psycho-Educational Treatment Modules proposed by TTUHSC

§

Minutes of Quality Council Meeting of July 12, 1999 conducted by UTMB

§

Administrative Segregation Review staff assignments and résumés

2

Re: TDCJ
March 19, 2002

The survey included interviews and informal discussions with prison
administrators, managers, correctional staff, and medical and mental health staff. I
interviewed staff working the day and evening shifts and observed daily routines on
the units. Environmental conditions were observed on both preliminary guided tours
and later over the course of my stay on the units.
During my visit I interviewed selected inmates and reviewed the medical
records of inmates culled from the Mental Health rolls and others based upon their
presentation when found in administrative segregation housing areas. A complete
list can be found in Appendix A.
The following conclusions are based upon review of institutional documents,
meetings and interviews with TDCJ staff, on-site observations, inmate interviews,
and review of a sample of medical records.
General Impressions
It has been well established that seriously mentally ill inmates are
disproportionately represented in correctional facilities in the United States. Once
incarcerated, inmates suffering from schizophrenia, schizoaffective disorder, bipolar
disorder, and major depressive disorder display predictable deficits in behavioral and
emotional control, maladaptive interpersonal styles, social skills deficits, and
distorted perceptions of their environments. As a result, they are less able to
conform their behavior to the rigid expectations of prison life and often fall into selfdefeating patterns of irrational opposition to the demands placed upon them.
Seriously mentally ill inmates are thus more prone to disciplinary infractions and
once segregated, react more negatively to the relative inactivity and sensory

3

Re: TDCJ
March 19, 2002

deprivation of 23 hour a day lock down. As external reality clues recede, their
mental functioning often deteriorates with concomitant restriction of their already
inadequate coping skills. In the absence of active mental health treatment, seriously
mentally ill inmates may become the “bottom dwellers” of the prison system, trapped
in segregation units by their illness and unable to adapt to the hard conditions found
at the deep end of the correctional system.
The current limited review of the mental health status of segregated inmates
in seven TDCJ prisons yields findings of 10 to 15% identified by TDJC as suffering
from serious mental illness.1 The clinical effects of segregation upon these seriously
mentally ill inmates varied substantially throughout the state depending upon local
factors such as mental health staffing patterns, quality and quantity of care, housing
design, and proximity to an inpatient setting. However, one factor was consistent
across all facilities visited. Seriously mentally ill inmates were subject to very
lengthy stays in segregation. Of the 68 inmates reviewed for whom the length of
stay could be roughly estimated from the medical record, the average length of stay
in segregation appeared to be 5.2 years with a range of one month to 17 years.2
Administrative Segregation
The structure of administrative segregation at TDCJ presents obstacles to the
effective treatment of severely mentally ill inmates and has created de facto long1

The PAMIO program at the Clements Unit was excluded from this calculation since it houses exclusively severely mentally
ill inmates.
2
Segregation lengths of stay were estimated based upon dated medical records including pre-segregation mental health
evaluations, weekly mental health segregation rounds, and quarterly mental health progress notes and segregation clearance
statements. Precise data could come only from classification documents. These have been requested of the attorney
general’s office at the time this report is due. Supplemental findings will be issued to define the lengths of stay with greater
precision. (“Length of stay” as used here is technically a misnomer since this calculation only sums the number of days spent
in segregation to date. True length of stay data was not available and could only be calculated by summing the total number
of segregation days upon release to the general population or from prison. This latter calculation would by definition produce
significantly higher numbers.)

4

Re: TDCJ
March 19, 2002

term mental health housing. Administrative segregation employs a three level
system that top TDCJ officials explained to us is meant to deter disciplinary
infractions by already segregated inmates. By policy, inmates can enter
administrative segregation on any level depending upon the nature of the infraction.
Once they are assigned to administrative segregation may be dropped to level III for
any disciplinary case involving assault.3 Level II and III assignments entail
significant restrictions and deprivations as compared to Level I.
Inmates on Level I are allowed a minimum of 7 and a maximum of 12 hours
per week of out-of-cell time in relatively barren recreation areas. They do not have
access to television or even in-cell education, by state law. Their visitation privileges
are similar to those of general population. They are provided desserts with meals
and are allowed personal property and commissary privileges similar to general
population.
Approximately 30% of all administrative segregation inmates are assigned to
Levels II and III. Most of the seriously mentally ill inmates interviewed were on one
of these levels. Level II inmates are allowed a maximum of four hours of out-of-cell
time per week. Level III inmates are permitted only three hours of out-of-cell time
per week to generally single-person recreation enclosures, plus daily showers which
are in-cell at high security units and out-of-cell at other units. Commissary and
property privileges are severely restricted compared to Level I. Level III inmates are
not allowed to buy or possess shampoo, toothpaste, deodorant, or commercial soap.
Inmates are allowed two visits per month on Level II and one visit per month on
Level III. None of these can be contact visits. Inmates on Levels II and III do not get
3

Assault in this instance has been broadly defined to include public masturbation and vague threats.

5

Re: TDCJ
March 19, 2002

dessert with their dinner. Any administrative segregation supervisor can place them
on property restriction or food loaf for 24 hours for any reason.
Once assigned to Level III, inmates must remain free of major disciplinary
cases for 90 days to be eligible ascend to Level II.4 TDCJ policy does not prescribe
a period for Level II, but 180 days was a common minimum on the units visited.
Hence it appears that an inmate assigned to Level III must remain free of major
infractions a minimum of 9 months before he can return to Level I.
Whether or not this system effectively serves as a deterrent to the typical
administrative segregation inmate, its deterrence value for seriously mentally ill
inmates is dubious. Inmates suffering from schizophrenia, schizoaffective disorder,
bipolar disorder, psychotic depression, and severe forms of personality disorders
have, as a part of their illness, poor impulse control, delusional thinking, volatile
emotions, distorted perceptions of their environments, as well as gross social skills
deficits. They typically suffer serious disabilities in terms of planning for future
events and learning from experience. Once segregated, they may lose the ability to
track the passage of time. Disoriented and confused psychotic inmates often
misinterpret the muffled voices of staff and other inmates in ways that confirm their
fantasies and fears, or they may suffer outright hallucinations. Those suffering from
paranoia typically misconstrue the motives of others in ways that prevent them from
acting in their own best interest. Once cut off from external reality cues due to social
and sensory deprivation, psychotic inmates may become autistic and lose the ability
to differentiate events occurring inside versus outside of themselves. Their behavior

4

In contravention of this TDCJ policy, the Clements Unit was requiring 180 minimums at Level III.

6

Re: TDCJ
March 19, 2002

can become bizarre and erratic. Hopelessness, apathy, and disturbances of volition
are common symptoms of these disorders.
Given these predictable symptoms, it is plain that seriously mentally ill
inmates differentially lack the ability to understand, internalize, and react
appropriately to the disincentives of this level system. Seriously ill inmates are
overrepresented in the lower levels of administrative segregation and the long
periods spent mired there can be attributed to the serious symptoms of their mental
illness. In a circular fashion, the extreme social and sensory deprivation of
segregation in turn exacerbates those same symptoms that have kept these inmates
stuck at the bottom.
Many of these symptoms could be ameliorated through aggressive mental
health treatment. Most of these self-defeating behaviors are susceptible to behavior
therapy interventions that have been proven effective in correctional facilities.
However, one of the effects found in reviewing administrative segregation
assignment at TDCJ is a de facto reduction in access to mental health care. In
effect, the inmates most in need of mental health services were least able to get
them.5
The outcome is a system where seriously mentally ill inmates often enter
administrative segregation early in their sentences and either start at or quickly fall to
Level III where they become trapped by the effects of their mental illness. These are
the “bottom dwellers”, many stuck in segregation for the duration of their sentences.

