HEALTH

MANAGEMENT

ASSOCIATES

Report on the Medical and Mental Health
Programs

of the Dallas County Jail
Confidential Draft

February, 2005

]
J

fNTRODUCTION

]
]
1

I

In December of 2004, the Dallas County Commissioners Court contracted with Health
Management Associates (HMA) to perform a comprehensive review of the medical and
mental health services at the Dallas County Jail. HMA, in order tp assure a highly
qualified and comprehensive review, subcontracted with Dr. Michael Puisis, a specialist
in correctional health with significant experience in both operating jail health services
and reviewing such programs across the country, to perform the on-site analysis.
Dr. Puisis worked in leadership positions in health services at Cook County Jail in
Chicago for more than a decade and spent his last five years there as its medical director.
From 1996-1999, Dr. Puisis was the Regional Medical Director for the State of New
Mexico for Correctional Medical Services and, since 1999, has worked as a consultant on
correctional health care for jail and prison systems across the country. He is a member of
the National Commission on Correctional Health Care Task Force for the revision of the
Standards/or Health Services in Jails, has been a consultant to the US Department of
Justice onjail and prison health conditions, is a reviewer for the Centers for Disease
Control (CDC) for the Prevention and Control o/Tuberculosis in Correctional Facilities,
and is a member of the National Commission on Correctional Health Care's Physician
Panel on Clinical Practice. Finally, Dr. Puisis has published widely on correctional health
issues. In addition to Dr. Puisis, the HMA team on this project has consisted of Pat
Terrell and Dr. Terry Conway, two Principals with significant experience in the
organization of public sector clinical services operations.
The report that follows is presented in two parts: 1) the first-person assessment of Dr.
Puisis, and 2) the recommendations that were developed from both the intensive analysis
of Dr. Puisis and interaction between Dr. Puisis and the rest of the HMA team. In the
process of the review, only the main jail complex was studied, but conclusions drawn
from the evaluation at the main complex can be generalized to the detention facilities as a
group. While this evaluation did not address the female care at the George Allen Center,
most of the recommendations apply to that facility as well.
The report is formatted to answer a series of questions posed as deliverables in HMA's
contract with the Commissioners' Court. Each question is addressed in Dr. Puisis'
review, followed by a discussion of the relevant issues. The recommendations at the end
of the report define a set of objectives that will provide direction for specific actions
needed to improve the health care services in the Dallas County Jail.

2

FINDINGS ON THE MEDICAL AND MENTAL HEALTH PROGRAMS
OF THE DALLAS COUNTY JAIL

Michael Puisis, DO
January, 2005

3

[ was asked to review medical and mental health care at the Dallas County Jail by Health
Management Associates (HMA). Health services were to be reviewed in a structured
format answering multiple questions regarding care delivery. [n order to accomplish this
task, [ toured the jail and spent eight days in the facility, reviewed documents and
medical records, interviewed staff from the jail and medical vendor, and interviewed
various officials whose agencies interact with the jail in the delivery of medical or mental
health care.
Medical and mental health care services are provided at the Dallas County Jail under a
contract with University of Texas Medical Branch (UTMB) and, thus, UTMB was a
significant focus of this analysis. Both the Sheriffs staff and UTMB were very
cooperative in my review. While UTMB would not provide financial information related
to its allocation of County funds (the budget) and [ was not able to obtain pharmacy
utilization information, there was full cooperation of the medical and administrative staff
in every other aspect of this process. I want to thank the many people who, though
understaffed, obviously work extremely hard in a very challenging environment with a
dedication that I found admirable. I hope they are able, if they review this report, to
separate the criticisms of the program from their clinical dedication to their work.

OVERVIEW

Since October of 2002, Dallas County has provided medical care through a contract with
the University of Texas Medical Branch. This contract is supported by County funds
through Parkland Hospital and is monitored by Parkland Hospital. The contract is
comprehensive and capitated on a per diem per inmate charge. Parkland Hospital, in
tum, is reimbursed by the contractor for inpatient, outpatient and emergency care at a
negotiated rate. The medical vendor is obligated for all costs of care, including pharmacy
costs, excluding some specific psychotropic medication costs.
The Dallas County detention facilities include several adult jail facilities and three small
juvenile facilities. This review only included a review of the main jail complex (Lew
Sterritt and North Tower). Excluding the three small juvenile facilities, the census and
capacity at the adult facilities during the time of my review was:

Main Complex
North Tower
Lew Sterritt
Other Jails
Suzanne Kays
George Allen
Unassigned
Totals

Census

Capacity

2954
1268

3292
1478

937
693
212
6064

1008
791
6569

4

Most medical and mental health high-acuity patients are housed in the main complex
which is the reason for focusing on those facilities. The Lew Sterritt Center includes an
intake area and a medical infirmary (which is basically protected housing) and a West
Tower that contains most of the closed and open behavioral observation cells for inmates
with mental illness. The North Tower facility contains some female housing, including
female observation cells, and a large male general population. The George Allen Center
has a female infirmary and houses both men and women of all classification levels. The
Kay Center is a minimum-security facility with a single nurse station. Data from 2002
show that in that year, 58% of inmates were discharged within 5 days of incarceration.
Approximately 11% of inmates stay longer than 60 days. From current census data, 14%
of inmates are females.

DELIVERABLE #1: Closely observe the intake processes at Dallas County Jail and
lor other Dallas County detention centers (i.e. how quickly and accurately is mental
illness or potential suicide detected, how detainees are directed into care after such
detention, etc.).
Typical Expectations for Intake Screening
Conducting intake screening evaluations is an important activity that establishes the
clinical requirements necessary to safely house inmates in a jail. This process addresses
immediate health care needs of new inmates and assists in assigning special housing via
classification.
Intake evaluations should include both arrival screening and intake physical
examinations. The purpose of arrival screening is to assess the general well being of an
inmate immediately upon his/her arrival in the institution in order to identify emergency
conditions that must be addressed upon entry to the facility, facilitate the immediate
transfer to a hospital or to identify chronic conditions, medications and any current
problems that should be identified in this population. This screening is typically
performed by nurses or other health care personnel and should both identify conditions
that require immediate attention prior to completion of a physical examination and
ensure continuation of prescribed medications.
Arrival screening typically consists of a structured history of the inmate's mental health
and medical problems and previous therapy, visual inspection, vital signs, and entry
testing for tuberculosis (Mantoux skin test) with follow-up. Patients identified by this
screening as requiring immediate attention are referred for mental health or medical
physician review or examination as indicated. Inmates should also, at this time, be
instructed on how to access health care and mental health care once they are housed in
the facility.
Intake assessments or examinations are evaluations that should take place anywhere from
hours to two weeks after an inmate arrives in the facility. The time period between
arrival and the examination is determined by whether the detainee has a medical or

5

mental health condition that requires evaluation. Inmates are prioritized based on safety
concerns and evaluated in a time-frame appropriate for their clinical condition(s).
Intake physical examinations have two purposes:
1. follow-up of recognized medical or mental health problems identified on
arrival screening, and
2. performance of age, population, and gender specific interventions and
examinations that update appropriate screening, counseling and other
interventions (vaccinations) that are appropriate for the inmate.
Follow-up of known or new medical or mental health problems are one major focus of
intake physical examinations. This activity ensures the prompt continuation of necessary
medication for all inmates with chronic medical and mental health conditions.
Physicians, psychiatrists or licensed prescribers should be available on-site or on-call to
initiate necessary medication whenever required for a newly arrived inmate. The arrival
screening process should identify the priority of examination for newly screened inmates.
Persons with more potentially serious chronic illness (i.e., active psychosis, suicidal, type
I diabetes mellitus, coronary artery disease, cancers under treatment, etc.) should be
referred from arrival screening for early physician appointments to establish a baseline
examination for their chronic mental or medical illness, as well as completing their age
and gender specific health appraisal interventions. Typically, a physician or psychiatrist
should examine persons with serious medical or mental health problems. This should
occur early (e.g. within 24-72 hours of arrival).

The Dallas County Jail Reception Screening Process
The Dallas County reception screening process is ineffective. Officers perform medical
reception screening in the Dallas County Jail. There are no medical policies governing
how officers screen and officers receive no training to screen. Officer screening is not
accomplished under medical supervision. The scant medical policy on reception
screening that does exist merely indicates that once an officer identifies a problem, he/she
refers it to the intake nurse.
Detainees are interviewed by screening officers prior to obtaining a booking number.
Ostensibly this is done so that if an inmate is identified with a serious medical problem,
they can be referred to Parkland Hospital before they are incarcerated or can be referred
for mental health diversion. There are financial considerations in this practice. If an
inmate is sent to Parkland prior to booking, then UTMB is not liable for the cost of
hospitalization.
Two separate screening questionnaires are used by officers as they screen inmates. Both
are self-report forms. One is the Mental Disability/Suicide Intake Screening Form and
the other is the Medical Screening Form. These forms are not incorporated into the
medical record.

6

]

]
]

]

]
]
]

]
]

]
]

Detainees line up for reception screening by officers in multiple lines perpendicular to a
long counter behind which screening officers stand. At anyone time, as many as four
lines of inmates present at this counter. Detainees stand in one of these lines
approximately one to two feet apart. Officers standing behind a counter perform the
Medical Screening Form as the inmates approach. The arresting officer may be standing
immediately behind or next to the detainee. The detainees that I observed were in the
process of taking clothing off, giving articles to arresting officers, and attempting to pay
attention to the transfer of their property to officers. The room where officer reception
medical and mental health screening occurs is a large room that is both the screening area
and throughway for the entire intake process. Dozens of staff are walking around,
inmates are sitting on benches waiting in line, people are loudly talking, there are
multiple conversations creating a chaotic situation. [n this context, officers are publicly
asking questions about health concerns, whether detainees are suicidal, past psychiatric
history, etc. Neither the officers nor the detainees appeared to be paying much attention
to the task. This screening process is not a specialized officer assignment and any officer
may fill this role. There is no training for any of these staff. It is highly likely that
officers would not understand the reasons for asking the screening questions that they are
to ask inmates. One officer in the intake area admitted to me that is was likely that
inmates with certain medical or mental health conditions might not want to reveal it
under these circumstances.
Public screening for medical or mental health purposes is not appropriate and for this
reason, standard-settingregulatory agencies (e.g. National Commission on Correctional
Health Care) require that clinical encounters occur in a setting of privacy so that accurate
information is obtained.
In the Dallas County system, officers currently identify approximately 30% of detainees
as having a medical or mental health problem and refer them to the Registered Nurses in
intake for further evaluation. However, several independent staff members in mental
health as well as in medical services cited that, in their opinion, officers miss
approximately 25-35% of detainees with medical problems. These detainees may
subsequently be identified.ifthe detainee requests health services once incarcerated. The
25-35% of detainees with health problems missed by screening officers is approximated
based on the number of health requests that subsequently reveal medical conditions.
However, substantial numbers of detainees are discharged within the first several days of
incarceration. Only approximately 42% of inmates remain in the jail longer than 5 days.
Therefore, it can be estimated that a considerable number of persons missed at intake do
not even place a request because they are discharged prior to being able to place a
request. This would mean that the number of detainees with health conditions missed by
screening officers exceeds 35%, conservatively. Because it can not be predicted with any
certainty that those missed will be released from custody, it is very likely that persons
with serious illness are missed who will remain in the facility. This assumption was
ultimately validated upon my subsequent review of medical records.
When an officer identifies a mental health or medical problem at this reception screening,
the detainee is referred to one of two Registered Nurses in the intake area. Nurses review

7

]
J
J
J

I
I
I

the officer screening fonns and then perfonn a Central Intake Evaluation Fonn that
includes vital signs, a pulse oximeter reading and a number of questions regarding history
of medical or mental health illnesses as well as medication history. Nurses will decide
which of these patients will need subsequent evaluation or special housing. The paper
fonns of persons who are referred by nurses in intake for follow up evaluation are then
scanned at a later time into the electronic medical record. This group represents only a
small percentage of persons entering the jail (approximately 30%). The electronic
medical record is not utilized in the intake area.
The self-report questionnaire completed by a nurse is perfonned in a clinic room in which
both intake nurses are simultaneously interviewing inmates. Officers enter the room with
the inmate and can listen to the interview or even stand next to the inmate. On the day I
observed intake, officers were standing nearby, sometimes listening and sometimes
talking to other staff or simply waiting for the nurse to complete the interview. This
situation does not ensure confidentiality and may result in less than accurate history.
After this evaluation, the nurse makes a decision about whether the detainee is referred
for specialized medical or mental health housing. Only persons referred by a nurse at
intake are followed up by mental health or medical staff on subsequent days. This is why
the intake process is so important.
As of December 8th, I was told that there were 89,000 persons incarcerated, for an
approximate average of260 persons a day, during 2004. In November, approximately 81
persons a day (30% of those daily incarcerated) were referred to a nurse for medical or
mental health problems. Of the 81 persons evaluated by nurses daily, approximately half
(41 per day) were referred for a secondary evaluation or placed in special housing. The
breakdown was:
•
•
•
•

26 referred daily for mental health evaluation;
13 referred to the infinnary;
1.5 referred for follow up medical evaluation;
0.5 per day referred emergentiy to Parkland Hospital.

The decision regarding whether a medical follow up will occur is made at a later time by
nursing staff in the housing units. Mental health referrals are somewhat different. All
those deemed by intake nurses as having mental health problems will be referred to
mental health liaisons for a follow up evaluation. However, intake nurses receive no
training in mental health screening and perfonn their evaluation based on legacy
expenence.
Using the November statistics for Dallas County, approximately 30 people a day (10% of
persons incarcerated) are referred for mental health screening. The American Psychiatric
Association (APA) estimates that approximately 20% of persons in correctional facilities
have serious mental illness. Given that clinical staff at the Dallas County Jail estimate
that officer screening misses about 15-30% of those with mental illness and that many
people are discharged before they are discovered, it is highly likely that between 10 to 25
people a day go through the intake process without having their mental illness identified.

