A Method For Tracking Severely Emotionally Disturbed Children And Adolescents Through A State Mental Health System

High Plains Applied Anthropologist No. 2, Vol. 21, Fall, 2001
166
A Method for Tracking Severely Emotionally Disturbed Children and Adolescents
Through a State Mental Health System
Michael H. Moynihan, Peter W. Van Arsdale, George Kerin, Eric Curton, and William
Madura1
Introduction
Cross-cultural issues affecting children and
adolescents are gaining increased mention within the
United States, in part because of tragic incidents such
as those at Columbine High School in 1999. These
issues also are raising concerns in part because of
demands being placed on the human services system as
mental health, social service, and youth corrections
agencies vie for scarce state resources.
Intersecting these issues, which often are tracked
“in the aggregate,” are those which involve individual
youngsters, many of whom “don’t fit” within readily
definable service categories. Social anthropologists,
sociologists, and practitioners have been concerned
with ways to better track what happens to these
youngsters as they move – seemingly in fits, starts,
waits, and stops – through the maze of public and
private service providers. The case of “Hannah’s Dark
Journey” (Imse 2000:7), in Colorado, certainly
exemplifies this for us. Through no fault of her own,
her parents divorced, her favorite grandfather died, she
and her mother became homeless, she was sexually
assaulted, and a judge took Hannah away from her
mother. For the next ten years, Hannah was placed in
21 different foster homes, treatment centers, mental
hospitals, and juvenile jails at a cost to the taxpayers of
almost $300,000. While Hannah has finally received
intensive mental health care, her odyssey is far from
unusual. In fact, it is because of cases very similar to
Hannah’s that prompted the authors of the present study
to devise a method of graphical tracking of children as
they move through the child service system. We
believe that this method will be of interest to field
researchers working in other institutional settings as
well.
Design
The overall study design was to develop a pilot
study as a pre- post longitudinal study with repeated
measures. Starting April 15, 1994, the authors assigned
the first three children admitted to the five respective
child and adolescent treatment teams at Colorado
Mental Health Institute at Fort Logan to the pilot study
after the children and their parents agreed to participate
and signed informed consent forms. Fort Logan is one
of two public mental health hospitals in Colorado and
during the last three years it has averaged
approximately 1000 inpatient admissions annually. The
pilot study did not involve the random assignment of
children to treatment conditions, except in the sense
that the first three children assigned to the five
treatment teams during the study period became part of
the study. Since the overall number of subjects was
small and their outcome data cannot be used to
determine that one treatment is better than another, the
main purposes of the pilot study were:
• to test the usefulness of the various
research instruments;
• to observe possible trends in client
tracking and outcomes, including
financial cost analyses, suggestive for
further study and;
• to provide working time frames for
project staff prior to engaging in a
proposed larger, more comprehensive
study involving 100 to 200 children.
Moreover, by developing a pre-Fort Logan archival
profile of each child’s psycho/social history, the authors
were able to use the analytic framework presented in
the Vermont System for Tracking Client Progress
(Buchard, Gendebien, and Hinden, et al. 1992), as well
as a related technique recently discussed by Van
Arsdale (1996). This provides both retrospective and
prospective analysis of certain elements of the child’s
treatment outcomes – analyses that will be useful in the
larger study of children following psychiatric
hospitalization.
Method
As noted above, the overall goal of this research
project was to develop a method to track the movement
of children throughout the mental health system. The

High Plains Applied Anthropologist No. 2, Vol. 21, Fall, 2001 167
authors assumed this could be done in a linear, point-to-
point manner. However, collecting historical data on
mental health treatment and placements of the children
proved to be very difficult. Indeed, when children
arrived at Fort Logan for psychiatric treatment, their
case files/histories were usually incomplete or non-
existent. Some children’s case histories were
incomplete six weeks after they had arrived. Even with
all appropriate medical and legal clearances and signed
consent forms, researchers often met resistance from
previous mental health providers regarding whether
they would release a given child’s case files. Not only
did this resistance hamper the data collection process,
more importantly, it also limited useful information
necessary for the child’s treatment. While we were able
to complete the retrospective profiles of each of the
fifteen children we studied, many of the other
children’s case histories we looked at had gaps ranging
from minor to major.