5

Acuity is not the issue here. The need expressed here has to do with the direct injurious effects of their illnesses on the
sufferers and their life circumstances. Hospitalized patients may be suffering from through a more acute phase of their
illness, but the therapeutic environment of the hospital lessens the immediate impact and interrupts self-defeating cycles.

7

Re: TDCJ
March 19, 2002

They behave irrationally, have frequent crises, and cycle in and out of prison
psychiatric hospitals.
Mental Health Service Model
Traditional mental health care systems are comprised of a minimum of three
elements. Acute care is intensive round the clock hospital level service for patients
suffering from the acute phase of a mental illness where symptoms of psychosis,
imminent suicide risk, or dangerousness justify intensive and intrusive measures and
curtailment of personal liberties such as forced medication and the application of
locked door seclusion or mechanical restraints. Sub-acute care is typically provided
outside of hospital settings for people suffering from severe and chronic conditions
that require intensive case management, psychosocial interventions, crisis
management, and psychopharmacology in a safe and contained environment in
order to prevent painful and dangerous deterioration in their conditions that would
otherwise lead to repetitive cycling in and out of acute care hospitals. Outpatient
care is provided in the general community and typically involves supportive
counseling, psychotherapy, and other palliative interventions for relatively healthy
individuals experiencing psychological symptoms due to mild forms of mental illness
or adverse reactions to difficult life circumstances. Outpatient care may also be
appropriate for chronically mentally ill inmates whose symptoms are under control or
have gone into remission to such an extent that they can function relatively normally
in segregation or general population. (See Appendix B for descriptions of Mental
Health Service Models from other departments of corrections)

8

Re: TDCJ
March 19, 2002

The contracted mental health providers for the TDCJ appear to offer the
necessary level of acute care through a sizable prison hospital system.6 Periodic
sweeps over the past two years have located and removed many seriously
decompensated inmates from administrative segregation areas thereby improving
provision of acute care services to this segment of the population. TDCJ and its
contractors have committed to continuing these sweeps on a biennial basis with the
assistance of an outside consultant.
Sub-acute care within the TDCJ system is limited to two programs at one site.
The state developed a non hospital program, the Program for Aggressive Mentally Ill
Offenders (PAMIO), in 1991 at the Clements Unit. A separate service now called
the Personality Disorder/Aggressive Behavior Unit evolved out of PAMIO to serve
inmates with Axis II disorders.7 Beyond these two programs, the TDCJ offers no
sub-acute care to inmates who require this level of care.
Outpatient care provided within administrative segregation units is inadequate
in its own right, but would not even under the best of circumstances be sufficient to
meet the needs of inmates requiring sub-acute care. The deficiency is not in the
quality but in the type of treatment provided. Outpatient care is appropriate for
relatively healthy inmates having difficulty adjusting to prison life or chronically ill
inmates who have become essentially asymptomatic. Under normal conditions, the
quality of outpatient care would not be relevant to the needs of sub-acutely mentally

6

Although the acute care system was not surveyed for this report, opinions of plaintiff’s experts in 1999 as well as currently
stated program descriptions and census figures indicate sufficiency at this level of care.
7
The “Personality Disorder/Aggressive Behavior Program” was formerly known as the “Step Down” program and is so
called in the Notice of Filing of Defendants’ Intervention Plan for Seriously Mentally-Ill Offenders in Administrative
Segregation.

9

Re: TDCJ
March 19, 2002

ill inmates any more than the quality of sub-acute care would be relevant to the
needs of acutely ill inmates. However, in administrative segregation units of TDCJ
of sub-acute care is nonexistent and inmates requiring this level of service are
offered outpatient care instead. As a result, inmates in this group of sub-acutely
mentally ill inmates are harmed by their long-term assignment to a housing status
where the appropriate level of care necessary to treat their serious illness does not
exist.
The mental health care providers have acknowledged the absence of subacute care for segregated inmates and have proposed new programs meant to
supplement existing outpatient services for inmates within administrative segregation
who do not meet criteria for acute inpatient care. They have not proposed new
facilities, programs, or housing for segregated inmates requiring sub-acute mental
health care to be provided in some form of restricted housing other than regular
administrative segregation.
Unit findings
Overall, the quality of outpatient mental health care delivered to inmates
surveyed in segregation ranged from adequate to virtually nonexistent. As
mentioned, sub-acute care is simply absent from the system outside of PAMIO and
the Personality Disorder/Aggressive Behavior Program at Clements. In most prison
mental health delivery systems, severely and chronically impaired inmates not in
need of acute care hospitalization receive sub-acute care in the form of enhanced
mental health services in specialized custody settings to prevent painful and
expensive decompensations. In the TDCJ model, these inmates were found to be

10

Re: TDCJ
March 19, 2002

receiving the bare minimum of outpatient care in regular administrative segregation
housing. This reactive model has allowed and even required inmates to repeatedly
deteriorate to the point of rehospitalization in order to receive mental health care
beyond the outpatient level. The medical records surveyed at all units revealed
numerous examples of minimally treated mentally ill inmates decompensating in
segregation, coming to the attention of unit mental health staff during a crisis,
transferring to an inpatient setting for treatment, only to be released back to
segregation to have the cycle repeated. Outpatient care itself is barely adequate in
segregation due to low and variable caseloads, inadequate and uneven staffing,
absent or irrelevant individualized treatment planning, serious and persistent
problems with medication administration, and most importantly, the substitution of
monitoring for treatment. These conditions were found in varying degrees at all
segregation units surveyed. Training, supervision, and enforcement of policies and
procedures were uniformly weak. Where adequate care was found, it was directly
attributable to the presence of one or two dedicated staff members.
Basic identification, medication, and monitoring of mentally ill inmates in
segregation constitute almost all of the mental health services rendered in
administrative segregation, but these were found to be inconsistent from unit to unit.
Although Allred, McConnell, Estelle High Security and Coffield were observed to be
doing a good job finding and tracking severely mentally ill inmates, they each
constituted examples of isolated personnel factors determining the quality of service.
In each of these facilities, the mental health administrators had developed efficient

11

Re: TDCJ
March 19, 2002

but idiosyncratic ways of functioning with inadequate resources and all utilized
frequent inpatient transfers to care for their chronic and severely mentally ill inmates.
Allred for instance has a staffing level of 8.5 full time equivalent (FTE) mental
health clinicians that is relatively high compared to other units visited and maintains
a caseload of 148 out of 950 or 15% of the inmates in segregation.8 The apparent
treatment model at Allred is to cast the widest possible net and then spread staff
resources thinly. Inmate interviews and medical record reviews reveal a lower acuity
threshold for obtaining mental health services than found elsewhere, but also
widespread complaints that individualized treatment is not available. Furthermore,
inmates report slow responses to requests submitted for mental health attention.
Although a significant number of severely mentally ill inmates were found in Allred’s
segregation housing, all appeared to be stable and receiving the most basic care9.
However, basic outpatient care at Allred, as at all segregation units visited, consisted
almost exclusively of brief 90 day visits for medication management, cursory rounds
conducted cell side weekly or sometimes monthly, and crisis management. This
passive treatment model of waiting for inmates to deteriorate to the point of needing
hospitalization essentially omits active treatment during the sub-acute phase of
chronic mental illness. Outpatient treatment necessary to promote stabilization,
symptom management, and adaptive functioning was simply not available to

8

Mental health staffing levels cited are not exclusive to administrative segregation, but to each facility as a whole, i.e. Allred
has a total census of 3,150 with 8.5 mental health clinicians.
9
The pattern of crisis referrals prior to our February 3, 2002 site visit is noteworthy. After an average of 8.3 crisis transfers
from segregation and general population per month from March, 2001 through December, 2001, Allred transferred 22
inmates for crisis management in the month preceding our site visit. Since the TDCJ mental health treatment model is
designed to find and remove seriously ill inmates rather than treat them on site, a snapshot taken following a purge would
underestimate the level of acuity occurring naturally at the facility. Data available on-site did not allow for analysis of
composition by housing area so that the numbers of referrals from administrative segregation versus general population could
not be determined.