8

I
J
J
)

I
I
I
I
I
I
)

I
I
J

I
]

These missed opportunities to pick up illness result in detainees having to place sick call
requests and increase the evaluation burden later in the clinical process and can result in
detainees being unsafely housed.
The paperwork related to detainees who will be referred for evaluation are placed in a
central location and subsequently delivered to either medical units or to the mental health
unit. All mental health referrals are delivered to the 3rd floor mental health offices on the
3 rd floor of the Sterritt building.
Mental health liaisons are the primary mental health evaluators. Typically, they have
Bachelor level training. Referrals from intake nurses are delivered to a clerk in the
mental health offices who distributes them to mental health liaisons. Mental health
liaisons do their routine work on weekdays. On weekends, limited staff is available to
perform screenings (only serious problems -e.g. suicidal inmates - are examined) and,
therefore, on Mondays most of the weekend referrals have accumulated and result in a
higher number to be seen. This results in a catch-up process in which liaisons attempt to
complete evaluations of referrals.
Before the liaisons can evaluate the patients, they must locate them. Accommodations
for person with mental and medical illness will be described later in this report.
behavior
However, serious mentaJly ill can be housed in anyone of a number of closed
t
observation or open behavior observation cells. Liaisons look up the location of each of
their patients individually on the custody inmate tracking system. On a daily basis, their
entire caseload of patients may have changed locations. This occurs unpredictably and
without notice, as will be described later. Once they have located inmates on their
caseload, they must then locate an officer to escort them to the housing unit so that they
can interview the inmate. No one is permitted to see inmates who are housed in closed
observation cells for clinical evaluations without an escort. There are about four times as
many clinical staff as there are officers. The result of this disparity is that there is
competition for officer time, creating a drag on the efficiency of the mental health
workers and reducing the number of individuals that can be evaluated in a day.
In addition, those inmates missed at intake screening, those with newly diagnosed
conditions, those who have problems with their medications or with appointments all
write requests for care that the liaisons have to evaluate. For the liaison for the 3rd and 4th
mental health floors with the majority of mental health patients, there are approximately
25-30 inmate health requests on a daily basis. About half of these inmates already have
appointments. Those who have appointments are not seen even though they may have a
need to talk to someone about some aspect of their care before their appointment. Those
without appointments must be located and are then placed in the queue to be seen. They
are seen after the intake referrals are seen. However, because the liaison is engaged
seeing intake referrals and has difficulty in getting an officer for escort, many of these
individuals who place requests to be seen are discharged from the jail before the liaison
can see them. Some of these may have had mental illness that was missed at intake
screenmg.

)

I
J

9

]

]
]

]
]

]
]
]
]
]

1

1
]

After the liaison sees an inmate, the inmate may be moved. Additionally, a nurse or
custody staff may have assigned an inmate to one of the closed observation cells and no
one may know they are there. One liaison I talked to prints a list once a week of all
inmates housed on units that he is assigned to. He then tours the units and individually
questions the patient regarding whether the patient has a problem and whether he has
been seen. There is no method to determine an aggregate caseload for any of the
practitioners. Many practitioners have developed their own unique methods of
attempting to track patients that are housed on units to which they are assigned. Even
though these checks are performed, there is no guarantee that someone who was missed
will be seen because of the aforementioned backlogs and custody restrictions. Further,
the list printed up from the custody computer contains no clinical information; it only has
housing information. So the weekly rounds are blind clinically. The purpose is only to
identify a crisis patient by interviewing patients, in which case the liaison will look them
up.
Thus, the steady state of this system is one in which there is incomplete intake screening,
delayed subsequent evaluations and a steady attrition of unscreened inmates by virtue of
being discharged from the jail before they can be evaluated.
Additionally, the electronic record is useful for looking up clinical records in one of the
offices, but the system adds additional work for mental health staff. All mental health
patients must be seen cell side. Therefore, all notes are written on single progress notes
or pre-printed forms and are later scanned into the electronic record. However, as
clinicians are seeing patients, their evaluations never occur with the benefit of a medical
record. While the electronic record is a useful way to retain records, the current
arrangement defeats the main purpose of having a medical record--that clinicians can
review past evaluations as they are seeing patients. Looking up patient locations,
scanning paper copies of progress notes into the record, and maintaining personal lists of
patient caseloads consumes up to 30% of liaison time.

DELIVERABLE #2: Evaluate the capacity ofthe intake and other jail staffto diagnose
detainees.
Intake

]

]
]
]
]
]

There is no one assigned to the intake screening process licensed to diagnose either
medical or mental health conditions. In this respect, there is UUno capacity to diagnose
illness at intake. The screening process exists not to identify disease but tp assign
housing to inmates. To a certain extent, nurses make referrals and some patients are
referred to Parkland Hospital, to mental health and to the infirmary. Many inmates do
not have their chronic illness identified through the intake process. For those whose
condition is identified, the subsequent evaluation of the status of their disease is
dependent on nurses in the housing units. Follow up evaluation seldom occurs by a
physician; this only occurs if a second nurse on the housing unit refers the patient to a

10

]
]
]

]
]
]
]

]
]

1
]

]
]
]

]
]

physician. Significant numbers of persons with clinical conditions never receive an
evaluation by a clinician other than a nurse.
The intake nurses are Registered Nurses capable of performing assessments but are not
trained or licensed to diagnose. Occasionally they perform minor treatments, but mostly
they make housing assignments (infirmary, closed or open observation, general
population, etc.). Practically, nurses manage intake without direction from any advanced
level practitioners (nurse practitioner, physician assistant, psychiatrist or physician). For
inmates who have new problems, or a condition that appears complicated, the nurses in
intake can contact the on-call physician to discuss the problem over the phone. However,
in reality, consultation with a physician occurs for only a very small number of inmates.
For the month of November as an example, the nurses made 81 calls to the physician on
call. This volume amounts to 2.7 calls a day out of 260 admissions to the jail. As will
be discussed in a later section, significant numbers of persons in a jail situation can be
expected to have a chronic disease. If looked at from the perspective of persons referred
to the nurse from custody officers, nurses call a physician for 3% of inmates referred to
them by custody officers.
The National Commission on Correctional Health Care estimated expected rates of
selected chronic illnesses. For the following diseases the average rates nationally were:
asthma--8.5%, diabetes--4.8%, and hypertension--18.3%. Other common conditions
include epilepsy, alcoholism (and alcohol withdrawal syndrome), other substance abuse
withdrawals, sequelae of traumatic injuries and infectious diseases (HIV, tuberculosis,
sexually transmitted disease, etc.). The cumulative rate of persons with a chronic illness
or other medical condition is unknown but can be expected to be somewhere around
30%. In Dallas County, intake nurses screen on average 81 patients a day. Twenty-six
are referred to mental health. If all of the others are assumed to be medical patients, then
this would equate to approximately 21 % of incoming inmates being identified as medical
patients, below the expected number anticipated. As with mental health conditions, it
appears that many detainees with medical problems are missed by officer screening.
The nursing review of inmates sent to them consists of a review of the officer screening
questionnaires and completion of another medical questionnaire with vital signs. The
medical questionnaire includes a listing of medications. The physical examination only
includes vitals signs. There is no formal required visual inspection of the detainee. The
examination is not conducted in private; it is in a room shared by two nurses. Two
simultaneous examinations are being conducted and there are officers present as well as
visitors to the officers. Thus, inspection is only incidental to the interview. For asthma
patients, a pulse oximeter reading is evaluated. This is not a useful screening evaluation
of asthma; expiratory peak flow testing of persons with asthma is not done. Pulse
oximeter testing should be ordered by a physician because nurses in this system do not
understand the purpose of this test and using it for asthma can result in dangerous
mistakes being made.

]
]
J

11

]
]
]
]

]
]
]

]

l
l
I
]

1
J

I
]

I
I
)

In the intake area, an examination of an inmate behind a curtain can be performed, but it
appears to be the exception rather than the rule based on my observation of the process.
In this arrangement, the effectiveness of screening for serious medical conditions is
significantly reduced.
After intake nurses complete paperwork for inmates referred to them, they send that
documentation for persons with medical conditions to nurses in the housing units to
which the inmate is being sent. The following day, every day, nurses on those units
receive a pile of referrals from intake. On weekends, these back up somewhat because a
physician is unavailable to review referrals. Nevertheless, an assigned nurse evaluates
every one of these referrals. This takes a considerable amount of time. The intake sheet
is scanned into the medical record and the nurse may write a brief note for selected
patients. Some patients are seen by the nurse.
There is no policy or procedure for how inmates who come into the jail at intake are to
receive their medication or be referred to a physician. Only certain persons will be
started on medication from intake. The operating practice appears to be that if a patient
brings medication in with them in a labeled container, intake nurses will allow them to
have several days worth of medication. If they do not have medication with them, intake
nurses will either call a physician or note the medication on the intake sheet for follow up
in the housing unit. Given that only approximately 2.5 calls per night are made to a
physician, not very many patients on medication have prescriptions written from intake.
For the majority of patients on medication, the nurse on the housing unit will send an
electronic reminder to a doctor indicating that certain patients need medication
prescribed.
The nurse is the person who also evaluates intake referrals, makes any necessary
notations, may see some patients, and may refer them for physician review. Physicians
will thus see only a fraction of incoming persons with chronic illness. Based on
discussions with staff and further validated by medical chart review, it appears that only
about 25% of incoming inmates with a chronic illness are actually physically seen by a
physician following intake. For most patients, their care involves physicians or nurses
reviewing an electronic version of a scanned intake form written by a nurse who is seeing
patients in a setting without privacy and not having performed a physical assessment.
This practice of review of electronic record is the dominant form of clinical encounter at
the facility and accounts, I believe, for the fact that virtually none of the medical chart,S I
reviewed reported physical examinations of the patients. This practice may be a result of
lack of clinical staff, a problem which I do believe exists. However, an examination of
the effect of the electronic record on reducing clinical examinations should be performed.
Even in the most efficient version of this process, a physician will prescribe medication
for a patient the day following intake without having evaluated them.
Patients with chronic illness also may never be adequately evaluated in this system.
Physicians must not only review intake reminders, charts, and prescribe medication for
intake patients but are also responsible for seeing patients who request to be seen in sick

12

]
]
]
]

I
)

I

1
J

I

1
1
]

l
1
]
]

]
]

call via the kite system that will be described later. In one physician's case, he is also
responsible for managing all of the tuberculosis patients. This amount of work results in
physicians managing all illness by reviewing electronic reminders that nurses send them.
The net result of this practice is that only nurses manage face-to-face care of patients and
virtually no one obtains a clinical examination appropriate for their condition. [n
addition, the nurses' description or documentation of the patient's illness may not be
accurate. This may result in patients with serious illness being described less seriously
by nurses and therefore ignored by busy physicians. It appears from chart reviews that
this is occurring as a usual pattern.

Sick Call and Evaluation of Inmate Requests for Care
Sick call is a process in which inmates in the jail population are evaluated based upon
requests for care or in follow up of already established care. Access of this type is
fundamental to the 8th Amendment Constitutional rights of inmates. The sick call process
in Dallas County Jail is not adequate and its steady state is that inmates, by attrition, leave
the jail before being seen more frequently than being appropriately evaluated. The
system suppresses utilization by virtue of the multiple barriers to access.
Inmate health requests are called kites in this system. Inmates request care by filling out
a form stating their reason for requesting to receive care. The distribution of these kite
forms to inmates and the delivery of completed kites are mediated by officers. In most
correctional systems, both for privacy and to ensure that officers do not destroy or mislay
an inmate's request, inmates are permitted to have access to paper kite forms and to a
locked container in which to place their kite. The reasoning is that if only health care
staff have access to the locked containers, delivery directly to a health care person would
be ensured. However, because officers handle kites in the Dallas County Jail,
confidentiality is not ensured and transfer of requests to health care is dependent upon the
trust of officers to transmit the information to medical staff.
In most correctional systems, the number of inmate health requests is typically around
10% of the incarcerated population on any given day. It is important to put into
perspective the fact that inmates have no access to any over-the-counter medication or
freedom to seek any type of care or assistance without first going through medical staff.
For that reason, they seek help through the medical services. These requests therefore
include assistance for minor complaints mixed in with complaints for care for very
serious illness. The manner in which a request is written does not often portray the
seriousness of the actual problem. Therefore, it is imperative that each request be
evaluated in order to determine the needs of the inmates.
The exact number of kites inmates fill out is not an officially recorded statistic, but some
nurses maintain the number of kites received. Because of the lack of secure transmission
of inmate requests from inmates to health care staff, inmates may be requesting care via
kites that health care staff do not always receive. Official statistics are only kept for the

13

]
]

]
]

]
]
]

]

]
]
]
]

]
]

number of kites scheduled for appointments and responded to in writing which may not
necessarily include all of the kites inmates fill out. [reached this conclusion because of
the low percentage of kites (4%) received by the health care staff compared to an
expected number (lO%) that is seen in a typical correctional facility. [n this system, with
imperfect intake screening, lack of physician examinations for those with chronic disease,
and with delays in getting medication, it should be expected that the numbers of kites
should be much higher than the 10% typically seen in other correctional systems because
more people will attempt to obtain necessary services.
As an example of the Dallas County system, [ looked at the North Tower sick call and
kite process. Nurses in the North Tower perform sick call, evaluate kites, and respond to
emergencies. They are so overwhelmed by the number of requests for care relative to
their staffing level that they have accepted seeing only a percentage of inmates who need
care. Nurses evaluate kites by going to the housing units where inmates reside and doing
a brief evaluation at a desk near the control tower. For clinical conditions, this is not
acceptable practice, and is only acceptable as a crude form of triage. There is no privacy
and no acceptable place to examine patients. For the North Tower, nurses receive
approximately 120 health care requests a day. This represents only approximately 4% of
inmates, which is lower than typically seen in correctional settings (making me believe
that the process of officers handling sick call requests is not optimal and may be a barrier
to inmates getting their request to health care staft).
In any case, nurses have staffing to evaluate these health care requests only two days a
week. Correctional health standards require that these requests be evaluated within 24
hours. Requests that are not evaluated are not entered into the electronic record; only
those seen are entered. On the two days the nurse sees patients she will see
approximately 70 persons in about 4 hours. This is approximately three and a half
minutes per patient, a very brief evaluation given that this includes the time it takes to
travel between housing units and to move patients through a line. The quality of clinical
evaluation can only be crude triage sufficient to determine whether a patient should be
seen in sick call in the clinic. The remainder of the shift will be spent in entering data
about these visits into the computer. However, since nurses receive approximately 120
requests a day on a 5-day-a-week basis, nurses are only evaluating about 25% of requests
that they receive. About 40% of requests are responded to in writing; the remaining 60%
are triaged for scheduled appointments in sick call.

]

Sick call is a clinical evaluation that the nurse performs in the clinic so that an
appropriate evaluation can occur with privacy and with appropriate equipment.
There are multiple barriers to inmate access in this process. The first barrier is simply the
number of scheduled visits relative to the available staff to see patients. All requests for
nurse sick call are entered into a computer and then enter a queue. Each day a clerk
prints out the queue of patients who are still incarcerated but have not yet been seen.
Many more people are scheduled than can be seen. Nurses work two shifts, seven days a
week and see about 30 people per shift. Statistics on the actual number persons seen was
not available; only the number of persons scheduled. Nurses indicated to me that patients
continually leave before they are seen.