Despite these difficulties, one factor allowed us to
develop a linear time frame to track the children
through the mental health system. The method is
simple in concept and design, but difficult in
implementation and analyses. In essence, the method
is Follow the Money. The method does not require
judgement whether the mental health monies should
have been spent, if too much or too little was spent, or
who made the payments. Our focus was to simply
follow the money trail. Even with this fairly
straightforward system, some difficulties were
encountered. Different systems and accountability
within systems often obscured the trail. Also, during
the course of the study, a major shift in Medicaid
funding occurred in Colorado.
Data Collection
Prior to designing any of the data collection
instruments, the authors of the present study analyzed
the Fort Logan Medical Record forms (approximately
200 pages of admission records, assessments, treatment
plans, consultations, progress notes, physicians’ orders,
prior history, therapies, legal status, correspondence,
financial status, etc.) for content. From this, we
developed a data collection guide to aid researchers in
finding information in the Medical Record forms.
Following the Vermont system for Tracking Client
Progress (Bruchard, Gendebien, and Hinden, et al.
1992) and expanding the content areas to include more
detail, the authors completed a demographic worksheet
on each child. In addition to basic demographics, this
residential and service worksheet includes: prior
residential services and restrictiveness of living scale
(Hawkins, Almeida, and Fabry, et al. 1991); the child’s
age at time of service; the number of times the service
was rendered; length of stay in any of the residential
services; the cost per day for the specific service; major
life events; prior child/family services; educational
services; referral source; and psychiatric and/or
delinquent characteristics. To establish baseline
comparisons, the authors collected and charted these
data from the Medical Records for one year prior to
each child’s admission to Fort Logan.
The expansion of the Vermont System for Tracking
Client Progress complement, the methods of event
history analysis. For example, Barton, Pillai, and Dietz
(1992) note that in the field of program evaluation as
applied to human services, researchers must distinguish
among “events,” “spells,” and “states.” That is, an
event is a dramatic, qualitative occurrence that results
in a change of circumstance (e.g., admission to a
hospital). A spell is the duration of time between
events (e.g., post-admission to discharge). A state is
one’s condition (e.g., being hospitalized). Events
predominate (including referrals) in the present study’s
research design in large part because of the data
available and the investigative methods used.
Following Barton et al. (1996), the authors of the
present study are able to infer a “risk set;” the group of
youngsters who are at risk of experiencing the types of
events in question.
Upon admission to Fort Logan, the Child Behavior
Checklist (CBCL), Caretaker version (Achenbach
1991) was administered for each of the 15 children
from the child’s most recent residential setting. Other
psychological and educational test results were
collected from referral sources, which included “Level
of Functioning Scales,” then known within Colorado as
PES 7A & B and PES 10. Data were also collected on
medications at admission, medications at discharge,
prior treatment history, treatment plans, discharge data,
and diagnosis at discharge. Additionally, the authors
designed a pilot study measurement schedule to detail
when specific instruments (e.g., PES-7, Child Behavior
Checklist, Piers-Harris Self-Concept Scale [Piers 1984],
demographic profile, Alliance Test, and the PES-10)
were to be administered and by whom, while the child
was in Fort Logan, at discharge, and at three-month
intervals following discharge.
The authors developed a post-discharge outcome
indicator checklist to measure the impact of treatment
and the child’s ability to reintegrate into the
community. The researchers were to complete the
checklist at three, six, nine, and twelve months

168
following discharge, a task not fully completed owing
to subsequent resource constraints. This checklist
included such items as the number of out-of-home
placements, length of stay in a variety of inpatient and
outpatient settings, psychological test scores, police
contacts, and other treatment outcomes treatment staff
determined to be important for the specific child.