12

Re: TDCJ
March 19, 2002

outpatient inmates in administrative segregation.10

Since chronically and seriously

mentally ill inmates are left to decompensate until they are captured by the mental
health monitoring system, the main role of outpatient mental health providers in this
system is not to provide treatment, but to look for inmates who have deteriorated
beyond a certain threshold and refer them out.
At Allred, the mandated weekly mental health rounds in segregation have
shown significant fluctuations in quality due to recent personnel changes. A review
of medical records substantiates staff and inmate reports that the quality and
regularity of segregation rounds dropped dramatically when one key clinician retired
in the fall of 2000. Prior to that, these rounds were conducted door to door as
required and produced meaningful information for the staff monitoring for
deterioration. Inmates report that for the past year, mental health rounds have been
conducted in the “nursing style” which involves a clinician stepping into the housing
area, shouting, “Psych”, speaking briefly with any inmate who shouts back that he is
mentally ill, and then departing the cellblock. This change in practice due to change
in personnel is typical throughout the system whereby the quality of services is the
product of varying levels interest and effort on the part of staff. Although this
practice was not directly observed, documentation of mental health rounds has
shown a clear deterioration in quality over that past 18 months.
The Coffield Unit has the second highest staffing level with 7 FTE mental
health clinicians for a total census of 3,161. Despite this, their administrative

10

TDCJ currently maintains three specialized programs meant to address narrow aspects of this need. However, while the
PAMIO program has operated since 1991, AMPP at the Lewis Unit and the In-Cell Psycho-education Program at the Smith
Unit are in their embryonic phase and operate on a small scale. These attempted remedies will be discussed in a following
section.

13

Re: TDCJ
March 19, 2002

segregation caseload is significantly lower with 66 of 696, or 9% of segregation
inmates receiving care. Interviews and record reviews reveal that the staff at
Coffield are providing a higher quality of care than found elsewhere, but to inmates
within a more restricted range of acuity. They are doing a good job of identification,
assessment, and monitoring of the most disordered inmates. Furthermore, Coffield
was one of only two units visited that provides regularly scheduled individual case
management in offices rather than at cell side. The responsibility for mental health
rounds is rotated among staff to reduce habituation, resulting in meaningful
observations that enhance monitoring. Interviews with the 25 most seriously
mentally ill segregation inmates found none who were decompensated or otherwise
in need of a higher level of care. Dramatic changes in the nature of crisis referrals at
Coffield, as compared to other units, over the past three years demonstrates the
impact of isolated personnel changes. During a three month period spanning July
through September, 1999, Coffield staff transferred 135 inmates from the facility as a
whole to crisis beds. Of these, 129 were for self-mutilation and seven were for
psychosis. This stands in stark contrast to data from the three months spanning
December, 2001 through February, 2002, during which only 28 inmates were
transferred, nine for self-mutilation and 19 for psychosis. While it is commendable
that self-mutilation has diminished so dramatically and surveillance of psychotic
decompensation has improved, the absence of sub-acute care is still evident in the
need to transfer on average six psychotic inmates per month to inpatient care.11

11

Although data was not reported in a manner that allowed for analysis of administrative segregation versus general
population crisis referrals, the supervisor of mental health services at Coffield stated that he believed crisis referrals come
disproportionately from segregation.

14

Re: TDCJ
March 19, 2002

The only other notable deficiencies found at Coffield were in medication
administration and the filing of laboratory results. Pill call practices observed here
and elsewhere are not consistent with stated TDCJ policy and are generally unsafe.
Medications were observed being dispensed cell side in paper cones passed
through the bars of the cell door with no attempt to observe the inmates. The nurses
observed sometimes did not even break stride as they proceed down the tier. This
practice allows inmates to do whatever they please with the medication, including
hoarding or disposal, and precludes any meaningful monitoring of compliance. This
happened despite the advantage presented by Coffield’s bar front cells. Nurses at
other facilities with solid doors were observed dropping the medication through the
slots, setting the medication on the slot shelves and walking away, and in one case
kicking the cones of medication under the cell doors. These practices hamper the
effective care of treatment resistant inmates and create a market for dangerous
drugs within the facilities. A reliable indicator that an inmate is disposing, hoarding,
or selling his medications can be found in laboratory results for medications such as
Lithium, Depakote, or Tegretol. However, even though lab tests are reliably ordered
by the psychiatrist at Coffield, the results were not filed three months after blood was
drawn in more than half of the records surveyed.12
The McConnell Unit employs 5 FTE mental health clinicians with one vacant
Mental Health Liaison position for a total of 2,800 inmates. Psychiatry is provided
via Telehealth due to difficulty recruiting qualified psychiatrists locally. There are 75
out of 500, or 15% segregation inmates on the mental health rolls. There is some
12

Obtaining critical laboratory tests was more problematic at other units where they were not reliably ordered by treating
physicians.

15

Re: TDCJ
March 19, 2002

evidence that McConnell staff are adequately identifying and monitoring seriously
mentally ill inmates in segregation and transferring them to inpatient care. Staff
report that they are able to occasionally see inmates in an office just off the cellblock
for crises or individual case management, but admit that most mental health visits
happen briefly at cell side. The custody staff at McConnell appear to use the mental
health providers to diffuse potential crises through verbal interventions cell side in a
commendable way not seen at other units. Mental health rounds are conducted
weekly and produce useful case management and referral information. However,
meaningful treatment beyond tracking and crisis intervention was rare. Out of 25
medical records reviewed, one had a barely adequate treatment plan while the other
24 contained treatment plans that were uniformly meaningless and outdated. While
it may be said that evidence of individualized treatment planning does not guarantee
good treatment, it is widely and in many cases officially recognized that the absence
of individualized treatment planning suggests seat of the pants treatment.13 This is
an important issue in a system with poor adherence to policy and weak enforcement
of standards. This lack of individualized treatment planning was uniform at all
administrative segregation units visited and accurately reflects the absence of
systematic treatment offered at the outpatient level.
McConnell staff relied more heavily upon crisis management services than
any at any other unit, sending out 91 crisis transfers from the facility as a whole from

13

Virtually all major accreditation and governance bodies overseeing organized mental health care in the United States
require individualized treatment planning. These include the Health Care Financing Administration, the Joint Commission
for the Accreditation of Healthcare Organizations, the National Commission on Correctional Health Care, the Federal Bureau
of Prisons, and the United States Department of Justice in its governance of local jails holding federal detainees.

16

Re: TDCJ
March 19, 2002

September 1, 2001 through February 28, 2002.14 Thirty of these came from the 500
administrative segregation beds while 61 came from the 2,300 general population
beds. Nearly half, 44, were repeat referrals. Nineteen were for psychotic
decompensation, and 14 of these came from segregation. One was a segregated
inmate with paranoid schizophrenia who bounced back and forth to the hospital five
times in six months. Eight of the 30 originally reported segregation referrals came in
the month preceding our site visit, and none of the seriously mentally ill repeat
referrals was in the facility upon our arrival. Conclusions based upon the fact that
these numbers are higher than elsewhere are difficult to draw since neither the most
seriously ill inmates nor their medical records were available for inspection.
McConnell has the only psychiatric observation bed observed at any of the units
visited except for Clements, giving them the relative luxury of isolating and
monitoring inmates in crisis for short periods of time in house. The numbers suggest
that the staff at McConnell are identifying seriously mentally ill inmates and
transferring them to the hospital when they decompensate. The staffing is simply
not sufficient to provide care much beyond monitoring, crisis intervention, and
hospital transfers.
Estelle High Security provided the best example of small scale staff
characteristics determining the quality of care provided. The staff consists of one
full-time counselor, .20 FTE of a psychiatrist, and .20 FTE of a psychiatric LVN

14

Data compiled 02/27/02 by McConnell mental health staff on-site indicated 91 crisis transfers for this period, however,
cumulative data provided by the Attorney General’s office on 03/21/02 adjusted this number to 105 crisis transfers. Of the 14
adjusted crisis transfers, eight were added to February, 2002. Although no clinical details are available regarding the 14
additional transfers, there is evidence to suggest that as many as 13 inmates may have been transferred from McConnell to
crisis management in the seven days preceding our site visit.