]

14

]
]

]

]

]
]
]
]

]
]

]
]
]
]
]
]
]

]
]

The availability of officers is also a problem in ensuring access. During the daytime
hours, officers will bring only one inmate to the clinic at a time. This is a tremendous
drag on the efficiency of staff. If both the doctor and the nurse are available for
evaluating patients, the process is even slower. For reasons that are not clear, officers in
the evening will bring inmates in groups that allow for greater efficiency and permits
nurses to see more patients on the evening shift. Nurses indicated that on evening shifts
they can see up to 70 patients a shift. The quality of these evaluations is not monitored.
Nevertheless, not all patients are seen who need to be seen and are reentered into the
queue, which is kept at a steady state only by attrition. On the day of my visit, nurses
were seeing patients who had initially placed requests in late November, or 6 weeks
previous.
Emergency responses appear to be of two types. When officers call nurses for an
emergency, all clinical activity in the clinic stops and the nurse responds to the
emergency. In addition to true emergencies, officers may call a nurse about a problem
with a particular inmate that is seen as important but not as important as a crisis. These
are requests from officers to see a patient because of the officer's sense that something
more serious is wrong. These might not be true emergencies but are a reflection of the
officer's attempt to bring a more urgent problem to the nurse's attention. Obviously, in a
system that cannot address routine problems in a satisfactory time frame, many problems
raise to the level of emergency. Yet, because officers are not expertly trained in medical
evaluation, the exact nature of the problem is never entirely clear. These evaluations
necessitate insertion of evaluations into a system which is already backlogged and nurses
attempt, with the time available, to address these "emergency" requests as soon as they
are able. I was told that nurses cannot see these "emergency" patients every day but try
to catch up with them by the end of the week.
The physician's role in this process is equally ineffective. The multitude of intake forms
that arrive each day to the nursing station are entered into the electronic record. The
nurse refers to the physician, by way of electronic reminder, if the patient needs
medication. The physician reviews these reminders as well as reviewing electronic
referrals from nurses for patient care issues. Because there are so many requests, the
physician reviews all of these requests electronically and makes decisions about whether
to see patients based on the perceived needs of the patient. Between referrals from
. intake, nurse referrals for care and kite requests for services, the physician has to review
approximately 50 charts a day, and the nurse indicated that this could be as high as 100.
Approximately half may have chronic illness, according to staff. However the physician
can only physically see about 10-15 people a day, meaning that almost 70% of the people
who should be seen actually are seen. These types of defective systems result in negative
feedback loops in which inmates who have problems are not seen and therefore generate
more requests which are not evaluated and so on and so on.

]
]

J

15

]
]
]

]
]

]

]
]
]

]
]
]

]
]
)
J

1
]
J

DELIVERABLE #3: Review and analyze deaths and suicide rates and attempted
suicides at Dallas County Jails and other Dallas County detention centers. Evaluate any
and all suicide prevention plans ofDallas County detention centers.
Nationally, deaths in custody are not well studied. The United States Department of
Justice began collecting crude data on deaths in local jails beginning in the year 2000, but
these data have not yet been published. Death rates in correctional facilities, but
particularly in jails, are difficult to determine because inmates are incarcerated and
discharged so unpredictably that an accurate denominator for the rate is difficult to
determine. Customarily, the average daily census is used as the denominator, but in the
Dallas County Jail, where over 50% of those incarcerated leave before 5 days, the
denominator of average daily census may bias the death rate even further. In addition,
the exact age, sex and race of all inmates needs to be accounted for if the death rate is to
be compared to any other death rate to avoid misinterpretation. Death rates for young
persons (the typical incarcerated person is young) are much lower than the average death
rate for Americans in general. Also death in the young is typically by accident; motor
vehicle accidents and homicide are the major causes of accidental death in the young.
These types of accidental deaths do not generally occur in jails. Therefore, the death rate
in jails should be expected to be lower than an aged matched civilian population and
lower than the overall civilian death rate and it will be difficult to calculate a rate at the
Dallas County Jail except for crude estimates.

In Dallas County Jail, deaths are recorded by cause of death, although there is no formal
mortality review so it is difficult to determine the accuracy of these designations and
therefore to what degree deaths are occurring that are preventable. As an example, for
the year 2004, three of the eleven deaths were listed as myocardial infarction and another
three of the eleven were listed as cardiac arrest. For this age group, six of eleven deaths
resulting from cardiac events would not be expected. Therefore the actual cause of death
has probably been reported inaccurately and the actual causes of death may have been the
result of a preventable death that should have been thoroughly reviewed. In fact, upon
reviewing one of these deaths, the patient probably died of electrolyte abnormalities
resulting from complications of end-staged liver disease that were known to staff but not
treated in a timely manner. The reported cause of death did not reveal this fact.
Review of deaths, including suicides, should be performed for every death to include
calise and contributing causes of death, a coroner's report, identification of any problems
surrounding the care of the patient, determination if the death was preventable, and
corrective action steps to correct problems identified. All deaths should result in an
autopsy. Also, mortality review should be a first step of peer review of clinical staff
when indicated. This information should be protected so that publicity and
sensationalism does not interfere with improvement of clinical care. Clinical and
administrative staff may not honestly review their practice if they know that their work
will be subject to public scrutiny. These reviews do not occur for deaths in the Dallas
County Jail.

16

]
]

]
]
]
]

]
]
]

]
]
]

]
]
]
]

In the year 2003, there were 13 deaths and in 2004 there were II deaths of inmates of the
Dallas County Jail. If the average daily census is used as the denominator, the resulting
death rate would be lower than that seen in the Texas Department of Corrections (216 per
100,000 vs. 294 per 100,000). However, the fact that 58% of inmates are released within
5 days may artificially lower the Dallas County number. For the year 2003, there were
three reported suicides yielding a rate of approximately 50 per 100,000. Department of
Justice statistics for jails in 1999 show a suicide rate on average for all jails of 54 per
LOO,OOO. This rate had been dropping steadily and no comparable data is available for
the year 2003. For the year 2004, there was one suicide in the Dallas County Jail. There
are no Department of Justice statistics available for this year but is should be expected
that the one suicide was below the national average.
Texas is the only state in the United States where jail standards require that county jails
maintain procedures for six critical suicide prevention components, including staff
training, intake screening, communication, housing, supervision and intervention. While
I did not review every one of these elements, intake officer staff received no training in
suicide screening and the communication between mental health staff and correctional
staff is informal at best. The main problem with this program lies with the interactions
and collaboration between custody and mental health staff.
UTMB mental health policy on suicide prevention was written in February of2004 but
not one of the multiple staff! interviewed was aware of any UTMB policy in any area
governing mental health care, including suicide prevention. In addition, that written
policy does not address major components of the suicide prevention practices. Suicide
prevention should be under the direction of mental health staff, yet, uniformly, the mental
health staff I interviewed did not agree with existing suicide practices, especially the
stripping inmates of all clothing.
Mental health staff do not formally review suicides. Staff I interviewed did not even
know how many inmates had committed suicide. The numbers of attempted suicides are
not tracked either. So there is no formal mechanism to meaningfully assess or review the
suicide prevention program in place.
The current practice of suicide prevention consists of placement of anyone who suggests,
or appears to be contemplating, suicide into a "suicide cell," which is a closed
observation cell in which the inmate is under continuous lock down. There are two tiers
designated as suicide tiers and there are 8 cells in each tier that are classified as suicide
cells. These cells are visually checked by an officer, ostensibly at fifteen minute
intervals, although this check is not formally logged. Of those individuals interviewed,
all admitted that there is a very low threshold to placement in a suicide cell. This is both
positive and negative. On the positive side, staff are alerted to the potential for suicide
and refer for specialized housing for those individuals so suspected as suicidal. On the
negative side, this assessment is seldom performed by a mental health staff member so
the sensitivity of the classification is probably poor.

J
j
j

17

]
]
]
]
]

]
]
]

]
]
]

]
]
]

l
)
)
J

]

There are no statistics on the numbers of persons placed in suicide cells who are actually
suicidal, but mental health staff told me that it is their perception that few of the persons
placed in these cells are actually suicidal. Because inmates who are placed in these cells
have all their clothes taken and are given a paper gown, the treatment is less than optimal.
They must remain in the cell until a mental health evaluation occurs and they can be
removed through classification. This is a crude process that has positive aspects but
could be improved by increased supervision of mental health staff in the placement of
persons into these cells and in the supervision and management of these individuals while
in suicide cells.
The actual management of suicidal inmates includes 15-minute checks on the inmate by
officers and daily (Monday through Friday) psychiatrist visits. On weekends, suicidal
patients are not evaluated by a psychiatrist and therefore must wait until the first working
day to have an evaluation. If the patient in a suicide cell is psychotic, the liaison will see
the patients daily if necessary but no interventions are done except to make a psychiatrist
appointment. All interviews are in public between the bars of the celL Suicide patients
are naked with a small piece of paper to cover themselves up. Anyone can place a patient
into suicide watch, including custody. These cells, similar to any closed behavior
observation cell, can have their classification changed by the classification officer so that
not all persons in the suicide cells are suicidaL Intoxicated people, for example, are put
in suicide cells as well, confusing the mental health staff regarding the reason for
placement in a cell typically construed as suicidaL One of the mental health staff
conjectured that 75% of people on suicide watch aren't suicidal (it was her belief that
placement in a suicide cell is usually for punishment or because of intoxication).
3PI is a suicide tank. The order to place someone in the suicide tank, a form completed
by the mental health worker (PA or MD) on call. However, the clinician doesn't even
know that their name has been used to make this placement. In effect, therefore, suicide
watch doesn't require a physician order.
One problem that staff encounter is that they tour the unit without benefit of any record as
to why the patient has been placed in a suicide celL They must depend entirely on the
interview with the inmate to try to uncover why the inmate is being housed there. If the
inmate was placed in the cell from intake, the intake note will not yet be scanned into the
record and is unavailable to the psychiatrist seeing the patient. Thus, clinical decisions
are made without benefit of knowing what has previously occurred to the patient. For
example, on the day of my visit, there were 5 patients in 3PI on the suicide unit. It was
not entirely clear why the patients had been placed in the cells based on interviews. If a
patient has disorganized thoughts, there will be little to no reliable communication as to
the patient's reason for being in the cell. Mistakes will result.
I checked the medical records for the five individuals in 3PI:
I. One patient had no information in the EMR related to why he was in a suicide
cell.
2. The second patient had no documents in the EMR.
3. The third patient had no documents in the EMR

18

I
)

]
]

I
I
]
]

]
]

]
]
]
]

l
]

]
]

I

4. The fourth patient had no mental health problems in October when he came in.
He got in a fight and said he might kill himself and the psychiatric nurse put him
on the unit. The physician assistant had given a phone order for an antipsychotic
medication without an evaluation.
5. For the fifth patient, there were no notes in the EMR relevant to this suicide
watch. He was incarcerated March 3rd , 2004.
Four point restraints are not used in the Dallas County Jail and there is no place in which
to apply four point restraints. There is a restraint chair in intake. Officers use it without a
direct MD order in order to physically restrain combative inmates. I was told it-is
infrequently used.

DELIVERABLE #4: Evaluate similarities and/or other connections between mental
health care services and medical health care services for their effectiveness,
appropriateness, potential for duplication of services, etc. & determine the appropriate
health care provider to use for mental health care and medical care at the Dallas County
detention centers.
Staffing plans should evolve out of the requirements of policy and procedure and the
mission of the health care program. Every correctional facility will have varying barriers
to care, different needs, and different missions to provide care to inmates. Therefore, the
mission of the health care services should be established. Policy should be developed to
carry through on the stated mission. From this, appropriate staffing levels can be
determined. The mission of the medical program at the Dallas County detention facilities
is unclear. Most staff I spoke with have a survival mentality. They see their main
purpose as trying to address the emergency of the day rather than being engaged in a
health care program.
The Dallas County Jail medical program does not have existing policy that is in current
use. UTMB developed a policy and procedure manual in February of 2004, but no staff I
talked to acknowledged that there was a policy and procedure manual that they use.
Some senior staff did not know one existed. Policy in that manual is not adhered to and
there are some areas for which policy should be developed or improved. I also did not
get a sense of the level of services desired by the County in its jail, including the public
health mission or dental care. The development of a mission by the County followed by
development of policies should be a high priority for medical and mental health services.
Effective policies require that staff receive training related to the implementation of those
policies and procedures on an ongoing basis so that expectations are clear. A system of
auditing against policies should be performed on an ongoing basis to continually improve
services as well as a means to monitor deliverables of the contract. None of these
processes are currently performed in an effective manner.
Also, medical and mental health staff do not always coordinate their work together.
There are no joint policies on sick call that are shared between mental health and medical
staff. When a person on the mental health unit has a medical problem, it appears that

19

)

1
1
1
1
]

1
]

]
]
]
]

1
]

]
]

]
]

I

there are no arrangements for the patient to be seen for the medical problem. It also
appeared difficult for patients on medical units to be evaluated for mental health
problems. The psychiatrists used nursing staff to assist them in transcribing their
prescriptions into the electronic record because of the time it otherwise took them to do
that. After these nurses were pulled to perform sick call on the mental health unit it
appeared that not all psychiatrists' prescriptions had signatures on them in the electronic
record. Mental health and medical do not appear to be working as a team.
If the main priority of an organization is simply quelling emergencies, development of
staffing plans will be ineffective. There is no question that there is a shortage of health
care staff at the Dallas County Jail. However, UTMB has developed a staffing proposal
in response to that perceived lack of staffing that is based on the existing practices at the
jail that were mainly to quell emergencies. In that regard, staffing proposals are not as
efficient as they could be and should have a different emphasis.
In addition, no financial information regarding budget lines is provided by UTMB to the
County so that an adequate determination can be made of the use of County funds in
providing medical services. This lack of transparency will make it difficult for the
County to adequately assess the reasonableness of any additional funding for staffing.
The ostensible reason for this lack of transparency is the UTMB considers its budget
proprietary. However, in my opinion, budget lines in correctional programs are not so
unique as to constitute anything proprietary or special. In addition, transparency reduces
any perception that excessive profit is being made.
UTMB is also permitted to use administrative and other staff to manage other contracts
they have. In principle, all budgeted hours for the Dallas County/ UTMB contract should
be dedicated to work at Dallas County facilities unless the County gives specific
permission otherwise. The Medical Director at the Dallas County Jail, for example, is the
named "cluster" Medical Director for UTMB, which means that he is managing a total of
17,000 inmates in 9 different facilities. How much of his time can realistically be
devoted to the Dallas County Jail? He is managing facilities other than Dallas County
facilities even though his full time position is funded through the County contract.
Several other jail staff are also used by UTMB to staff other jails or prison facilities.
Given the lack of staffing at the jail, this practice should be prohibited by contract
language, but it is not. All budgeted positions should be fully engaged .at their budgeted
hours in dedication to County work. There should be a means verify this and there
should be a system of quarterly adjustments to the contract that account for under-filling
of contracted positions, or approved excess staffing with pro-rated monetary adjustments.
UTMB has provided a proposed staffing plan that attempts to set new staffing ratios in
order to improve services. The proposed staffing plan of UTMB is, in general, structured
like a prison program and not a jail program, probably because the vast majority of their
work is in prisons. Prisons have stable populations with significantly smaller intake
programs than jails. UTMB has not taken this into consideration in structuring their
program or their staffing proposal. Also, as a general practice, UTMB proposes or
actually reduces higher level trained staff with less well trained staff. While there are

20

]
]
]

]
]
]
]

]
]
]
]

]
]
]

times when this is not only appropriate but also fiscally efficient, there are other times
when the substitution will result in less than adequate clinical care and should not be
done. For example, using pharmacy technicians to pass medication is legal but for
certain high risk individuals (e.g. those mental health patients on psychotropics or for
those with tuberculosis), nurses are better able to perform symptom assessment and
adverse drug reactions.
The most serious deficiency of staffing is the lack of qualified physician staff. The
quality and mix of physicians and mid-levels is important because higher acuity patients
are best managed by someone who has been trained to evaluate and treat the conditions
patients have. There is no Board Certified Internist on staff, even though many people
require this type of service and much of the morbidity and even mortality involves
conditions that should be managed by an Internist. Most of the medical care of seriously
ill patients is managed by nurses and physician assistants, who are either not trained to do
so or are less well trained and can not be expected to adequately manage these types of
conditions. There is also virtually no physician supervision of medical care provided by
nurses, nurse practitioners, or physician assistants. This is a dangerous practice.
I question the rationale for several of the proposed UTMB recommendations for staffing
changes. Specific concerns include:
1. Increasing administrative staff should not be approved unless existing staff are
fully engaged at the jail and are not working other non-county UTMB programs.
2. Given the problems with intake screening, reduction of intake nurses and
substitution with emergency medical technicians is a questionable decision. The
intake screening process should be significantly revamped and improved.
3. Decreasing the ratio of physicians to physician assistants (from approximately 1:1
to 1: 1.7) will only increase the unsupervised practice that appears to result in poor
outcomes.
4. Maintaining tuberculosis screening staff at four persons is insufficient to conduct
screening in the manner stipulated. The County should realistically consider what
staffing is necessary and compare the cost of this staffing to the cost of instituting
an x-ray screening program.