To work as field researchers, student interns were
recruited from the University of Colorado at Boulder
and, later, from Denver University. These students
collected the majority of information from the Medical
Records files, worked with the treatment staff at Fort
Logan, and, when appropriate to the measurement
schedule, interviewed and tested the children on the
various psychological instruments. Treatment staff at
Fort Logan were responsible for conducting some of the
psychological tests. In addition to signing agreements
of confidentiality each student researcher completed
detailed time analyses of data collection and data
analysis on a child-by-child basis. Initially, three
student researchers from the University of Colorado
worked with four children each and one worked with
three children to complete the final data collection
phase. During the time period of the study (April 15,
1994 to May 15,1996), the four student researchers
from the University of Colorado logged 310 hours in
data collection and analysis, and the five student
researchers from Denver University logged 195 hours
in data collection and analysis. In all, the student
researchers spent 505 hours collecting data, with an
average of 34 hours per child.
Financial cost analysis came to play a central role in
the pilot study. For the most part, the analyses utilized
audited unit cost data rather than data on “expenses per
person.” An economic analysis was not conducted due
to the large number of estimates and assumptions that
would be required. For medications and certain other
cost categories, expense data were utilized (i.e., that
which was actually paid out) rather than expenditure
data (i.e., that which refers to obligations incurred,
whether paid or not). Charge-based data were used in
some instances since it approximates costs.
Government benefit data were not collected, and thus
only factors in indirectly, as with the hospitalization
costs covered by Medicaid. Virtually all unit cost data
were current as of fiscal year 1995-96; thus cost
adjustments in accordance with Consumer Price Index
(CPI) were not necessary.
Population
Relatively precise numbers regarding those children
and adolescents in need of mental health services in
Colorado are unknown. As of 1990, the last year for
which complete state-wide census data are available,
Colorado Mental Health Services estimated 5 percent of
those children aged 11 years or younger, or
approximately 35,000 children in the state, had the
most serious mental health problems and were in need
of mental health intervention (Heller and Coen 1996:2).
Only about 21 percent of these – some 7,300 children
(including repeat admissions) – are served by the public
mental health system during the course of any year
(Heller and Coen 1996). From vital statistics available
to our research team, the authors estimate in Colorado
that approximately 30,000 to 40,000 adolescents aged
12 through 17 are in need of mental health services.
About 10,000 to 12,500 of these adolescents are served
by the public mental health system during the course of
any year. Through age 17, most of these children come
from families living below the poverty level.
System of Services
Colorado has four major service systems for at-risk,
troubled youngsters: the mental health system, the child
welfare system, the youth corrections system, and the
private system. Discussion here focuses on the three
public systems. Prior to their admission into Fort
Logan, each of the 15 children involved in this pilot
study were processed through a mental health center.
The mental health centers are responsible for following
their respective children while they are in the hospital,
as well as being responsible for the children’s mental
health services after hospitalization. The social
services system has the financial responsibility for the
children’s out-of-home placements. The youth
corrections system has the responsibility for the child if
he or she has been adjudicated of an offense,
temporarily placed in a detention center, or committed
for a long-term period in a youth corrections institution.
With the advent of managed care in August 1, 1995,
Medicaid (the federal government’s health-care
program for the poor) became capitated and mental
health centers could set limits on mental health services
to children and adolescents. At this time, fifty-one of
Colorado’s sixty-three counties were included in a
Medicaid capitation pilot program, which included
approximately 69 percent of the Medicaid population.
Areas initially not included in this pilot program were
Denver and Larimer Counties, and ten counties in
northeast Colorado. Before the system became

capitated, the placement of children outside of the home
was the responsibility of the child welfare/social
services system, including paying for the out-of-home
placement costs. After the system became capitated,
children who were placed out of the home continued to
be the major responsibility of the social service system,
but the capitated mental health centers could use their
Medicaid funds for supplemental services after the
child was hospitalized. All 15 children included in this
pilot study were admitted prior to capitation. By
August 1, 1995, the mental health centers included in
the Medicaid capitation pilot program that referred
children to Fort Logan came under the capitated
system, and thus, they became responsible for paying
for their children’s hospitalization at Fort Logan. Many
of the capitated mental health centers, when notified
that they had received their managed care awards,
began taking their children out of Fort Logan’s
treatment programs as early as June 1995.