17

Re: TDCJ
March 19, 2002

dedicated to the 500 bed High Security Unit.15 Although the mental health rolls are
light, 44 of 500 or 9% of segregated inmates, little indication was found of
unidentified or untreated seriously mentally ill inmates in segregation. The mental
health staff is vigilant and refers seriously mentally ill patients to the hospital with low
threshold criteria. There were 69 crisis transfers during the six month period
spanning September 2001 through February 2002.16 Out of 45 of these for which
clinical details were available, only one inmate had multiple referrals. Mental health
follow-up for inmates on the caseload is excellent. As at Coffield, inmates are seen
reliably for individual case management held in offices off the cellblock and weekly
mental health rounds are rotated among staff to prevent complacency. The one
counselor and his off site supervisor are both well suited to working with inmates in
segregation. Both are conscientious and caring, and neither engages patients in
iatrogenic power struggles. Requests for mental health interviews are addressed
quickly with sound outcomes. Extensive reviews of the medical records of patients
seen on sick call for mental health issues that were determined not to be in need of
follow-up yielded no false negatives. Inmate interviews in segregation yielded no
previously unidentified psychotic inmates.17 Inmates rate the mental health staff
high on caring and reliability. It appears that the care provided to mentally ill inmates
at Estelle is the product of the skill and dedication of one conscientious counselor.

15

Staffing data must be analyzed differently at units where administrative segregation is housed exclusively in High Security
Units. For example staffing for Estelle, Lewis, and Smith count staff that provide service only for those inmates assigned to
administrative segregation and close custody whereas staffing at Robertson, Allred, McConnell, and Coffield count staff that
provide services to both segregated and general population inmates.
16
In contrast to other units, the rate of crisis management transfers at Estelle did not rise in the month preceding the site visit.
17
It is worth noting that canvassing for mentally ill inmates involved only one psychologist covering 500 inmates and was
limited to less than four hours. Furthermore, the architecture at the high security units (Estelle, Lewis, Smith) complicates
discovery of unidentified mentally ill inmates. The cells have thick, solid doors with glazed apertures. There are no
windows to the outside so that an inmate in his cell with the light off cannot be adequately assessed.

18

Re: TDCJ
March 19, 2002

As at other units, individualized treatment plans in the medical records were devoid
of meaningful guidance and many were as much as four years old. Psychiatric care
is the only other notable weakness at Estelle. The paucity of hours and the lack of a
consistent presence leads to situations where inherited diagnoses are continued
blindly due to insufficient opportunity for comprehensive assessment. Cell side
interviews yielded a significant number of likely misdiagnoses yielding unnecessary
treatment with potentially harmful medications.
In contrast to the above mentioned units, Robertson, Lewis, and Smith Units
were found to be failing at even the basic services. The deficiencies at the
Robertson Unit may be attributed to local issues. However, very serious failures
were found at the Lewis and Smith Units which illuminate systemic problems.
Compared to other units of comparable size, Robertson had slightly fewer
staff with 5 FTE mental health positions, two of which were vacant, to cover a prison
housing 2,800 inmates. Of the segregated inmates only 19 out of 504 or 4% were
identified as in need of services. The psychiatrist and psychologist on-site saw
nothing wrong with this and expressed certainty that the mental health needs of the
segregated inmates were being addressed. Medical record reviews and cell side
interviews did not bear this out.
Of the 19 segregation inmates on Robertson’s mental health rolls, 7 were
found to be significantly neglected. These included a 45 year old man with chronic
paranoid schizophrenia found in a decompensated psychotic state. His thinking was
grossly disorganized and his speech was irrelevant. He appeared confused,
agitated, and paranoid. The medical record indicated that his antipsychotic

19

Re: TDCJ
March 19, 2002

medication had been discontinued on September 24, 2001 due to refusals. There
could be no question that his decompensation was long and tortured. He clearly
belonged on an inpatient psychiatric unit. A 28 year old man with schizoaffective
disorder and mental retardation had his antipsychotic medication discontinued on
January 4, 2002 following a verbal altercation with a nurse over drawing blood for
laboratory tests. Upon interview, he appeared floridly psychotic and deteriorating.
He reported that he had repeatedly requested to be placed back on medications and
had offered to have blood drawn from his right arm. He showed his left arm from
which the nurse wanted to draw blood. His left arm was severely mutilated from
multiple self-inflicted lacerations. A 37 year old man with chronic paranoid
schizophrenia was found in a floridly psychotic state despite receiving long-lasting
injectable antipsychotic medication once a month. He presented with severe
Parkinsonian side-effects from his medication. He reported that mental health staff
conducted cursory rounds once a month, but did not inquire about medication sideeffects. A 34 year old man with “Inhalant Psychosis” was found not receiving
treatment due to refusal. He presented as dirty and disheveled with paranoid
delusions, inappropriate affect and bizarre speech. He required inpatient care but
was instead receiving no psychiatric care other than questionable mental health
rounds once per month. A 34 year old man with chronic paranoid schizophrenia was
found appearing psychotic in his cell. He refused to be interviewed. A review of his
medical record revealed that he was not prescribed any psychotropic medications.
Two other inmates taking antipsychotic medications appeared to psychiatrically
stable, but presented with moderate to severe Parkinsonian side-effects that were

20

Re: TDCJ
March 19, 2002

not being adequately addressed. These side-effects are painful and debilitating,
requiring immediate medical attention. Despite being on a remarkably small
caseload, none of the above patients were receiving adequate mental health care.
In the brief time allotted to canvassing administrative segregation, seven inmates
were found with serious mental illness who either had not come to the attention of
the mental health staff or had been seen and denied services. One of these had
self-inflicted a two very serious lacerations requiring more than 50 staples on
January 21, 2002. When interviewed by the writer, the inmate reported having
current suicidal ideation and a concrete plan to kill himself. When asked, he
volunteered that he had a razor blade in his cell with which he intended to commit
suicide. A subsequent cell search produced the razor which had been hidden under
his mattress.
The Robertson Unit had by far the lowest incidence of crisis management
referrals over the past 12 months compared to comparable facilities. This is likely
due in part to the fact that the mental health staff does not conduct the mandated
weekly segregation rounds and does not respond appropriately to requests for
mental health treatment. A more thorough review would be required to fully assess
the impact this neglect is having on mentally ill inmates housed in segregation at
Robertson. However, based upon the available prevalence rates at comparable
facilities and the low rate of identification and hospital referral at Robertson, it is
probable that a significant number of seriously mentally ill inmates have been and
are being medically neglected in administrative segregation there.

21

Re: TDCJ
March 19, 2002

The Lewis Unit was selected for site visit in order to evaluate the
“Administrative Segregation Maintenance Psychiatric Program” (AMPP) begun there
in November, 2001. This program was initiated by UTMB staff to provide some
portion of the missing sub-acute care necessary for chronic, seriously ill inmates
who do not require inpatient hospitalization. Staffing for this program as described in
the Notice of Filing of Defendants’ Intervention Plan for Seriously Mentally-Ill
Offenders in Administrative Segregation, dated December 14, 2001 was consistent
with in-brief discussions held with UTMB staff on site. However, discussions with
AMPP staff revealed that no other mental health staff had been assigned to provide
care for the approximately 800 administrative segregation and close custody
inmates housed at Lewis. The AMPP staff reported that since their arrival in the fall
of 2001, they had been providing all of the mental health care for the High Security
building. When asked about this, the Director of Mental Health Services for UTMB
asserted that AMPP had been staffed based upon the premise that TDCJ had been
prohibited from housing mentally ill inmates at Lewis other than inmates who
developed psychiatric disorders after admission or those referred expressly to the
AMPP program.18 The Director produced a TDCJ inter-office communication dated
March 15, 1995 from the Health Services Liaison to the Classification Committee
detailing units which lacked sufficient mental health staff to maintain outpatient
caseloads. Lewis (and Smith) were designated as such. The Classification
Committee was asked not to assign inmates with PULHES19 scores higher than
S1AP to these units since there were no mental health staff on site. A review of
18

The AMPP census was 19 at the time of the site visit.
PULHES is a TDCJ classification tool designed to categorize severity of risk in various domains. A PULHES score of
S1AP would indicate that the inmate had no psychiatric problems.