]

5. The infirmary staff is grossly understaffed and does not include a physician who
is in charge of this unit. This unit should be directed by an Internist. The existing
arrangements create liability concerns and are dangerous.

]

6. The Suzanne Kays facility with over 600 inmates has only a halftime mid-level
provider. This is inadequate.

J
]

1

7. While nursing is currently understaffed under any circumstances, future staffing is
somewhat dependent on correctional practices. Most important of these is the

21

J
]
]
]
]

]
]

ability of officers to transport inmates for scheduled appointments and
evaluations. If correctional staff maintain current practices in movement of
inmates, significantly higher numbers of nurses will be necessary to adequately
care for patients.
8. The mental health program staffing consists of a response to security's need to
absolutely control inmate housing as well as any movement of the inmate. The
lack of appropriate housing for necessary clinical programming as well as
inadequate programming results in a staffing plan that is a response to a bad
program.
9. The electronic medical record and the continual movement of inmates by
correctional classification staff actually increases the burden of work on clinical
staff by forcing them to do significant clerical work to look up locations and to reenter data into the electronic record that has already been captured in a different
format. This is a system problem that might be solved by re-designing the
system. Alternatively, more staff will be required to address these inefficiencies.

]

]
]

]
]
]
]

]
J

For jails, the most important program elements of any effective health care system are:
accurate and early identification of those with medical and mental illness, continuation of
existing treatment regimens or initiation of necessary care, and appropriate placement of
those with a higher probability of harm if they are housed in general population. With
this in mind, specific goals of the Dallas County jail would be the following:
I. Improve the screening of inmates so that those with mental or physical illness are
diagnosed.
2. Establish an acute crisis unit for mental health patients that are severely disturbed
or suicidal. Admission and discharge to this unit would be by order of a
psychiatrist. Movement within this unit would be under direction of a
psychiatrist. The unit should allow for establishment of a therapeutic milieu.
3. Create an intermediate (step-down) care mental health unit for disturbed patients
and as a transition unit before sending an inmate with mental illness to general
population. The rules for entry and discharge and management would be similar
to an acute unit but the patient mix and therefore the therapeutic aims of this unit
would be different from the crisis unit.
4. Establish an infirmary unit that is managed similar to a skilled nursing unit or an
accredited correctional infirmary. This unit should not merely be for protected
housing but should include the capability of intravenous therapy for selected
patient.

]

5. Develop and implement a tuberculosis control plan and public health program and
develop realistic staffing for that plan.

]

22

l

]
]
]

6. Establish a realistic inmate health request (kite) and sick call process and
determine realistic staffing from that.

]

7. Establish a plan for managing patients with chronic physical and mental illness
for persons in general population and similarly develop a staffing plan from that
plan.

]

8. Review the mission of dental care and develop a staffing plan for that purpose.

]
]
]

]
]
J

J

]
J

I
J

]

]
)

Policies should be written to accurately describe how these missions will be
accomplished and then a staffing plan can be developed. Several general principles
should be used in guiding staffing choices.
1. The care of the most complicated medical patients should be provided by Board
Certified Internal Medicine physicians, not mid-level providers. Patients with
HIV infection should be managed by a physician with expertise in AIDS care.
Patients with tuberculosis should be managed by someone with experience in
managing tuberculosis.
2. The infirmary unit should include required physician notes on a daily basis for
persons who are not in the infirmary for purposes of housing only. All patients in
the infirmary should have a physician note weekly, at a minimum.
3. Patient encounters for clinical care should include a physical examination. This
will change staffing numbers because it appears that physical examinations are
.
not done for lack of time/staffing.
4. All patients (severely disturbed) on the acute crisis unit should have a psychiatrist
evaluation daily (with a note).
5. A PhD level psychologist or a psychiatrist should be the managing director of the
acute stabilization unit and intermediate care units.
6. All patients on an intermediate care unit should have a psychiatrist or
psychologist note monitoring progress on a weekly basis. If a psychologist is
monitoring the patient weekly, a psychiatrist should re-evaluate medication as
often as necessary but every month at a minimum.
7. The choice of staff member to pass dose by dose medication should take into
consideration the types of patients. For example, if pharmacy technicians pass
medications for tuberculosis (directly observed therapy) or for severely disturbed
mentally ill, they should receive specialized training in how to monitor side
effects and progress of disease and have a mechanism to report their findings to
the treating physician. Otherwise nurses should pass medications to this
specialized population.

23

]
]
8. The intake function should be re-designed so that early diagnosis with physician
evaluation of seriously ill patients occurs promptly after reception. This will
reduce overall work and liability to the county. For mental health patients, earlier
diagnosis and placement must occur as well.

]

]
]

]
]

9. Nurse staffing for the infirmary should be developed in accordance with what
services are to be provided on the infirmary. At a minimum, the infirmary should
always have a registered on the unit. If intravenous therapy is contemplated on
the unit a greater ratio of nurses needs to be present. If all persons with diabetes
are to be housed in protected housing in the current "infirmary", then staffing
needs to be adjusted so that capillary blood glucose monitoring can be performed
on a routine basis. At a minimum, all true infirmary patients should have a
nursing note with vitals daily. For infirmary borders (special circumstances)
vitals can be ordered less frequently.
IO. All sick call evaluations should be performed by registered nurses, at a minimum.

]
]
]

1

I
I

I
I

II. Mid-level providers should not be primarily responsible for monitoring the most
severely ill or disturbed patients. Their role should be restricted to monitoring
stable chronic disease and chronic mentally ill patients, and conducting sick call
evaluations as referred from nursing staff and to collaborating with physicians on
a team in managing certain individuals (e.g. tuberculosis patients, HIV patients,
diabetics, etc.).
12. The process of data input into the electronic record should be evaluated to ensure
greatest efficiency of higher cost provider time.

DELIVERABLE #5: Observe the physical housing accommodations (i.e. use of
isolation, physical restraints, etc.) of medical, mental illness patients and mental health
services in Dallas County jails and/or other Dallas,County detention centers.

Requirements of a medical and mental health program include identification of and safely
housing those who are seriously medically ill, mentally ill, suicidal, in medical or mental
health crisis, severely disturbed patients who are not in crisis, and those who have routine
medical or mental health problems that require follow up. Where these persons are
housed will increase or decrease the likelihood of life-threatening sequelae for individual
inmates. The type of housing arrangement is also integral to the treatment for suicidal,
severely medically iII, psychotic, severely disturbed patients, and tuberculosis patients.
There are significant physical barriers to inmates' access to mental health and medical
care in the Dallas County Jail. Housing arrangements in the Dallas County Jail actually
promote deterioration of clinical status. In addition, housing assignments are made for
the convenience of the custody staff rather than in order to safely house medically and
mentally ill persons. While housing assignment is done for the convenience of custody

24

]
]
]
]

]

]
]

]

]
]
]
]
]

1
]
J

J
J

1

staff, the housing assignment of individual inmates makes providing clinical care
excruciatingly cumbersome for medical and mental health staff to the point of
significantly reducing the efficiency with which they can manage patients. In addition,
housing assignments are so unsafe for some persons that they are life-threatening.
Jail cells that are designated for use by medical and mental health patients are assigned
by a sheriffs officer as part of classification. There are 4 medical or mental health
specialized housing designations: infirmary, suicide, closed behavior observation and
open behavior observation. Suicide is a variant of closed observation. A suicide cell is
one of the cells located in a tier in which an officer is assigned to monitor. Closed
observation cells are single cells in which inmates are locked up 23 hours a day alone.
These are mostly used for mental health patients but can house medical patients. Open
behavior observation cells are a tier of cells in which inmates are permitted to congregate
in the day room during parts of the day; they are used for mental health patients.
Infirmary cells are part of a single complex maze of tiers that are close to a nursing
station.
The cell assignment process is initiated at intake by the nurse identifying only those
persons identified as having a special need with a stamp that indicates suicide, infirmary,
closed behavior observation, or open behavior observation. This sheet then is forwarded
to a classification officer. In order to assign a cell, the classification officer will simply
look up on a computer for an open bed in the designated area and assign the inmate to an
open bed. The assignment is made in a matter of seconds. Except for suicide cells and
tuberculosis cells all cells within a class are considered equal regardless of where they are
located. This is a dangerous practice. For example, several of the closed observation
cells of each tier are completely out of sight of the corridors and can only be visualized
by walking into the recesses of the tier. Because officers make closed behavior
observation assignment without any consideration of acuity, it is only a matter of chance
as to whether a severely disturbed psychotic inmate is assigned to a cell where he/she can
be easily seen versus a cell that is hidden from view. The officer makes this assignment
without understanding or knowing the acuity of the patient. This can result in severely
psychotic inmates being out of visual sight, a practice that can result in poor clinical
outcomes. It is more important that clinical staff be able to see a patient than it is for
custody staff to see a patient because clinical staff are trained to recognize signs and
symptoms of mental illness while custody staff are not. No inmates are in continual
visual sight of clinical staff.
With the exception of infirmary cells, these specialized cells can change designation at
any time at the convenience of custody, based on the number of available cells. These
cells are therefore "virtual cells" in that the configuration of open observation cells and
closed observation cells is never the same month-to-month and are changed by
rearranging cell types by the classification officer on the computer. As an example, if a
tier of 13 cells has 4 closed behavior observation inmates occupying cells and another tier
of 13 cells has 9 closed behavior observation inmates in it, the classification officer can
combine these inmates into a single tier and free up the other tier for custody use. This
practice occurs on a daily basis. Thus, patients, especially those with mental illness (even

25

]
]
]

]
]
]

]

]
]

]

]
]
]
]

]
]

J
]
.~

those severely disturbed), are involved in a potentially life-threatening reassignment
process after which mental health staff must continually search and re-search for their
patients who are being moved about the jail by the classification staff without any regard
for their clinical status.
All of this is occurring in front of a computer screen far away from the site of care and
without any consideration from the mental health or medical staff except for the general
recommendation to house in a closed observation cell. Because of this situation, mental
health staff spend up to 30% of their work day in simply searching for their patients.
Needless to say, many patients become lost in this process to the detriment of clinical
care.
Closed behavior observation accommodations also have significant and serious clinical
consequences. For the most part, all of these types of housing arrangements are antitherapeutic and may be harmful. Suicide cells and closed behavior observation housing
with 23 hours in-cell lock up and the lack of stimulation (especially for those cells out of
visual contact) result in extreme isolation equivalent to a super-max prison. These types
of arrangements have known to result in psychotic behavior. To use these types of
arrangements for known psychotic inmates can only make treatment more difficult and
may prolong their disease or increase the severity of symptoms.
Accommodations for the purposes of mental health clinical examinations are equally
problematic. There is not a single proper clinical examination room in which a
psychiatrist or other mental health staff can conduct a clinical evaluation of a patient. I
have never been in a correctional facility where this has been the case. All clinical
evaluations are conducted in inmate housing units in the open and without any aural or
visual privacy. This arrangement is for the convenience of custody staff. It is
inconvenient for officers to have to have to bring these inmates to a remote clinic. There
are also insufficient officers to accomplish this even if it were required. Over time,
everyone has adjusted to this abnormal situation to the detriment of clinical care.
These types of accommodations also result in a lack of access of inmates to clinical care.
The rate-limiting step in obtaining access to inmates for the purpose of clinical
examination is the availability of escort officers. For inmates in these closed behavior
observation cells (of which there are well over 200 at any time in the North and West
towers), all mental health and medical staff must have an officer escort them when they
desire to see a patient. For the North and West towers of the Lew Sterritt Tower, there
are approximately 20 mental health staff (8 liaisons, 2.5 psychiatrists, 3.8 mid-levels, 5.5
RNs) who compete for the time of3 correctional officers who are assigned to this escort
task. This situation excludes medical providers. Compounding this problem, clinical
staff are not permitted to see inmates during meals and shift changes, further reducing
time available to see inmates. Only approximately 7 hours per day is available to see
inmates for mental health evaluations on a routine basis. Hypothetically, given these
constraints, over a 7 hour day, 3 officers continually' working would be spending
approximately 9 minutes per patient, including the time it takes to open doors and move
between tiers and cell blocks and engage different providers who desire their services.