19

22

Re: TDCJ
March 19, 2002

medical records quickly produced 40 severely mentally ill inmates exclusive of
AMPP scattered throughout the facility. All of these inmates were admitted with
PUHLES scores greater than S1AP. Many had PUHLES scores of S3NT upon
admission, indicating moderate to severe psychiatric problems. Several had
PUHLES scores of S4PT indicating psychiatric problems requiring inpatient
treatment. Ten inmates were found to have been transferred to Lewis directly from
psychiatric inpatient hospitals. One inmate was transferred to Lewis while on suicide
monitoring at another unit. Cell side interviews turned up a substantial number of
psychotic inmates. One man had a history of frequent psychotic decompensations
leading to inpatient admissions. Another had been transferred for inpatient care
twice during his stay at Lewis. A third man with paranoid schizophrenia had been
transferred from Lewis to an inpatient unit in February, 2002, and returned to Lewis
within the past seven days. One severely ill interviewee reported that his cell mate
had been wildly psychotic for months, screaming at the mirror throughout the night,
refusing to bathe, and ranting about paranoid delusions. Upon inspection, the cell
mate was found in a state of florid psychotic decompensation. He appeared
confused and agitated with incoherent speech. His medical record revealed that he
had denied the need for psychiatric care and therefore had not been treated for his
psychosis. None of the inmates identified were receiving adequate care whether
they were on the mental health rolls or not. This is not surprising since the staff
assigned to treat them (prescribe medication and conduct cell side rounds once a
month) were hired specifically to staff the AMPP program and had been given these
responsibilities beyond their AMPP duties. Many of these inmates were admitted to

23

Re: TDCJ
March 19, 2002

Lewis many months before the AMPP staff arrived in the fall. It is unclear what
mental health staff had provided services at this unit nominally designated for
“nonmentally ill” inmates only. An average of 8.1 inmates per month were
transferred from all of the Lewis Unit to hospitals for crisis management from March,
2001 through January, 2002. Interestingly, the number of crisis transfers nearly
doubled (15) for the month of February, 2002. None of this reflects negatively upon
the AMPP program except to the extent that the proposed staffing numbers are
effectively diluted by the need to provide routine mental health services to a
relatively large non-AMPP segregated population. Unfortunately, the staffing
proposed for the AMPP program as it is currently configured could not possibly be
sufficient to cover the needs of this other population in addition to the highly
specialized requirements of the AMPP program. (This will be discussed further in
the program descriptions to follow.)
The Smith High Security Unit houses 800 inmates, 450 in administrative
segregation with the remainder in double bunked close custody. Approximately 130
out of 800 or 16% are receiving some form of mental health care.20 However, this
relatively high percentage may not be meaningful. First, the mental health staffing is
grossly deficient in two ways.21 Only 1.5 FTE’s are allotted to the High Security
facility in the form of one registered nurse and a half-time physician’s assistant. This
number is insufficient in own right given the size of the overall population and the
known level of acuity. Neither of these clinicians have the training or experience to

20

At the time of the site visit, 12 segregation inmates were enrolled in the In-Cell Psycho-education Program. These were a
subset of the 130 on the mental health rolls. This program will be described in detail below.
21
This staffing might be adequate if the Smith Unit was in fact restricted to “nonmentally ill” inmates as indicated in the
TDCJ memorandum of 3/15/95.

24

Re: TDCJ
March 19, 2002

provide the kind of psychiatric services required. The facility employs no
psychiatrist, psychologist, social worker, counselor, or mental health liaison. While
the full-time nurse is dedicated and conscientious, her psychiatric experience is
limited to one year at a state hospital. She is expected to provide virtually all of the
mental health services to a large population of severely mentally ill inmates in High
Security. She apparently receives no supervision from a clinician with corrections
experience and she has no clerical support. By policy, she is expected to conduct
mental health rounds on 130 inmates every week. Even conducting these rounds
once a month, as she reports she does, little could be expected in terms of
thoroughness. She could hardly be expected to provide even case management or
crisis intervention.
The physician’s assistant works half-time as a medical provider and half-time
prescribing psychotropic medications. His psychiatric supervisor is located
approximately 100 miles away and is infrequently on site. The results are
predictably appalling. Of the 101 inmates prescribed psychotropic medications, 55
were receiving sleep medications of either Amitriptyline or Doxepin often in high
doses. Medical record reviews clearly indicated that these medications were
prescribed for sleep despite universally held beliefs that doing so is contraindicated
in correctional settings. They are prone to abuse and lead to tolerance and
withdrawal. Most importantly though, these medications are lethal in relatively small
dosages and are a common means of suicide. Given the absence of reliable
observation of inmates actually swallowing their medications and the inevitable black
market that develops in facilities where these medications are prescribed, one could

25

Re: TDCJ
March 19, 2002

expect their unrestricted abuse. This was in fact indicated by correctional incident
reports indicating quantities of pills either found on cell searches or taken during
suicide attempts.22 One might expect that vigilance would at least be found in
monitoring of blood levels for these medications to assess compliance and toxicity.
This was not the case. Of 25 medical records reviewed for this purpose, 15
contained no orders for lab tests. Five records contained recent lab results showing
no presence of medication in the blood and two more detected low levels. One
record indicated that blood was drawn on February 12, 2002 but as of March 12, no
results filed in the chart. Just two of the 25 yielded lab results in the expected range.
Prescription practices for other psychotropic medications were equally
questionable. For instance, a 40 year old man with an extensive history of explosive
behavior, frequent disciplinary infractions, and psychotic decompensations who was
prescribed a high dose of a mood stabilizer and the highest allowable dose of an
antipsychotic medication, after having no completed lab results since 1999, had no
presence of medication in his blood when tested in January, 2001. Despite this, no
further lab results could be found in his medical record since he was transferred to
Smith in March, 2001. All psychotropic medications were abruptly discontinued on
March 4, 2002 after the inmate failed to, “get back from door during pill pass on 3/3
and 3/4.” A 28 year old man with explosive behavior and psychotic
decompensations taking a very high dose of a mood stabilizer plus an antipsychotic
and a tranquilizer had nothing detected in his blood on August 22, 2001. These
results were not commented upon in the medical record and the medications were
22

For instance an inmate attempted suicide by swallowing “20-24” Doxepin tablets on January 28, 2001. Another
inmate ingested, “an ample amount of unknown red pills” in a suicide attempt on December 14, 2001. On
December 9, 2001 an inmate attempted suicide by swallowing 30 pills.

26

Re: TDCJ
March 19, 2002

continued with no further lab work ordered to date. A 32 year old man with Bipolar
Disorder and a history of multiple crisis referrals who is prescribed a fairly high
dosage of a mood stabilizer had none detected in his blood on November 13, 2001.
A notation was entered on February 6, 2002 and labs were appropriately if belatedly
reordered. When the inmate was next seen for routine follow-up on March 7, 2002,
no mention was made of the second abnormal finding. The inmate’s medication
continued unchanged and no further labs were ordered. A 30 year old man
diagnosed with Schizoaffective Disorder is prescribed an antipsychotic medication, a
mood stabilizer, and a sleep medication. He last had labs ordered in December,
2000 and was last seen by the physician’s assistant at Smith on October 4, 2001.
The inmate refused his next scheduled psychiatric follow-up appointment on January
31, 2002.23 There are no mental health notations in this inmate’s medical record
since October, 2001 despite his extensive history of psychotic episodes leading to
psychiatric inpatient admissions. The required lab test for this inmate’s mood
stabilizing medication is at least eight months overdue as well.24 And finally, a 20
year old man with Explosive Disorder had required labs ordered for his mood
stabilizer 12 months late on November 15, 2001. As of March 11, 2002, no results
were in the medical record and the inmate is three weeks overdue for his 90 day
follow-up.
In short, the mental health care provided at Smith is of very poor quality and
there is too little of it. The high security architecture and the limit on time allotted for
canvassing door to door precluded a comprehensive survey of mentally ill inmates
23

Even this missed appointment was late. A 90 follow-up would have been scheduled for January 2, 2002.
Commonly accepted practice guidelines for inmates taking Carbamazepine call for serum levels every six months, or more
frequently if there are indications of toxicity, poor compliance or when dosage is changed.

24

27

Re: TDCJ
March 19, 2002

not on the mental health rolls. However, it was possible to locate one profoundly
psychotic man housed in administrative segregation. This involved a 39 year old
man admitted to Smith from Coffield on September 19, 2001 where he had been
treated with a high dose of antipsychotic medication for the diagnosis of Psychotic
Disorder NOS. Since the patient’s medication was discontinued shortly before
transfer, the nurse at Smith did not pick up the psychiatric history upon chain review.
This occurred despite descriptions of extensive prior treatment and an inpatient
admission for bizarre behavior and psychotic decompensation as recently as
January 22, 2001. As a result, no referral was made to mental health. The inmate
was observed on rounds by the psychiatric nurse on October 1st and 11th, 2001 after
custody staff commented upon the inmate’s bizarre behavior. The nurse noted that
the inmate was, “delusional, disorganized, agitated, labile, with rapid speech, flight of
ideas, and loosening of associations.” Despite this, she noted that he was in, “no
apparent distress.” She nonetheless suggested that he see the physician’s assistant
for a medication evaluation. However, the inmate declined this offer and no further
mental health notations were made in his medical record. On the day of the site
visit, the inmate was highly agitated with prominent paranoid delusions. He was
grossly disoriented with rapid speech, loosening of associations, clang associations,
and apparent responses to internal stimuli. This inmate would stand out as severely
impaired on any psychiatric inpatient unit, but was receiving no mental health
services while being locked in a windowless box 24 hours a day for six months.

28

Re: TDCJ
March 19, 2002

Special Programs
The TDCJ has proposed several special mental health programs aimed at
addressing the sub-acute needs of segregated inmates. These include the Program
for the Aggressively Mentally Ill Offender (PAMIO) and the “Personality
Disorder/Aggressive Behavior Program” at the Clements Unit, the Administrative
Segregation Maintenance Psychiatric Program (AMPP) at the Lewis Unit, and the InCell Psycho-Education Program at the Smith Unit.2526 Although the programs at
Clements are well established, the latter two programs are best described as in their
embryonic stage of development and have very small caseloads. Although
impressive claims were made about the scope of these new programs in the Notice
of Filing of Defendants’ Intervention Plan for Seriously Mentally-Ill Offenders in
Administrative Segregation, it is not yet apparent that TDCJ has committed the
necessary resources to ensure either their efficacy or survival.
PAMIO
According to the program description disseminated by TDCJ, PAMIO was initiated in
1991 after the plaintiffs and defendants determined that administrative segregation
was an anti-therapeutic environment for persons with serious mental illness. PAMIO
was designed with the goal of helping mentally ill inmates, “adjust to prison and
control their behavior to the point that they no longer required segregation.” The
program is now established with its own building and a census of approximately 425.
25

The PAMIO program straddles the line between acute and sub-acute care. It is considered inpatient care by TDCJ and like
an acute care hospital requires voluntary consent. In many other ways however, it functions like traditional sub-acute care.
For instance, the program description states that PAMIO in not an acute inpatient unit and does not take “doctor to doctor”
referrals.
26
The “Personality Disorder/Aggressive Behavior Program” was formerly known as the “Step Down” program and is so
called in the Notice of Filing of Defendants’ Intervention Plan for Seriously Mentally-Ill Offenders in Administrative
Segregation.

29

Re: TDCJ
March 19, 2002

Staffing is good with three FTE psychiatrists, seven FTE psychologists (three
Ph.D.’s and four M.A.’s), four FTE social workers (bachelor’s level), four FTE
recreation therapists, two FTE expressive therapists, one FTE occupational
therapist, plus 24 hour a day nursing coverage. A copy of the PAMIO Unit Policy
and Procedure Manual of July 2001 provided on site offers ample evidence of the
thoroughness and maturity of the program. On site observations of both individual
and group therapy supported claims of a high quality of care. The line staff
observed all appeared to be well trained and supervised. Medical record reviews
suggest that PAMIO staff are consistently offering the services promised. The
treatment interventions appear to be conceptually correct for addressing the
problems of aggressive mentally ill inmates. Most importantly, PAMIO has
developed a program of environmental interventions and contingency management
that goes beyond individual counseling and psychotropic medication in the
suppression and management of maladaptive behaviors caused by chronic mental
illness.
To their credit, PAMIO administrators have recognized two program
shortcomings over the past decade and have adapted by spinning off subunits to
address specialized needs. A subgroup of PAMIO patients with severe and
persistent psychotic disorders was known to deteriorate shortly after graduating from
PAMIO and returning to their sending facilities. Over time, a pod was set aside (F
Pod) for their long-term treatment and maintenance without the expectation of
graduation. This, in effect, created a small but important locale dedicated to subacute care for men with intractable schizophrenia and schizoaffective disorders. The

30

Re: TDCJ
March 19, 2002

broader PAMIO environment supported a much higher level of care than currently
available anywhere in the system outside of psychiatric hospitals. Patients were
afforded intensive case management, individual and group therapy, 24 hour
psychiatric coverage, on-site crisis management, and enhanced opportunities for
social skills training and other psychosocial interventions as a result of greater out of
cell opportunities compared to traditional administrative segregation. Without this
level of care, the patients assigned to F Pod would be bouncing back and forth
between inpatient facilities and outpatient services in administrative segregation,
entailing frequent psychotic decompensations, numerous disciplinary infractions,
major uses of force, and potential injuries to themselves and those around them.
A second significant adaptation came in the form of the “Personality
Disorder/Aggressive Behavior Unit” formerly known as the “Step Down” program
housed in E Pod at Clements. While this program is housed contiguous with PAMIO
and shares staff with PAMIO, it is not technically a part of the program. The
Personality Disorder Unit evolved to serve those seriously mentally ill inmates whose
personality disorders precluded their effective treatment at PAMIO, but whose crisis
management needs were of sufficient intensity as to make them unmanageable in
traditional administrative segregation.27 Left to their own devices, these inmates
engage in frequent self-mutilation, parasuicidal behavior, intense and unpredictably
violent behavior, and bring out the worst possible responses from custodians and
caregivers. The treatment model developed at E Pod has shown that through
relatively inexpensive interventions and environmental alterations, these aggressive
27

The inmates assigned to the Personality Disorder Unit typically display a constellation severe symptoms from the cluster of
antisocial, narcissistic, and borderline personality disorders (Cluster B).