26

]
]
]
]
]
]
]

]

]
]
]

]
]
]
J

J

]
]
~

This does not take into consideration the informal coordination amongst clinical staff as
they compete for these services. It is, therefore, impossible for clinical staff to
adequately assess the numbers of high-acuity patients on a daily basis. Evidence from
interviews with clinical staff as well as from chart reviews demonstrates the reality of the
lack of interval clinical evaluations even for severely disturbed patients or newly
incarcerated patients.
For inmates housed in these closed behavior observation cells, treatment is further
worsened because clinical staff are not permitted any contact except verbal contact with
the inmate through the food port. If the clinician desires to have the patient evaluated out
of the cell, a second officer must be present. This limitation virtually ensures that
inmates are rarely, if ever, examined out of the cell. There is no privacy for these
encounters. These are humiliating and anti-therapeutic encounters that merely give the
psychiatrist an opportunity to perform a partial assessment of the patient. One of the
major therapeutic modalities for schizoaffective disorders, for example, is psychosocial
interventions. This never occurs. The extreme enforced physical isolation of these
patients harms them. Custody concerns are dominant. It appears that the primary clinical
consideration is prevention of suicide. Therapy is not a priority.
Accommodations for those inmates in open behavior observation cells are also poor.
These arrangements are ones in which inmates can leave their cells and congregate in an
open day room for part of the day. These cells are used to house disturbed patients who
are not acutely psychotic or suicidal. The acuity or diagnoses of inmates is not
considered when making these assignments. The group of inmates who will be assigned
to particular cells is purely by chance. This situation can result in bad outcomes. For
example, if a schizophrenic psychotic inmate who talks to himself has just been stabilized
in closed observation and is being discharged to an open behavior cell, he may be
reassigned to an open tier with other inmates with manic disorder, depression or persons
who were intoxicated. The mix of patients may result in inmate-on- inmate interactions
that are harmful or violent. These units are not supervised in any meaningful way and
not under any type of clinical control. There is no group or special therapy on the units.
They are merely additional cells in proximity to mental health offices.
For all of the above reasons, correction of treatment of the mentally ill requires staffing
changes, changes to the orientation of custody staff toward the mentally ill, housing
changes and consideration of deficient officer'staffing.
Accommodations for medical patients are not much better than for mental health patients.
Some closed observation cells are used occasionally to house a medical patient. For
example, female TB patients are housed in North Tower closed behavior observation
cells.' However, for all remaining medical patients, the only option for special housing is
to be housed in the infirmary.
The accepted definition of a correctional medical infirmary is a designated area in which
inmates are within sight and sound of a health professional. These units should be
managed under written policies and procedures by a Registered nurse, 24 hours a day.

27

]
]
]
]

]
]

]
)

Admissions and discharges to the unit should only occur only under order of a physician.
Physician rounds on patients are specified at certain intervals for different classes of
patients. A medical record should be available on the unit and that part of the record that
is part of infirmary care is separate from the record for the remainder care. These units
are typically for inmates requiring skilled nursing care or advanced physician care that
does not rise to the level of requiring care in an acute care hospital. The groups of cells
that are designated as an infirmary in the Dallas County Jail do not meet these
specifications. The cells that are called the infirmary unit are merely sheltered housing
in which the most acute medical patients are housed.
The infirmary is a complex maze of nineteen separate housing units, almost all of which
are out of sight or sound of nurses. A nursing station and administrative offices are
located in a corridor that is surrounded by infirmary cells, but nurses cannot look into any
single cell unit from their station. Most of the cells are physically separated from the
nursing station by a winding maze of corridors. In addition, nurses must have officer
escorts whenever they need to access any inmate. There are three officers assigned to
cover the infirmary unit, but at anyone time only one is usually available to escort
inmates for infirmary movement or clinician visits to the inmate housing units. The
officer escort requirement restricts access of staff to patients and creates a barrier on the
infirmary unit for clinical care.
Within the infirmary, cell blocks have been assigned for alcoholics, diabetics, mobility
handicapped, hearing handicapped, and for other medical conditions. These 14 cell
blocks comprise approximately 180 beds. Infirmary housing is treated similar to any jail
housing with the exception that it is next to a nursing station so that inmates can,
theoretically, be watched more closely. Nurses do not round on patients daily except to
visually inspect the cells. Patients place kites (requests for care) rather than have nurses
check on them daily, as is the practice in a typical infirmary. The level of services is
minimal. There are no specialized policies and procedures for this unit. Nurse staffing
for these 180 inmates is 8 registered nurses for all three shifts, equaling approximately 2
nurses per shift, excluding any vacation or day-off coverage. Considering that nurses
must distribute all insulin and most medication dose by dose, and that all interaction with
an inmate must be accompanied by an officer escort, the barriers to access are so severe
that care on this unit is not much different than general population units. Nurses can
walk the corridor and look in on people, but clinical i~teractions are severely limited. For
this reason, most monitoring of patients is visual through the glass of the tier, accounting
for the dearth of vital sign assessments and monitoring (peak expiratory flow monitoring
or capillary blood glucose testing, and physical examination by providers) that should
occur on a unit of this type.
The unit does have a clinical examination room in the corridor adjacent to the nursing
station, however, most actual clinical encounters are conducted cell side or in the
dormitory style cell. The lack of access of nurses to the patients makes using intravenous
therapy or any other labor-intensive monitoring or therapeutic care plan very difficult.
In addition, the remoteness of some of the cells places potentially ill inmates out of
contact with medical staff. Several of the cells are ostensibly negative pressure isolation

28

]
]
]
]

]
]

]
]
)

I
I

cells, important for the effective isolation of infectious patients. However, as will be
described later in this report, these cells are not truly negative pressure cells. There is a
negative pressure sputum induction room, but it is no longer used for that purpose.
Finally, the current dental unit is not adequate. This unit consists of a single chair and
light, both of which are in poor repair. The room is not clean. The only procedure
performed is extractions. There is no sink, no equipment except hand tools, and no
receptacle for the patient to expectorate in the event this is necessary. The dentist
operates with rudimentary tools and equipment. This situation is very crude and
inadequate.

DELIVERABLE #6: Evaluate and analyze the psychotropic drug usage ofdetainees for
effectiveness and appropriateness & review pharmaceutical data and processes (e.g.,
utilization, cost, comparison ofdiagnoses to drugs prescribed, etc.).

I was unable during my two separate visits to the Dallas County detention facilities to
obtain any pharmacy utilization data, despite requesting it on both visits. For this reason,
I was unable to analyze psychotropic or other drug utilization or the effectiveness and
appropriateness of that utilization. Nevertheless, there were several observations that
should result in follow-up action.
Patients who come into the jail, who are on medication for chronic illness or serious
mental disorders, do not have their medication promptly restarted. The current system is
dependent on whether an inmate brings labeled medication into the jail with them (which
will seldom happen) and on the review of reminders that physicians perform on a
Monday through Friday basis. This type of system will inevitably result in missed
medication and delays, especially when an inmate comes in on a Friday and physicians
are unavailable until Monday morning.
Also, patients returning from Parkland Hospital often come back to the jail on medication
that is not covered on the UTMB formulary. As a result, there may be delays until their
medication is re-initiated. The formulary used by UTMB should be approved by the
Parkland contract monitor to prevent these problems. In addition, an improved system
should be in place to coordinate medical transfers returning from Parkland Hospital to the
jail. This will be covered in a subsequent section.
The formulary that exists for psychotropic medication reduces reliance on atypical
psychotropic medications and SSRI medications other than generic fluoxetine. While the
use of these drugs should be monitored on a continual basis, changing medications that
inmates come into the jail on should not be summarily changed until a clinical evaluation
has occurred.

29

J
]
]
]
]
]

]
]
]
]

]
]
]
]

)

j

]
]

I

DELIVERABLE #7: Observe the processes for medical intake and assessment
(particularly focusing on screeningfor tuberculosis and other communicable diseases) at
Dallas County jails and/or other Dallas County detention centers.
In 2003, there were L4,874 cases of contagious tuberculosis reported in the United States,
with L,594 (l L%) in Texas. Correctional facilities have consistently higher rates of
tuberculosis than the free-world population. In Texas, 112 cases of tuberculosis (7% of
Texas' cases) were reported from a correctional facility at the time of diagnosis. This is
second in the country only to Arizona in terms of the proportion ofTB cases diagnosed in
correctional facilities. The majority (53%) of reported tuberculosis cases in the United
States are now accounted for by foreign-born individuals and approximately 26% of these
are from Mexico. Hispanics account for approximately 28% of all cases of reported
tuberculosis in the United States. Both of these population groups are heavily
represented in Dallas County detention facilities. The expectation is that there would be
a high prevalence of tuberculosis in the Dallas County Jail.
Dallas County detention facilities would therefore be classified as high-risk correctional
facilities by Centers for Disease Control and Prevention. For a high-risk facility, the
Centers for Disease Control (CDC) recommends that all inmates coming into correctional
facilities receive screening for symptoms of tuberculosis (cough, weight loss, night
sweats, etc.), screening for active tuberculosis disease, and screening and treatment of
latent infection when it can reasonably be coordinated with local public health
departments upon discharge from the correctional center.
Screening for symptoms consists of asking questions about tuberculosis symptoms
(cough, weight loss, night sweats, etc.) of all inmates coming into the facility. Those
inmates that provide positive responses would be evaluated by a clinician for other
evidence oftuberculosis. Often, this process is linked to public health databases so that
those persons already known by local health departments as having TB can be identified
when they come into the detention facility. New York jail facilities, for example, have a
linkage with the city health department TB database so that inmates who are suspicious
for TB can have their TB record checked if one exists. Symptom screening for
tuberculosis is not performed in the Dallas County Jail.
In addition to symptom screening, all inmates should be screened for active tuberculosis
disease. There are currently two methods for doing this. One method is to perform
Mantoux skin tests on all inmates who enter the facility. Those inmates with positive
tuberculin tests would then obtain a follow up chest radiograph to assess for active
tuberculosis disease. This process requires substantial coordination in order to be
effective. Inmates are distributed all throughout the jail after intake. Additionally, they
are frequently moved by custody for a variety of reasons and they are often not available
in their housing unit because of court or other visits. The skin test process therefore
requires that health care staff locate the inmate. After placement of the skin test, the staff
must return to the inmate in approximately 72 hours to read the test. If the test is positive
(indicating an infection), a follow up chest radiograph must be done to determine if the
inmate has active tuberculosis disease.

30

]

]
1
)

]

]
]

]

]
]
]

]
]
]
J

]
]
J

I

In comparable jail systems, the proportion of inmates with positive tuberculin skin tests is
approximately 15-20%, or higher. This will result in large numbers of inmates requiring
chest radiographs. Those with abnormal chest radiographs should be isolated in negative
pressure rooms until the diagnosis ofTB has been excluded. Because of the time
requirements of the screening process (locating inmates, 72 hour delay to read the test
after placement, necessity of a follow up x-ray), the skin test method of screening for
active disease has been known to take approximately two weeks before an active
tuberculosis case is detected and many persons will enter and leave the jail unict'entified.
This also results in many potential exposures of other inmates and staff.
The principal aim of tuberculosis screening in a jail is to identify infectious persons and
to reduce exposure to other inmates and staff. Therefore the inherent delays and 1
inefficiencies of skin testing have prompted some large jail systems to use alternate
methods. Some facilities have initiated programs of screening all incoming inmates with
a chest radiograph. Cook County Jail, Los Angeles County Jail and Harris County Jail all
have programs of using radiographic screening for tuberculosis. At Cook County Jail,
the time required for diagnosis of tuberculosis was reduced from approximately 17 days
to 2 days using screening radiographs and the numbers of persons identified with
contagious tuberculosis tripled.
The Dallas County Jail TB screening program is basically non-existent. Symptom
screening for tuberculosis is not performed at any point in the intake screening process.
This can be corrected by including symptom screening questions in the intake
questionnaire. In addition, inmates are not questioned as to whether they have had a
previously positive tuberculin skin test in the past. To screen for active disease, UTMB
performs skin testing but does not do this for all inmates and delays the process so that
effectively many persons are discharged before the test can be applied, read, or followed
up on as indicated. Also, there appear to be errors either in recorded skin test results or in
performance of this test.
When the jail health services were operated by the Dallas County Health Department, 13,
staff members were assigned to the TB program to screen for active TB disease. This
number has been recently reduced to 4 (a nurse, one patient care assistant, and two
certified medical assistants), yet they appear to be applying and reading the same number
of skin tests as the 13 individuals from the Health Department did previously, although
the number of recorded positive skin tests is now dramatically reduced. This simply does
not make sense.
The UTMB TB staff work Monday through Friday. They obtain a list from the custody
computer of persons who have been booked. The custody computer is used because it
can provide the location of the inmates. They do not screen persons unless they have
been in the jail for at least 3 days. After identifying those who have been booked and
incarcerated for 3 days they go to the inmate's cell and apply the skin test in the cell.
This system is not very effective and is evident in· reported statistics.

31

]
]

]

]
]
J

I
I
I
I
I

The State Department of Health sent a letter to UTMB in October of this past year
warning them that they were non-compliant with tuberculosis reporting requirements of
the Health and Safety Code of State Statues. The jail then began reporting statistics.
Despite not knowing the number of active tuberculosis that occurred prior to October of
2004, the jail did report TB skin test data for the past year. These data indicate that for
the past eleven months, 27,856 TB skin tests were applied and 22,574 skin tests were
read. Of these, only 274 were positive greater than 10 mm. This rate of skin test
positivity is approximately 1%, a result that, in my experience, is simply not credible. It
can be expected that the rate of skin test positivity would be somewhere between 15%
and 25% or higher.
In a discussion with a University of Texas Southwestern (UTSW) Infectious Disease
specialist at Parkland Hospital who is head of the Infection Control Department, the
employees at Parkland Hospital have a 20% skin test positive rate and house staff
(resident physicians) had a 2% positive rate. It is simply not believable that house staff
have a greater rate of tuberculosis skin test positive reactions than inmates in the jail by a
factor of 2, and that employees at Parkland have a rate that is 20 times greater than the
inmates. The jail data are either erroneous or result from bad technique.
In 2000, the Centers for Disease Control from Atlanta, in cooperation with the Dallas
County Department of Health and Human Services, used DNA fingerprinting to attempt
to see if there were any similarities between tuberculosis cases that had occurred in
Dallas County and the surrounding areas in the years around 1998. Not all cases were
tested. The study identified a cluster of cases (cases with identical DNA fingerprintingindicating transmission from a common source) which they called cluster 242. There
were 76 total cases in cluster 242. The most common ecology risk factor of potential
transmission was having been incarcerated in the Dallas County Jail. The study found
that 28 of the 76 total cases in the cluster (37%) had been incarcerated at some point,
indicating that the jail was not only a high risk TB facility but that is might be playing a
role in amplifying tuberculosis spread into the community. Nineteen of 41 patients
whose tuberculosis was infectious at the time of diagnosis either were or had been
inmates at the jail, although only 3 of the nineteen were diagnosed at the jail.
This situation appears to be an ongoing problem. From January to October of 2004,
Parkland Hospital reported 83 cases of tuberculosis, with 19 of these having had been
inmates at the jail at some point and 7 of the 83 were admitted directly from the jail when
they were diagnosed. These data reflect that tuberculosis remains a major problem at the
jail, that patients with active disease are probably missed by screening and the statistics
reported by UTMB indicating a I % skin test positivity rate are not credible and reflect a
broken screening program.
Using the UTMB jail skin test data and jail admission data, it can crudely be estimated
that UTMB screens about 30% of incoming inmates but about 70% of those staying more
than 5 days. This misses significant numbers of persons. But the test itself appears to be
performed so poorly that it cannot adequately screen for pulmonary tuberculosis.

32

]

]
]
]
]
]
]
]
J

]

]
]
\ ..