31

Re: TDCJ
March 19, 2002

behaviors can be significantly reduced over time. Importantly, assignment to the
Personality Disorder Unit is a custody placement and does not require voluntary
consent. While this treads a fine line in terms of consent to treatment and the right
to refuse intrusive interventions, the alternatives are either untenable in the case of
voluntary admission to a hospital or inhumane in the case of their frequent
abandonment by mental health staff in outpatient settings and the typically draconian
custody measures heaped upon these inmates at the most restrictive end of
segregation. These inmates, often referred to as ‘the worst of the worst’ by
corrections administrators, have demonstrated significant behavioral improvement
as a result of treatment on the Personality Disorder Unit.28
AMPP
The AMPP program at the Lewis Unit is designed to provide sub-acute care
to roughly the same inmate population currently being served at PAMIO’s F Pod.
The UTMB program description aptly portrays the process whereby administrative
segregation inmates are discharged from inpatient facilities in good remission only to
decompensate as outpatients at their assigned unit. AMPP was expressly intended
to serve as a maintenance program that would interrupt the admission-dischargecrisis-admission cycle. The program description aptly depicts this circular model of
treatment as, “staff time consuming, costly and sub-optimal for the offenders’ mental
health care.”
Although the program is conceptually appropriate and benefits from
knowledgeable and dedicated staff, caution is deserved. First, as mentioned in the
28

I interviewed a former E Pod inmate at currently housed at Coffield who complained vociferously about the services he
received at Clements. He nonetheless reported that he had since gone 27 months without a disciplinary infraction and had no
crisis referrals since his discharge.

32

Re: TDCJ
March 19, 2002

above discussion of the Lewis Unit, the promised staffing levels may be illusory
since staff hired specifically for AMPP have been assigned to provide routine care
for an additional 800 segregated and close custody inmates in Lewis High Security.
Even at face value, there is no reason to expect that current or projected staffing
could adequately serve both purposes. Without the statistically necessary staff
dedicated solely to Lewis High Security and a firewall between the two services, the
AMPP program could be expected to degrade into an understaffed, highly restrictive
dumping ground for uncooperative seriously disordered inmates. Second, despite
assurances that the AMPP program was located at Lewis because of the support of
its warden, placing a mental health program at a high security facility creates distinct
limitations. The architecture is a formidable obstacle to providing mental health
care. Lewis’ sound deadening cells with thick doors and glazed apertures make
communication between staff and inmates difficult at best. Group treatments
designed to enhance social skills are necessarily conducted in four adjacent cinder
block holding cells in a hallway. Under these circumstances it is hard to find the
“group” in group treatment since the inmates cannot see or effectively hear one
another and the therapist can only address one inmate at a time. Furthermore, the
profound social and sensory deprivation inherent in Lewis’ windowless cells devoid
of any outward signs of the passage of time and with virtual 24 hour a day isolative
lock down may prove to be deleterious to inmates prone to disorientation,
hallucinations, and states of delirium.
It is unclear in terms of scope whether the AMPP program can make a
significant contribution to the need for sub-acute services in the long run. Nineteen

33

Re: TDCJ
March 19, 2002

inmates were enrolled at the time of the site visit. The program description allows
for expansion to a maximum of 60 inmates.
Staffing projections allow for increases tied to census figures. However,
these increases are not designed to maintain the staff to inmate ratio observed at
the time of the site visit. The ramping up of staff is proposed as follows: 15
offenders to 4 staff; 30 offenders to 5 staff; 60 offenders to 6 staff. In other words,
the inmate to staff ratio goes from roughly 5 to 1 at the present time up to a high of
10 to 1 when all the bed are full.29 Given the cell isolation and limited group
treatment options, it is hard to see where any economy of scale will develop.
In-Cell Psycho-Education Program
The In-Cell Psycho-Education Program demonstrated at the Smith Unit is still
in the pilot phase. It currently provides roughly one hour per week of contact to 12
administrative segregation inmates. Staff for this program do not work at Smith, but
drive the roughly 100 miles down from Montford Hospital in Lubbock every Thursday
afternoon to run the program.30 Each week three out of a pool of eight Montford
clinicians (social workers, recreation therapists, and occupational therapists) make
the trek to spend four hours conducting group interventions in conditions identical to
those described at Lewis. Inmates are encouraged to work on hand-outs between
sessions and these are assigned and reviewed during their out of cell sessions. The
psycho-educational modules are conceptually relevant to segregated inmates,
including such topics as anger, boredom, exercise, hygiene, sleep, stress

29

The AMPP proposal included in the Notice of Filing of Denfendants’ Intervention Plan for Seriously Mentally-Ill
Offenders in Administrative Segregation allows for “temporary reallocation of existing staff existing elsewhere in the
system” pending the hiring of permanently assigned staff.
30
Mental health staffing at Smith is insufficient to provide basic care let alone conduct special programs.

34

Re: TDCJ
March 19, 2002

management, and relaxation. Inmates are assigned specific modules based on
need with an average of 8 sessions per inmate. The materials are professionally
produced and the staff are enthusiastic and committed.
The future of this program beyond the pilot phase is problematic. The current
plan is to export the model to other units with large administrative segregation
populations. However, since none of these facilities are within easy driving distance
of Lubbock, existing mental health staff at each site would be required to learn the
techniques and then implement them in addition to or in place of their current duties.
Program fidelity would be promoted and monitored through periodic supervisory
visits from the program’s founders. Units such as Smith would, unfortunately, be
excluded due to insufficient staff to conduct the weekly sessions.
Under the best of circumstances this essentially unfunded program, which is
designed to provide about eight weeks of in-cell handouts and group interventions,
might be of limited utility to a large population of severely mentally ill inmates stuck
in segregation for an average of more than 5 years.
With the exception of the programs long established at the Clements Unit, this
last issue is relevant for all of the intervention options proposed to address the subacute needs of seriously mentally ill administrative segregation inmates. Even under
the best of circumstances the scope of the proposed services could not significantly
ameliorate the directly harmful effects and exacerbation of psychiatric symptoms
caused by prolonged segregation. It is also doubtful that any of these programs
could achieve the ultimate goal of helping a significant proportion of the seriously

35

Re: TDCJ
March 19, 2002

mentally ill inmates languishing in administrative segregation control their behavior
so that they no longer need segregation from the general population.
Conclusions
The administrative segregation units surveyed contained large numbers of
seriously mentally ill inmates. These inmates have lengthy stays in segregation as a
product of the interaction of their mental illness and the three tiered level system
employed to deter bad behavior. Even a relatively effective program such as PAMIO
has not demonstrated the power to successfully help seriously mentally ill inmates
master the skills necessary to string together the lengthy periods of behavioral
control required to meet the criteria for reclassification. Lengthy stays in
administrative segregation in the form practiced within TDCJ are manifestly harmful
and unfair to seriously mentally ill inmates who have proven to be incapable of
demonstrating “good enough” behavior without the necessary sub-acute care for
their emotional and behavioral disorders. The absence of sub-acute care and its
substitution with low intensity outpatient care in highly restrictive and isolative
conditions has led to a deformed mental health delivery system where inmates cycle
back and forth from outpatient care to hospitalization, spanning a systems gap
where sub-acute care would normally be.
The mental health providers have proposed a system that relies heavily upon
identification and referral of inmates who have been allowed to become psychotic in
segregation. The findings of this survey support the notion that aggressive sweeps
such as those conducted prior to this site visit can cleanse most psychotically
decompensated inmates, present in the moment, from segregation areas and

36

Re: TDCJ
March 19, 2002

transfer them to a higher level of care. To their credit, the defendants propose to
conduct biennial sweeps of administrative segregation and report that they have
retained an outside consultant to supplement this process. Although this partially
addresses the needs of segregated inmates who have already decompensated, it
does not address the needs of those seriously mentally ill left behind who, because
sub-acute care is not available and the outpatient care they receive instead is so
thin, are bound to decompensate on a later date.
Outpatient care as it is currently delivered to seriously mentally ill
administrative segregation inmates is grossly deficient in its own right, but would,
even if optimal, be inadequate in type for the needs of these sub-acutely ill inmates.
The supposition that outpatient care as provided by TDCJ could substitute for subacute care for seriously mentally ill inmates in segregation is belied by the poor
quality of the outpatient services observed and the poor condition of inmates seen
there.
Both the quality and quantity of care fall well below community standards on
all measures. Mental health staffing numbers range from poor to abysmal.
Monitoring is substituted for treatment. Necessary individualized treatment plans are
either missing or so poorly constructed as to be meaningless. Prescription practices
are sometimes dangerously inadequate. Medication administration practices are
dangerous and interfere with adequate care. Important laboratory tests are
inconsistently ordered, filed, and reviewed. Individual counseling and group therapy
are widely unavailable. The majority of mental health contacts take place at cell side
with no privacy or confidentiality. Mental health rounds are often conducted in a