I
,
;

)

I
1
]

]

I

Several conclusions can be drawn. The jail does not maintain accurate statistics for
tuberculosis. They do not adequately screen for this disease. This inadequacy results in
the jail being a focus of amplification of tuberculosis spread both within the jail as well as
into the community at large. Persons with TB from the jail are part of a larger pattern of
tuberculosis spread into the community that has been verified by very accurate data.
The problems with TB control extend beyond simply screening for tuberculosis and
include the clinical management of disease. In fact, the clinical management of TB
disease may be partly responsible for the potential for TB spread within the jail. The jail
personnel do not manage patients with active tuberculosis disease appropriately. After
persons are identified who have positive skin tests, the nurse refers them to a provider
designated as the TB physician. In the recent past, a nurse practitioner was responsible
for this function but now a physician has been given this responsibility. This physician
has no experience in managing patients with tuberculosis and he has so much else to do
that he actually spends very little time managing people with tuberculosis. There is a
nursing protocol to perform routine chest x-rays for persons with positive skin tests and
the physician is designated to follow up on these, but at the time of my visit there was
confusion regarding where the x-ray reports should go and it appeared that over the
recent time period there has been no follow up of abnormal x-rays. The physician
seemed unaware of any TB protocol for management of active disease. The physician
who manages TB does this as an additional assignment and indicated that he has
difficulty in getting tuberculosis smears back from the laboratory. In my review of
several tuberculosis records, I did not find anyone who had a documented tuberculosis
sputum smear in the record.
The records of tuberculosis cases demonstrate significant problems with TB identification
and management. One patient gave a history of previously positive skin test at intake.
Four days later, a chest x-ray was ordered that was performed 4 days after that (8 days
after intake) that showed probable tuberculosis. Cultures were ordered the following day.
There were no follow up notes and it appeared that the inmate was discharged from the
jail. I could not identify the cultures as done in the medical record. The same inmate
returned to the jail seven months later. He told the nurse at intake that he had been
treated for active tuberculosis in the past year. He received no other evaluation.
Symptom screening for his disease did not occur even though he had a history of active
disease. He did not have a review of his prior medical record nor was an x-ray obtained.
Either one of these would have identified active tuberculosis: Instead, he was placed on
the mental health unit because he had bipolar disorder. Over a week later the inmate
coughed up a large amount of blood and was sent to Parkland Hospital where active
tuberculosis was identified. There was no contact tracing of the persons exposed in this
case and multiple inmates and staff were unnecessarily exposed to contagious
tuber:culosis and were not tested to see if they had acquired the disease.
A second patient gave a history of HIV disease. There was no intake screening for
tuberculosis despite the fact that this inmate had a significant risk factor for tuberculosis
(HIV disease). Two days after incarceration, he told a physician assistant that he had
been on two drugs for tuberculosis in Fort Worth. There was no physical examination of

33

J
]
]
]
J

]
]

]
]
]
]

]
]
)

I
1
1
J

I

the patient. He did not get an immediate x-ray, physical examination, or smears with
cultures for tuberculosis. Five days later, without explanation, he was placed on two
tuberculosis drugs without isolation. There were no medication records demonstrating
that he received the medication. An x-ray was finally done 11 days after incarceration. It
showed a cavity suspicious for tuberculosis. He should have been immediately isolated
in a negative pressure room until smears were negative for tuberculosis. Instead, he was
kept in general population and has been incarcerated for 4 months and has not had a
physical examination, tuberculosis smear, or follow up x-ray. This is inadequate
management and may result in spread of tuberculosis.
Another inmate did not have a medical intake screening, so I assume that the officers
either did not identify any medical conditions or did not screen the patient. No
tuberculosis skin testing was done for this patient through the intake process. At one
point several months into incarceration, the patient became extremely ill and the illness
was not identified until the inmate was in extremis. The patient was admitted to Parkland
Hospital and found to have idiopathic thrombocytopenic purpura, a condition that
resulted in treatment with high dose steroids. Steroids can cause multiple complications
(including tuberculosis) that require that the patient be monitored. At a minimum at the
start of his treatment as a baseline, a tuberculosis skin test should have been done. In any
case, he had a significant disease for which he should have been monitored. Yet, for
seven months since discharge from the hospital, there was only one examination in the
medical record. Seven months after return from the hospital, the patient developed 104.6
fever with a pulse of 154 and a respiratory rate of 40. This was an extremely advanced
illness. He told a nurse that he had been dizzy and weak for 2 months yet he had not
been examined in follow up from his previous hospitalization. He went to Parkland
Hospital, where pulmonary embolism and contagious pulmonary tuberculosis were
diagnosed. The pulmonary embolism required therapy with a blood thinner for several
months. The tuberculosis required therapy with four different medications. The meaning
of this is that the inmate either had TB infection when he came into the jail which reactivated while on steroid medication in which case jail intake screening missed this, or
he newly acquired TB disease while at the jail. Either possibility is problematic. After
six ~eeks at the hospital he returned to the jail and was placed on a mental health unit.
He wasn't examined upon return from the hospital and didn't receive either tuberculosis
medication or his blood thinner for approximately three weeks. This was dangerous. For
a period of five months he wasn't examined. His anticoagulation (thinning of the blood)
statuS should have been checked monthly but over the five month period was only
checked once. The one time it was checked it showed inadequate blood thinning that
should have resulted in an adjustment of his medication. There have been no follow up
x-rays or tuberculosis smears that should have been done to monitor his tuberculosis
disease.
Another patient had a mental illness and arrived at the jail with packets of four
tuberculosis drugs from the Austin Department of Health but was sent to the George
Allen jail. A nurse at the George Allen jail identified the problem and wrote that the
patient "was to be housed in a contagious single cell," so the patient was transferred to a
single cell in the North Tower. Initially, the patient was started on only one tuberculosis

34

J

]
]
]

]

]
]

]
]

]
]
J

]
~,,.

]
\

i
\r

I'.•
~~

,

I)
]

drug (INH). Two days after incarceration, a physician from the Department of Health
communicated with the jail that the patient should be on at least 3 anti-tuberculosis drugs.
The jail records document only two anti-tuberculosis drugs were prescribed. I could not
determine how many drugs the patient actually took, but it appears that the patient only
took two drugs. The patient's history was that 3 weeks prior to incarceration, the patient
was diagnosed with active non-cavitary tuberculosis. Her tuberculosis smears were
positive. About ten days after incarceration, after having been transferred to the North
Tower, a nurse identified that the patient had not been taking medication since she had
come into the jail. Subsequently, the patient agreed to take her medication, but I could
not verify that she was taking more than two drugs. A two drug regimen would
inadequately treat the tuberculosis and might result in resistant organisms. About six
weeks after incarceration, a physician performed the first physical examination of the
patient. The patient was described as thin. If her tuberculosis had been treated, she
would be expected to gain weight. She no longer had cough. The doctor discontinued
isolation but did not do so based on negative smears, cultures or x-ray. In fact, no follow
up testing of her tuberculosis has occurred. Her mental illness was never addressed.
Two referrals from the physician to mental health were unanswered.
The conclusions from these chart reviews are the following.
• Tuberculosis disease management is inadequate.
• Directly observed therapy (the standard of care of tuberculosis management) does
not appear to be the standard at the jail.
• Sputum and culture testing is underutilized at the jail and there is no evidence that
most patients get appropriately tested.
• Management and follow up of persons with active tuberculosis disease is not
competently performed.
• Patients with serious disease are not examined at appropriate intervals.
An additional problem with tuberculosis management is that there is no Infection Control
work that is done at the jail. A nurse practitioner is named as the person in charge of
infection control but that is merely because she monitorsHIV patients. No real infection
control work is being done. Typically this would consist of monitoring and reporting
rates of contagious and infectious diseases including tuberculosis, monitoring the
negative pressure isolation rooms, performing contact tracing for persons who may have
been exposed to an infectious case of tuberculosis, monitoring tuberculosis conversion
rates of employees, tracking infections such as MRSA, syphilis, Chlamydia, gonorrhea,
HIV, and any other reportable disease.
In this regard, an infection control team visited the jail from Parkland Hospital several
years ago and identified that the negative pressure rooms were not actually functioning as
negative pressure rooms for several reasons. Doors were imperfectly sealed and included
food ports that disrupted containment. Because inmates could set off alarms indicating
that the negative pressure was not working, the rooms were disabled. Thus, when
inmates are housed "in isolation" at the jail, they are probably not in an effective negative
pressure room. This can contribute to the spread of disease.

35

J
J
]

J
]

J
J
]

I
]

]
]

)
)

Il
~.

l
.

In addition, if an inmate is diagnosed with active contagious tuberculosis at the jail and
has been incarcerated and undetected for any period of time, other inmates and
employees who have been exposed to this individual are not tested to determine if they
have acquired the disease from the infected individual. This is called contact
investigation and is a fundamental aspect of tuberculosis control in any institution and
should be instituted. It would be important for employees and inmates who have been
infected to be offered preventive therapy for tuberculosis.
The lack of infection control practices surface in other areas as well. The Infection
Control chief at Parkland Hospital indicated to me that methicillin resistant staph aureus
(MRSA) infection is a significant problem for inmates coming to Parkland Hospital and
that this becomes a problem for the hospital because it can spread within the institution
and then into the community. Not all areas of the jail consistently report MRSA
statistics, but for 5 months during the 2004 calendar year, over 200 MRSA cases were
reported per month. This is an extraordinarily high number of cases, even for a
correctional facility. In a site survey of the jail, the infection control team at Parkland
identified several potential sources of spread. In one area where inmates were strip
searched, chairs or other touch items which had potential for skin contact between
multiple individuals were not disinfected between use. Hand hygiene for medical staff
was very poor due to the lack of sinks to wash hands. The fact that most encounters
occur cell-side means that it is virtually impossible to wash hands between patient
contacts if such contacts occur. Alcohol based hand cleansers, which are currently the
recommended manner of hand disinfection are not used at the jail. Also, it is not clear
whether MRSA statistics are based on actual cultures or on presumed infection. This
should be clearly stated. I did not have an opportunity to evaluate charts of specific
cases, but did notice one episode of supposed MRSA in a record reviewed for mental
health purposes.
In that case, a patient in a mental health observation cell, placed a health request for a
"huge pimple, spider bite, or staph" on his back. A nurse saw him and wrote that he had
a possible staph infection. Without a physician examination or culture of the abscess, a
two-week prescription was written for bactrim, an antibiotic. There was no physician
signature on the prescription. The wound should have been cultured, and a physician
should have seen the patient.
Finally, there is no meaningful sexually transmitted disease screening performed at the
jail, even though there are most likely high rates of certain conditions for people coming
into the facility. This is a missed public health opportunity. Nationwide, jails that screen
for sexually transmitted disease provide a significant public service to their communities
in reducing the prevalence and spread of these diseases. This is particularly true for
Chlamydia, gonorrhea and syphilis infections. Suspected rates of these diseases should
be ascertained in conjunction with the Department of Health and a collaborative
screening program should be initiated that is fiscally possible. Currently, intermittent
screening of adolescents in the juvenile facilities is being done and for males 1-2% are
positive for gonorrhea and 15-20% are positive for Chlamydia. These diseases typically
are much higher in the female population. This represents a significant burden of disease.

~.

l

·36

J

]
]
]
]

]
]
]
]
]

]

]
]
]

I
]

1
]

I

DELIVERABLE #8: Review and analyze multiple data sources and conduct interviews
with key administration and/or management of Dallas County jails and/or other Dallas
County detention centers, which shall include key administration and/or management of
the Dallas County Hospital District, d/b/a Parkland Health and Hospital System
(hereinafter, "Parkland") for the effectiveness of the similarities and/or connections
between levels ofcare.

Many local agencies have the potential to interact with jail staff to improve care for
inmates at the jail. In the Dallas County Jail, there is a failure to establish viable linkages
with these key groups. This is a dropped opportunity to improve services. Most of the
agencies we interviewed indicated that they felt there was no outreach from UTMB to
them and that the relationship could be improved. There are almost no interactions
between the Dallas County Department of Health and UTMB regarding public health
issues at the jail. While Department of Health staff continue to visit the jail to perform
minimal sexually transmitted disease outreach, this work does not seem, to be coordinated
with UTMB staff to try to improve effectiveness. Administrative staff at the Department
of Health indicated that UTMB staff do not communicate with them.
Parkland Hospital has a case manager at the hospital who is assigned to case-manage
inmates. There are regular meetings between Parkland and UTMB, but issues of critical
importance such as tuberculosis at the jail, multiple re-admissions for selected patients, or
communications on other seriously ill patients seem not to be part of these regular
meetings. Including a collaborative clinical review of problem cases in the format of a
clinical conference may be beneficial, particularly to the jail staff.
Large numbers of inmates with mental illness are enrolled in NorthSTAR, the
coordinating agency for mental health providers who care for the mentally underserved in
the greater Dallas area. For the first four months of 2004, 31% of the inmates referred by
officers to nurses and subsequently referred on for mental health evaluation and treatment
in the jail were NorthSTAR enrollees. Nearly 70%ofthe NorthSTAR-enrolled inmates in
the jail are served by one provider, Dallas Metrocare Services. Based on interviews with
staff, this organization has both empathy and interest in the jail population. They have
assigned 2 case-workers to the jail to identify NorthSTAR members in the jail and to
serve as the case-managers for all NorthSTAR clients while they are incarcerated. These
individuals perform social work services for inmates while they are incarcerated, attempt
to arrange for aftercare when inmates are discharged and are involved in attempting to
facilitate information transfer to jail personnel regarding prior treatment their clients had
received at NorthSTAR clinics. Because such a large percentage of inmates are cared for
by this organization, and because of their clear dedication to the population, I would
strongly recommend a greater role for them in the mental health program for this
population. One of their programs, the Special Needs Offender Program (SNOP) is
specifically meant for mentally ill prison offenders who are discharged to the community
and is a potential resource for discharged inmates.

37

]

]
]
]
]
]
]
]
)

]

)
:.'

)

I
~t·
~~

J

I
I

1
]
]
[

An area where many organizations have attempted to collaborate in reducing the numbers
of persons incarcerated with mental illness is Project Divert. This jail diversion program
is a voluntary program in which recently booked inmates who have a mental illness are
identified as such and redirected to mental health programs in lieu of incarceration.
Three barriers exist to improving the numbers of clients involved in this program: the
criteria for patient selection based on the criminal charges is said to be too restrictive,
many of these persons may be homeless and unless housing is part of the aftercare
arrangement recidivism may not be reduced, and lastly, inmates must be enrolled in
NorthSTAR in order to participate.
Of inmates with mental illness entering the jail, only approximately 31% are enrolled in
NorthSTAR. NorthSTAR is the equivalent of a managed care organization for the poor
with mental illness.
Even though inmates are potential clients of NorthSTAR
organizations, inmates who are not already enrolled in NorthSTAR are excluded from the
diversion program, reducing potential clients by up t070%. The exclusion of almost 70%
of inmates from this program creates a situation where those who may most benefit from
diversion are excluded by virtue of lack of having a mental health provider. Project
Divert aims to serve the population of persons with serious mental illness (schizophrenia,
bipolar disorder, etc.). These persons by virtue of their disorder (and possibly combined
with homelessness) are less likely to be organized enough to enroll in any type of
ongoing mental health treatment. Consideration should be given to opening up the
diversion program to anyone with serious mental illness, regardless of whether they are
already enrolled in a mental health program and to enrolling them if accepted into
diversion. The effect of the current system is to ignore a large segment of the mentally
ill; those who have no mental health coverage or provider. This is exacerbated because
there is no advocate group for the uncovered, uninsured inmates with mental illness. The
Dallas Area NorthSTAR Authority (DANSA) is chartered to be an oversight authority
over mental health care in the greater 7 county area. However, they do not believe that
their oversight extends to inmates, even though inmates reside in the seven county area
over which they have authority. The effect of this perception is that inmates not enrolled
in a NorthSTAR mental health program (which is 70% of the inmates) do not have the
benefit of the oversight of DANSA and do not obtain the benefit of diversion.