37

Re: TDCJ
March 19, 2002

manner that requires inmates to publicly proclaim their illness in a hostile social
environment. Statewide standards of care are not enforced. Profoundly ill inmates
can be “lost” in the system. Psychiatrists are replaced by physician’s assistants and
Telehealth. Doctoral level psychologists are replaced by master’s level providers.
Master’s level social workers are replaced by bachelor’s level providers.
Although the psychiatric inpatient programs at Jester IV, Skyview, and
Montford were not surveyed, it is assumed, based upon prior knowledge, that they
are providing adequate acute care services to inmates housed there. The PAMIO
program is well established and was found to be providing quality care to the most
aggressive segment of the seriously mentally ill population in segregation.
The Personality Disorder/Aggressive Behavior Unit formerly known as the
“Step Down” program housed in E Pod at Clements has proven to be an innovative
and successful program for extinguishing severely maladaptive behavior displayed
by inmates with serious personality disorders. This program is the only model of
true sub-acute care within the TDCJ system. It is apart from the traditional
administrative segregation structure and does not require voluntary consent. It offers
enhanced case management, behavioral programming, crisis management,
cognitive behavioral training, and aggressive psychopharmacology within a secure
but therapeutic environment. It is not meant to service the needs of acutely ill
inmates the way PAMIO sometimes does and is less staff intensive than PAMIO or
hospital services.
The new programs proposed to create sub-acute care are not likely to fill the
current service gap. Both programs have significant limitations. Although the AMPP

38

Re: TDCJ
March 19, 2002

program has been well conceived, its implementation on a scale large enough to
address the demonstrated need is in doubt.31 It has been placed in a High Security
facility where the architecture presents formidable obstacles to psychosocial
interventions. Adequate staffing is problematic as the program grows, and is
threatened altogether by the need for AMPP staff to provide routine care to a sizable
segregated population outside of the program. The AMPP program could
conceivably provide a portion of the sub-acute needs if guaranteed resources and
autonomy and freed from the constraints of a High Security administrative structure.
The In-Cell Psycho-Education Program, in its pilot phase at Smith, could be a
valuable resource, but is not likely to have much of a mitigating influence on the
damaging effects of prolonged segregation of a large population of seriously
mentally ill inmates. When and if this program is exported to other units, it will go
without staff resources. It would therefore replace existing services rather than
supplement them in a zero sum gain. Since the program relies heavily upon inmate
completion of written handouts, it could not address the needs of functionally
illiterate inmates or those so ill that they could not concentrate on or understand the
material. This program promises valuable ancillary care, but has not been proposed
in a way that suggests sufficient intensity or duration of services to raise it to the
level of sub-acute care.
Recommendations
The TDCJ and their contracted mental health care providers must develop a
model for delivering care to sub-acutely mentally ill inmates currently housed in
31

At the time of the site visit the AMPP program was operating from their program description since policies and procedures
had not yet been developed.

39

Re: TDCJ
March 19, 2002

administrative segregation. This would necessarily involve the creation of an
effective system for care to segregated inmates that goes well beyond the current
level of outpatient service. There are two logical possibilities. Either create a
system separate from, but parallel to administrative segregation for seriously
mentally ill inmates who require separation from the general population or
dramatically enhance mental health services to inmates in their current locations by
filling the service gap with true sub-acute care.
The most straight forward and least expensive way to address the need for
sub-acute care would require treatment settings that, while separate and secure,
would be outside the administrative segregation structure. For these settings to
allow the kind of biological and psychosocial interventions appropriate for treating
serious mental illness, restrictions on movement and out of cell time would have to
be altered as they are at the Clements Unit. F Pod at PAMIO and the AMPP
program are good models for this level of care for inmates suffering from severe Axis
I disorders although care would be needed in selecting appropriate housing. Unlike
PAMIO, inmates should be assigned rather than voluntarily admitted and aggression
should not be a necessary selection criteria. The Personality Disorder/Aggressive
Behavior Program at Clements is an excellent model for habilitating the most difficult
and self-defeating segregation inmates, but would have to be conducted on a larger
scale, preferably at scattered sites. Inmates meeting strict diagnostic criteria for
serious mental illness and who demonstrate their behavioral intractability by crossing
a prescribed length of stay threshold in segregation, i.e. one year, should, under this
model, be diverted to designated mental health housing that maintains safety, but

40

Re: TDCJ
March 19, 2002

allows treatment in a therapeutic setting. Seriously mentally ill inmates identified in
the sub-acute phase of their illness and likely to be harmed by administrative
segregation by exacerbation of their symptoms would be diverted directly to this
alternative setting at the time of the Pre-Segregation Mental Health Evaluation.32
This outcome could be accomplished without the creation of new facilities by sorting
inmates into more appropriate mental health housing.
If TDCJ abhors the creation of dedicated housing for the mentally ill outside of
hospitals, then it must provide enhanced treatment to those sub-acutely ill inmates
trapped in segregation by their mental illness. Using the same measure of
behavioral intractability, i.e. one year in segregation, it would be incumbent upon the
mental health providers to offer enhanced treatment with firm minimum standards.
Individualized treatment plans should by necessity include specific behavioral goals
designed to actively teach/reinforces the cognitive skills and behaviors necessary to
progress through the level system. Treatment success could be objectively defined
as meeting the behavioral criteria reclassification.
Beyond these program options, TDCJ should consider making alterations to
the administrative segregation level system. It seems prudent and parsimonious to
change the rules such that prolonged stays are not the natural and inevitable
outcome for inmates with known impairments in emotional and behavioral control.
At virtually no cost to the state, TDCJ could shorten the intervals at Level II and III
such that fewer seriously mentally ill inmates would be trapped and for shorter

32

By policy (TDCJ Mental Health Services Policy IV J, Pre-Segregation Mental Health Evaluation) inmates are
currently evaluated by a psychiatrist prior to admission to administrative segregation in order to determine if the
conditions there would be harmful given the severity of the inmates’ mental illness. Since no sub-acute alternative
is currently available, only those most acutely ill inmates meeting the criteria for hospitalization can be diverted.

41

Re: TDCJ
March 19, 2002

periods at the most isolative and constraining end of segregation. This would allow
some portion of these inmates to be reclassified to a custody level where they would
not be subjected to social and sensory deprivation of such magnitude that it
exacerbates their mental illness and where they could have improved access to an
appropriate level of mental health care. Similar results could be achieved by
eliminating one level altogether. Either way, the original disincentive remains intact
for the vast majority of inmates in segregation, since the deterrence value of 30 days
of Level III restrictions in liberty and privileges is for all intents and purposes
phenomenologically indistinguishable from 15 days. Six months at Level II in the
current system, given the harsh atmosphere and negative reciprocal expectations, is
simply insurmountable for many seriously mentally ill inmates and effectively serves
as a barrier to their reclassification. Continuing the current level system as it is
currently structured will complicate any efforts to address the needs of sub-acutely
mentally ill inmates in segregation regardless of which service model is chosen. The
intensity and duration of the social and sensory deprivation inflicted at Levels II and
III are not only exacerbating current symptoms of mental illness, but are creating
intractable conditions that will require ever more resources for their future treatment.
Please call me if you have any questions.
Sincerely,

Keith R. Curry, Ph.D.
Licensed Psychologist

42