DELIVERABLE #9: Assess the existence of chronic disease management, specifically
observing and analyzingthe first five (5) days ofincarceration for ER utilization.
Inmates with chronic diseases, if identified, are primarily managed by nurses. As
discussed previously, many inmates with medical conditions are missed at intake
screening. Those who are identified with chronic disease and are on medication have
their diseases and medications listed on the Central Intake Evaluation Form. These are
sent to clinical areas and reviewed by nurses. The expected prevalence of chronic illness
in correctional populations exceeds the number of persons screened by the nurse in
intake. Thus, an unknown number of persons with chronic disease are missed at intake
but can be conservatively estimated at 10% of incoming inmates.

38

]
]

]
]

]
]

]
]
]
]
]
tt-

]r

)
K

On medical units, nurses enter all data of persons who require medication into the
electronic medical record and reminders are sent to physicians for those who are on
medication so that the doctor can prescribe medication. One exception to this process is
persons with epilepsy, for whom intake nurses typically call the doctor for a verbal order
for a prescription. However, this process for all the rest of the inmates with chronic
disease results in frequent delays of medication. For example, if a person comes in on
Thursday evening and a nurse writes a reminder to the physician on Friday, the physician
may not get the reminder until Monday and the patient may not get their medication until
Wednesday. The inability to promptly get medicines ordered is a significant problem and
was evidenced in chart reviews. The lack of continuity of medication along with the
absence of physician evaluations can only promote deterioration of clinical status
ultimately resulting in unnecessary hospital and emergency room visits.
Persons with chronic illness should be examined by a physician to determine the clinical
status of their disease. The process of arrest and incarcerations is disruptive for the
inmate. Frequently, when they are arrested they will not have their medication with
them. They may not have had consistent care in the community. Thus, their disease
status may have deteriorated from a usual state or may not be in good status. For these
reasons, it is important that inmates with chronic disease see a physician who knows how
to manage the illness they have. This does not occur at the Dallas County Jail.
At intake, if an officer sends an inmate to the intake nurse, the nurse does only a brief
assessment. The main function is to decide if the patient requires hospital admission or
special housing. The nurse will send persons with diabetes to the infirmary, pregnant
women to the Allen Center, and persons with disabilities to the infirmary. Based on the
nurse's judgment of the acuity of the patient, an inmate may be sent to the infirmary for
monitoring. But there are only approximately 180 beds in the protected housing
infirmary unit. If 20-30% of inmates have some form of chronic illness that means that
well over 1000 persons with some type of chronic illness are housed in the remainder of
the jail facilities.
The method of monitoring these patients is poor to non-existent. Their disease, including
medications, is listed on the intake form and delivered to the nurse's station in their new
housing unit. Subsequently, the nurse will review the sheet, send a reminder to the
physician for medications, and make a judgment about whether the physician needs to see
the patient. The physician will review the electronic reminders and the electronic record
charts of patients referred by the nurse. This can be a large number of charts. I was told
that this is between 30 to 60 records but can be as high as 100. Of these, the physician
may see between 10 to 15 patients a day. In this system, therefore, physicians are only
seeing a fraction of the persons with chronic illness accounting for the virtual absence of
physical examinations discovered on chart reviews. This is a significant lack of access to
care. In discussing chronic disease management with one of the physicians he indicated
that for chronic disease charts from intake he will only pick the most severe cases to
evaluate. He said mostly nurses manage these patients and he oversees through emails.
Most of the 10-15 patients he sees are actually not patients with chronic disease but are
episodic problems (back pain, boils, etc.). Of the 10-15 patients he sees he estimated that

39

]
J

]
]
]

]

]
]

]
]
]
','

".

~.'
1
j
r

~.'.;
f'
.....

~..

j

Rl'
fl"."

j
J
J
]

about 3-5 have chronic disease. He does very little lab testing on chronic illness patients
and, because of the infrequency of vis its, when he does prescribe medication, he orders
prescriptions for 12 months.
There are no policies, procedures or chronic care guidelines that are used in the
management of chronic diseases. There is no mechanism to track individuals who have a
chronic illness, there are no required interval visits, there are no required laboratory tests
used to monitor individuals with chronic illness, and there is no mechanism in place to
assign acuity to patients with chronic illness. Therefore, management of these
individuals is similar to that of all other inmates. Inmates with chronic disease are seen
when their condition deteriorates to an urgent status while the others usually are
discharged prior to being seen. This type of system, as evidenced by chart review, results
in disease deterioration to the point of requiring hospitalization in multiple individuals.
Management of chronic diseases mostly consists of physicians reviewing the electronic
record of nursing notes describing interactions with the patient. Very few persons with
chronic illness are actually physically examined.
Even when persons are hospitalized, upon their return they are not usually examined.
Their record may be reviewed but, based on chart reviews, it is not even clear that this is
occurring. This lack of attention to persons with very problematic disease results in
repetitive hospitalizations that, with sound ambulatory management, would be prevented.
Review of selected diseases illustrates the manner in which chronic diseases are managed
at the jail.
Asthma and hypertension are typically the most prevalent chronic diseases in the
correctional populations; asthma being more prevalent in the younger population with
hypertension becoming more prevalent as the population ages. For asthma, monitoring of
symptoms and peak expiratory flow rates are standards of care. Yet this is not done at the
Dallas County Jail. At intake, a pulse oximeter is used to evaluate persons with asthma,
but this test is not a useful monitoring test for this disease. Nurses at intake will not
provide an asthma rescue inhaler unless the oxygen saturation is abnormal, but this test
would not be abnormal in an asthmatic until the patient was so ill as to require intubation.
This is a bad practice that should be discontinued. Although peak flow meters are present
in the jail, I did not find a single instance in which they were used. One patient who had
ten admissions to the hospital for out-of-control asthma was Clearly not getting his
prescribed medication as ordered.
A nurse practitioner manages all HIV disease at the jail. She has about 100 patients on
her caseload with HIV infection at anyone time and about 30% rotate out of the jail
weekly. At intake, the medications get noted and the nurse assigns the patient to an mv
tank or to general population ifthe inmate wants. There is no female HIV tank. The HIV
tanks are in the west tower on the 5th and 6th floors and are usually full. The intake nurse
assigns housing. After patients arrive on the floor, as with other patients, the nurse will
write an email to the nurse practitioner documenting the arrival of the patient on the floor.
Nurses will also email to her regarding other problems the HIV patients have. She gets
about 40-50 emails a day. There will be 5-10 new patients a day.

40

Patients with chronic illness, including HIV, are usually not examined in a clinical
examination room. They are mostly evaluated cell side. So, similar to mental health
clinicians, the nurse practitioner must first locate her patients. The first task in the
~orning is to identify from the correctional computer the list of persons housed in the
HIV cells and she compares that to her existing list to identify new persons. Her
interaction with new patients is a cell side interview. If they know what their CD4 count
is and seem to know a lot about their disease, she will start medication. If not, she will
check with their civilian provider. She may do an HIV test to verify HIV positivity.
Most people get medication in a week to 10 days.
She sees as many people she can. She thinks she sees all patients who remain at the
facility by the end of week. Inmates are permitted to keep their HIV medication on their
person. There is a state program in existence the supplies HIV medication for a small copay fee available to inmates, but it appears that this program is not accessed. She does
not use UTMB's telemedicine service, even though it is available. She works on her
own. There is no medical supervision over her work. When she is off on vacation, the
coverage doctor is a retired surgeon who has no experience managing HIV patients. He
was not trained in HIV management but the nurse practitioner told me that he has a
"John's Hopkins HIV book," so at least he has a reference. She works with doctors at the
Amelia Hart clinic at Parkland when she has a problem and calls them about 2 times a
month.
Because of the large caseload of patients, the difficulty in finding and interviewing
patients and because she does not work every day of the week (she works only Monday
through Thursday), people are not re-started on their medication promptly after
incarceration. This can be a problem in promoting resistance. Not all individuals are
examined thoroughly and physicians are not involved in the management of persons with
this disease. Management of patients with HIV is relatively complicated due to the
variety of opportunistic infections that can occur and because of the complicity of drug
interactions that exist for HIV medications. In addition, the choice of what medications
to use is complicated. For these reasons, it is recommended that an HIV expert be
involved in the management of all patients with HIV infection. This does not occur. The
nurse practitioner can call a physician if she wants, but this occurs for only a small
number of patients. UTMB had plans to utilize telemedicine for HIV care but this has
not materialized. HIV patients at the jail-used to be followed in the Amelia Hart clinic, a
Parkland/UTSW staffed HIV clinic, but this is no longer occurring. Currently, persons
with HIV do not have adequate access to a physician.
Insulin dependent diabetics must be housed in the infirmary. There are several infirmary
tiers in which they can be housed. Patients on the infirmary are followed by a nurse
practitioner. Very few of these patients are ever physically examined by a doctor.
Almost no one gets a hemoglobin Ale, a blood test that is commonly used to monitor
long term control of diabetics. A surgeon who provides coverage is the responsible
physician for the unit, even though most ofthe patients have Internal Medicine problems.
Even if a physician were available to see patients, access to those patients is extremely

41

]
)
J

I
I
I
I
I
I
I
I
I

poor. During daytime hours, three officers are assigned to the infirmary. One is engaged
in transporting inmates, one is assigned to the officer station and only one is available for
escorting inmates to the examination room. Effectively, this prevents anything but cell
side visual inspections through the glass wall for most patients. Clinical staff are not
permitted access to inmates except with an officer escort. Because there are so few
officers, very few individuals are examined or evaluated. During evening shifts there are
fewer officers and the situation is worse. Thus, the housing unit into which the most
seriously ill patients are sent has barriers which prevent examinations, evaluations and
treatments resulting in poor outcomes. Chart reviews demonstrate the lack of physical
examinations even in seriously ill patients.
The nurses that I interviewed verified problems with getting patients seen on the
infirmary unit. There are occasionally problems with even getting inmates to the nursing
station. On some days, no inmates are brought out for vital signs. For example, nurses
monitor persons with potential for withdrawal with vital signs at regular intervals, but
nurses often have trouble getting to these patients. The mix of staff on this unit is also
not appropriate for an infirmary unit with large numbers of patients with significant
chronic illnesses. UTMB is actually redesigning the staffing ratios to less well trained
staff (patient care attendants and assistants instead of nurses). Some assistant staff are
useful and cost efficient, but the proportionate decrease in nurses capable of performing
assessments on a unit of this type is not prudent from a clinical perspective.

DELIVERABLE #10: Assess the potential for reduction of transfers from Dallas
County jails and/or other Dallas County detention centers to Parklandfor tests, specialty
care, inpatient admissions, etc., review a sampling of medical records for adherence to
AHRQ national quality indicators and review aggregate utilization data for clinic,
emergency, diagnostic, and inpatient transfers.
Utilization review of off-site medical consultation and hospitalization, if done at all, does
not result in any aggregate reports or conclusions. UTMB does not maintain a utilization
log of transfers to Parkland. This information is only available from the correctional
officers transport log. Information is also available from Parkland Hospital on the
number of hospital admissions, including the number of days for each admission and a
diagnosis. Also available was the Parkland Hospital Ambulatory Services Case
Management Program report detailing the number of scheduled and actuaf visits of
inmates to multiple clinics and specialty testing services. Several conclusions can be
drawn from this limited data.
For a one-year period (December 1, 2003 to November 30, 2004), there were 309
hospital admissions (25.75 per month) for a total of 2,182 hospital days for the year, or
181 days per month. The average length of stay is 7.06 days per patient. Of these 309
hospital admissions, 37 exceeded 15 days length of stay and 8 exceeded 30 days length of
stay. This is a long length of stay for many admissions and reflects persons being
admitted with late stage of disease necessitating longer hospital stays. This information
is consistent with findings on chart reviews in which inmates did not receive timely care

42

J
]
]

]

]
]
J
J

I
I
I
I
I

and ultimately required hospitalization with more serious disease necessitating longer
hospitalization.
The Agency for Healthcare Research and Quality (AHRQ) is a Federal agency that has
developed a set of measures that can be used with hospital inpatient diagnosis data to
identify potential preventable hospitalizations and thereby assess the quality of
ambulatory care. Using these indicators with the addition of 3 other diseases that are
usually preventable (fluid overload with end-stage renal disease, encephalopathy, and
cellulitis) some potentially preventable hospital days are identified. Cellulitis, in
particular, is a problem condition at the jail and may indicate hygiene conditions as well
as lack of early referral for evaluation of minor infections. Using these diagnoses as
indicators of ambulatory care at the jail is useful as a way to improve the quality of care
at the jail as well as reduce unnecessary hospitalization. These potentially preventable
days amount to approximately 22% of hospital days and indicate that there is
considerable room to improve ambulatory care.

DIAGNOSIS
CELLULITIS
ASTHMA
CHF
HYPERTENSION
GANGRENE
SEPSIS
OVERLOAD
FLUID
ON DIALYSIS
ENCEPHALOPATHY
KETOACIDOSIS
TOTAL

ADMISSIONS
40
11
9
4
3
3
2

HOSPITAL DAYS
262
86
34
26
II
48
5

2
1
75

5
486

9

Of some concern is the large number of cases exceeding fifteen days. These should be
reviewed to assess whether problems at the jail are deteriorating to the extent that inmates
are going to the hospital at a later stage of disease or whether the ability of the jail to
accept inmates back on the infirmary is adequate. Several charts that I reviewed
documented care given to patients who were re-admitted multiple times for the same
condition, indicating a failure to follow-up with an adequate management strategy,
necessitating repeated and unnecessary admissions.
These types of preventable
admissions' indicate poor quality of care, but also result in excess cost for the
hospitalization as well as for the security staff that must guard the inmate while
hospitalized.
Despite the opportunities that exist to review existing data to try identifying unusual
trends, reducing hospital stays and improving quality, there is no data that is used by
medical staff to review, in aggregate, the types of clinical cases that are being transported
to Parkland Hospital for hospitalization or specialty services. Parkland Hospital

43

maintains this data, but, to the best of my knowledge, there is no communication between
Parkland and jail staff to discuss utilization or any other issues.
Off-site appointments for specialty services mostly occur at Parkland Hospital clinics.
Typically, this type of data would be reviewed to identify trends in specialty care that
may indicate system problems with health care delivery, or to identify trends in specialty
care so that specialized care might be brought on-site. [n addition, review of these data
assists in identifying ways to reduce transportation to local area hospitals and specialty
clinics by improving the array of services onsite. Information from Parkland Hospital
shows that, for a year's worth of specialty appointments, inmates show up only 66% of
the time. Whether the inmate was still incarcerated is not identified. This type of data
could be used in a combined Parkland/Jail Quality Improvement manner to improve
overall show rates and identify why inmates are not arriving for their appointments.
Telemedicine equipment does exist at the facility and could be more fully used to reduce
hospital trips. Telemedicine was used with Parkland previously for orthopedics,
emergencies, and for HIV care. The extent of that use was not clear. Now no
telemedicine clinics are done.
In order to get a better sense of what problems existed with ambulatory care at the jail, I
reviewed several charts of persons who had admissions to the hospital that had a higher
likelihood of being preventable. Because there was no clinical data available on
hospitalizations from UTMB staff, the officer's transportation log was used to identify
cases. The cases chosen were randomly chosen from only two months of recent hospital
admissions. These reviews demonstrate that there are significant jail ambulatory care
management problems and many preventable hospital days as well as clinical quality
Issues.
One patient didn't have his chronic illnesses identified for over 6 weeks after intake
despite having many abnormal symptoms, vital signs, and clinical manifestations. After
he finally was clinically examined, multiple laboratory tests were ordered on two
different occasions several days apart, none of which were evaluated. For one of the
laboratory results, the chart documents that the laboratory had called to notify the jail that
the laboratory result was a critical value. Still the patient had no physical examination
follow up or follow up of his laboratory tests. Several days later the inmate was unable to
breathe, was admitted urgently to Parkland and died.
.
Another patient had several problems. One of his problems was hyperthyroidism. This
condition causes a fast pulse when it is inadequately controlled. This patient had a fast
pulse for over 9 months and his hyperthyroidism was basically not controlled or even
appropriately evaluated for almost a year. During this same time period, he developed a
leg infection that was complicated because he had diabetes. The infection was so poorly
managed that it necessitated 10 separate hospital admissions for 53 hospital days over a 6
month period. There were periods during this time when he was not even examined
while at the jail. This form of systemic incompetence was an extremely costly series of
preventable hospitalizations in addition to the clinical consequences to the patient.

44

Another patient had a history of severe asthma for which he had been intubated five times
in the past when he was a civilian. This indicates a very high-risk patient. This part of
his history was not picked up at intake screening. Medication he said he customarily took
was not provided to him for several days. When he did receive medication he only
received some of his medications. Over a three month period he had six separate hospital
admissions totaling 41 hospital days along with five other emergency room visits
necessary to manage his out-of-control asthma. Despite these repeated life-threatening
asthmatic emergencies, he was examined at the jail by a physician or physician assistant
only twice over a three month period. Peak flow monitoring, which is the standard
method of monitoring patients with asthma was not performed at the jail except once.
That single episode occurred when Parkland Hospital sent a peak flow monitor back with
the patient along with his medications. In fact, the jail does have peak flow monitors but
never uses them in the evaluation of asthmatics. This costly and potentially life
threatening series of episodes was probably entirely preventable.
Yet another patient had a history of hypertension, prior tuberculosis and heart disease
taken at intake. The intake screening missed the fact that he had advanced cirrhosis and
probable mental status alterations. His care was provided by physician assistants (one a
medical physician assistant and another psychiatric physician assistant). The patient was
examined initially by one of the physician assistants and given medication appropriate for
someone with cirrhosis of the liver but the examination was not thorough and the
documentation in the record did not have sufficient information to determine what
exactly was wrong with the patient and why the physician assistant had treated the patient
in the manner that was done. Shortly after that, the second physician assistant (the
psychiatric one) diagnosed organic brain syndrome and prescribed haloperidol a mtUor
psychotropic medication that has only marginal indication for this condition. In addition,
because the patient had cirrhosis, the haloperidol was contraindicated. The patient wasn't
evaluated again for about 3 months when he had collapsed. Initially, the nurse called a
mid-level provider who recommended that the patient be monitored. What this meant
was not clear from the documentation. He was sent to the infirmary. There is nothing in
the medical records that documents monitoring of any kind. The next day the patient was
admitted to Parkland Hospital. There was no examination or note in the record to
describe what had happened. He returned from Parkland about 2 weeks later and the
notes by jail physicians do not clearly describe what his medical condition is or what had
occurred at Parkland or what had occasioned his admission. One note indicated that the
medical staff were trying to get him into a nursing home. He had yet to receive a
physical examination (almost two months) since discharge from the hospital at the time I
reviewed the medical record.
Several conclusions can be drawn from these examples.
• There are many preventable hospitalization days that are both costly and result in
liability exposure to the County and UTMB.
• The lack of competent physician intervention in the care of seriously ill patients is
extreme.
• Patients seldom are physically examined when necessary.

45

•
•

•

Patients returning from the hospital are never examined upon return in a timely
manner, if at all.
Providers do not appear to review past medical records as they evaluate patients.
If providers are only focused on what they may be doing or what they have done,
they will miss important information documented by other providers. Seldom
does a patient see the same provider. In this sense, the electronic record appears
to actually be a barrier to continuity of care.
The quality of medical care is extremely incompetent.

Contract monitoring by Parkland Hospital is effective for several reasons. UTMB does
not fully disclose information that is necessary to evaluate program success. Financial
data is not provided by UTMB to Parkland. Performance objectives are described so
generically as to be not useful to perform program evaluation. In addition, the system of
care is not accurately and honestly addressed in these evaluations. For example, "Intake
Mental Health Referrals" is a performance objective and is listed as "target met." This
indicator is meant to measure whether any inmate needing psychiatric care is examined
within 7 days for routine care. However, because "many inmates are released before
mental health staff can access them," the review considered that the target was met. This
assessment, however, ignores the fact that the kite system and referral system is so
broken that few if any patients are seen via the referral process. To consider that
performance is adequate because patients are discharged before being seen is not
accurate.
Similarly, the "Intake Screening" performance objective is meant to measure whether
inmates referred to health services during intake screening receive a central intake
evaluation. This received a "target met" designation implying that persons with medical
conditions are evaluated via the intake process. Virtually none of the patients with
chronic illness received an adequate evaluation as evidenced in chart review. The
indicators miss the point. There has been an absence of review of critical cases in order
to identify problem areas.

46

· RECOMMENDATIONS FOR IMPROVING MEDICAL
AND MENTAL HEALTH SERVICES AT
DALLAS COUNTY JAIL

47

1. Medical autonomy must be established at the jail over medical programs meaning
that access to all medical and mental health units should be under control of
clinical staff and clinical staff should be the final decision makers on any clinical
decision. Inmates with severe illness, particularly mental illness, should not be
moved at the discretion of classification staff. This is best accomplished by
establishing specific units for crisis management patients as well as an
intermediate care mental health unit.
2. Monthly meetings that are formal and have minutes should exist between custody
staff and medical staff to identify problem areas of concern. These problem areas
should be studied and improvements attempted through a quality improvement
process.
3. A set of policies and procedures that realistically and practically describe the
practices at the jail should be developed, trained against and established at the jail.
These policies should be produced by medical leadership staff in collaboration
with other key staff members. For some policy, the correctional leadership will
have to participate and agree to the policy (e.g. intake screening, transportation,
etc.). These policies should include the Sheriffs (or designee) signature as
acknowledgement of acceptability. Unless the Sheriffs staff support these
policies they will not work.

4. If the current arrangement with a medical vendor continues (and that should be
thoroughly assessed), Parkland Hospital should create an audit system that very
specifically audits against contract expectations including clinical indicators for
care. These audits should be performed at specific intervals and would be a major
mechanism of contract monitoring. Monthly reports of key service areas (number
of inmate requests, number of requests seen, skin tests applied, etc.) should be
reported from the vendor to the contract monitor. The monitor should determine
what the scope of these reportable items.
5. All staff assigned to this contract should be fully engaged in providing services at
the Dallas County Jail and the services delivered should be clearly delineated in
the contract budget. A staffing plan, including credentialing requirements, should
be agreed to by the County. A system of quarterly adjustments should reconcile
approved over use of staff or under use of staff and would be reflected in
subsequent payments.
6. Medical staff should perform intake screening.
7. Medical screening should be conducted with aural privacy. Examinations should
also be conducted with visual privacy.
8. Medical screening should result in continuation of medication for inmates within
24 hours. For persons on insulin, therapy must be continued immediately.

48

9. Screening evaluations should all be included into the medical record, whether
they are performed by correctional staff or medical staff.
'10. An automatic system of referral should occur for any inmate with a mental health
or medical condition identified at intake that results in a physical examination by
a clinician appropriate for the level of illness. For mental health evaluations, this
will mean bachelor level or higher persons screen and refer complicated to a
psychologist who can complete the evaluation for complicated patients. For
medical patients a physician or physician assistant should evaluate and examine
patients with chronic disease early (1-3 days of booking) depending on the level
of acuity of their illness.
II. A crisis stabilization unit should be established for housing suicidal and severely
disturbed mental health patients that creates a permanent and fixed housing
arrangement for inmates until such time that they are discharged from the unit by
a clinician authorized to do so. Admission to this unit should be by a licensed
psychologist or a psychiatrist. Correctional staff must oversee security so that the
unit is secure but within those security restrictions, mental health leadership
should have authority and autonomy to establish a unit with a therapeutic milieu
so that inmates can be housed in a therapeutic manner. This unit should include
daily rounds by a psychiatrist and or psychologist and be monitored by mental
health staff on a continual basis during day and evening hours.
12. An intermediate mental health unit should be established in order to house
inmates with severe mental illness who do not rise to the level of the acute
stabilization unit. The oversight and rules on this unit should be similar to the
acute unit except for the indications for admission.
13. A correctional infirmary should be established that includes clinical autonomy
similar to the mental health units described above. This unit would include
unimpeded access to examination of inmates, daily rounds by physicians and
nurses with documented notes, and a level of staffing commensurate with the
desired goals of the jaiL A Board Certified Internist should manage this unit.
14. Sufficient clinical space should be identified so that there are sufficient clinical
examination rooms to examine patients.
15. Establishment ofa chronic disease program that include Board Certified or
Eligible Internist or Family Practitioners who manage patients occasionally
transferring care of less complicated patients to mid-level providers who assist
them. This type of program is best performed as a team with physicians, midlevel providers, nurses and clerical staff working as a team. This program should
include chronic care guidelines specific for the commonest illness at the facility
and the ability to track, and examine inmates on a regular basis or as often as
needed in order to manage their chronic illness.

49

16. All sick call requests should be scanned or otherwise incorporated into the
medical record.
17. The system of sick call should be staffed so that routine clinics can be timely
accomplished. Mid-level providers or physicians should see patients and examine
them in clinical settings appropriate for a clinical examination. Statistics should
reflect the number of requests submitted with the date, the date of first evaluation
and the date of follow up referral examination.
18. A system of confidential transfer of inmate health requests from inmates to health
care staff should be instituted. This is most often done by installing locked boxes
in which inmates may drop their health requests. Health care staff pick these up.
This eliminates custody staff from any involvement in the process except to give
inmates blank forms and writing instruments if they do not have them.
19. A system should be established so that medical patients on mental health units are
seen and examined and mental health patients on medical units are seen and
examined.
20. Symptom screening for tuberculosis should be instituted.
21. The screening for tuberculosis with skin testing should be improved so that it is
effective or consideration should be given to instituting chest x-ray screening of
all incoming inmates.
22. TB and HIV management should be under the direction of a physician
knowledgeable and experienced in the treatment of these diseases.
23. An infection control program should be established that tracks, contagious and
infectious diseases, reports appropriately to the Department of Health, ensures
reasonable contact tracing via an exposure control plan within the facility, and
establishes appropriate blood borne pathogen and other infection control policies
and procedures. This program should use the expertise of Parkland Hospital,
UTSW and the Dallas County Department of Health and should develop linkages
to these organizations in the follow up care of their patients.
24. Negative pressure isolation rooms should be inspected by a someone certified to
do so on a regular basis and maintained so that true negative pressure is assured.
25. Management of methicillin resistant staph aureus (MRSA) should be reviewed
and improved. Consultation should be obtained in this area. Significant
improvement of the hygiene of the facility will be necessary to thwart this
important and emerging pathogen.

50

26. Strong consideration should be given to screening selected persons for sexually
transmitted diseases in conjunction with the intake process. This should be
developed in coordination with the Department of Health and, if they are willing,
the UTS W/Parkland.
27. The process of nurses and liaisons providing front line evaluations of inmates is
burdensome due to the lack of correctional staff to move patients for
appointments. It should be established whether increased officer escort staff will
be available. After that is done, an assessment of nurse and liason staffing should
be done so that there is sufficient staff to evaluate inmate health requests and
conduct preliminary evaluations. Obviously, the system that is mutually agreed
upon by corrections and medical and mental health staff will have a bearing on
staffing considerations.
28. Problems with data entry and management in the electronic record should be
identified and studied through a quality management process. All patients should
be seen with a medical record. Because the record is electronic should not
prohibit this from occurring. Policy should be developed that delineates how
clinicians shall use the medical record.
29. The medical vendor should develop a written strategy (codified in policy) for the
contractor (County) specifying what the operational procedure will be in the event
the electronic medical record becomes disabled. There should also be defined in
writing an exit strategy that defines the condition and format of records upon
termination of the contract with the existing vendor or in the event the software
becomes unsupported.
30. Consider utilizing the MetroCare organization in collaboration with the
Department of Psychiatry at UTSW to assist in organizing and staffing the mental
health programs at the Dallas County Jail.
31. Pharmacy services should be evaluated. All prescribed medications should be
delivered within one to two days. Barriers to this should be identified and
corrected.
32. Utilization should be monitored in collaboration with Parkland Hospital. Patterns
of hospital admissions should be studied for trends in order to make
improvements that reduce overall hospitalization. Specialty service trends should
be analyzed in order to identify frequently used services so that those services can
be brought onsite, if possible, in order to reduce unnecessary transportation of
inmates. In this context, telemedicine is underutilized.
33. Mortality review should be instituted for all deaths. Under the current
circumstances, this is best done by outside experts. This is probably best
performed by Parkland Hospital as they are responsible for monitoring the

51

contract and could bring expertise to the evaluation. This should be done in the
format of a mortality review committee to which jail staff participate.
34. The jail should consider accreditation by the National Commission on
Correctional Health Care as this accreditation would help to structure reforms.
35. All inmates transferring from another facility or returning from hospitalization or
off-site specialty appointment should be examined by a physician with a day to
ensure that prescriptions and follow up care is initiated.
36. The mission of the dental program should be reviewed and augmented. If
available, a licensed dentist from Parkland or the medical school should review
the dental equipment and dental operatory and advise the jail on how to improve
the existing dental facility. Probably this unit should be rebuilt with adequate
equipment.
37. A request should be made so that inmates with mental illness can initiate
enrollment into NorthSTAR upon incarceration.
38. Dallas County Jail mental health leadership should continue to be involved with
the County in assuring the success of the mental health diversion program.

52