Literature Review and Research Rationale
Intensive inpatient psychiatric care for children and
adolescents has been utilized in the United States since
the 1920s. However, few rigorous follow-up studies
were attempted prior to 1958 (Gossett, Lewis, and
Barnhart 1983). The first such comprehensive study
was published by Masterson (Gossett, Lewis, and
Barnhart 1983) in that year. As Gossett, Lewis, and
Barnhart (1983:15) note, Masterson found that
“neuropathic” traits (like tantrums) in the early lives of
schizophrenic patients were not predictive of eventual
outcome. During the next few years, other studies
followed such as the Timberlawn Adolescent Treatment
Assessment Project (Grossett, Lewis, and Barnhart
1983:15) where post-discharge indicators were charted.
In this study, favorable outcomes were found in the
“Level of Functioning” analysis for over two-thirds of
those studied.
In 1986 the National Mental Health Association
(NMHA) clarified the need for coordination of services,
early identification and assessment of at-risk children,
delivery of appropriate community-based services, and
comprehensive research and evaluation. As of 1988,
some 20,000 children were being served in in-state
psychiatric hospitals (Gendebien, Hinden, and Pandina,
et al. 1992). Moreover, also during the 1980s,
researchers associated with the Colorado Division of
Mental Health expanded their investigation of treatment
outcome assessment (Ellis, Wilson, and Foster 1985).
By the mid-1990s, synthetic studies summarizing what
has been learned were appearing regularly (Kutash and
Rivera 1996).
Rationale
No longitudinal studies have been conducted at
either of Colorado’s two public hospitals – and few
anywhere – which have systematically tracked and
evaluated children and adolescents from pre-admission,
through the duration of hospitalization, and through the
first several months or more post-discharge. The 1998-
2000 Residential Treatment Study at Fort Logan is now
doing this. Useful anecdotal information has been
collected, as has some information comparing different
groups pre- and post-, but in Colorado no study
systematically evaluating a single large group has been
completed.
It is essential to understand predictive behaviors of
those studied. This ideally can lead to the development
and identification of indicators. Both “adjustive” and
“predictive” indicators are important. Since several
forms of therapy (i.e., interventions) are used at Fort
Logan and in other related Colorado programs, it is
important to understand the extent to which
interventions correlate with post-discharge outcomes.
This also would enable a systematic understanding of
failures (e.g., re-admission, suicide). At the broadest
level this also would allow the identification of typical
flow patterns from intake to eventual referral and post-
hospitalization placement (Forward, Moynihan, and
Wilson 1990).
Throughout this pilot study, one of the greatest
challenges was comparative. Curry (Goodrich
1994:280-281) writes that in studies of this sort,
“…because of the lack of a [formal] comparison group,
within-program and between-subject designs are
inherently incapable of addressing the question of the
effectiveness of a specific program treatment.”
Furthermore, Curry (1991) notes that, in adolescent
inpatient settings, the adjustment of the patient within
the program does not predict adjustment during follow-
up, but the degree of support and continuity of key
relationships following hospitalization does. According
to Grob and Singer (1974), another major challenge is
the validity of ratings of patient adjustment and
progress. Objective, operationalized research and
terminological definitions must be used to aid the rating
process. The targeted, tested, and diverse types of
instrumentation that would have allowed full-scale
comparisons and ratings originally proposed for this
pilot study were not utilized here.

170
Key Intervention and Outcome Studies
While there is no clear-cut correlation between type
of intervention and type of post-discharge outcome, the
recent study of adults by Chandler, Meisel, and Hu, et
al. (1995 and 1996) is of special importance. In a three-
year controlled study in California, these researchers
gathered data which compared two integrated service
agency demonstration programs, one urban and one
rural. Capitation was instituted and its goal was to
determine if the two integrated programs compared to
the programs that provided the usual services produced
improved outcomes for clients with severe and
persistent mental illness. In contrast to those clients
(randomly assigned) who received the usual services,
those served by the integrated programs were found to
have less hospital care, greater workforce participation,
fewer group and institutional housing arrangements,
and greater client and family satisfaction, among other
factors. Of particular interest in the present context, the
average mental health service cost per client was
substantially higher for demonstration clients than for
the comparison clients.
By contrast, an early research study by Grob and
Singer (1974) involved patients admitted to the McLean
Hospital near Boston in the early 1960s. The 36 males
and 31 females in this study were aged 13 to 19 and had
been in residence a minimum of three months. Among
the most important findings were that relatively few
pre-admission variables and in-hospital variables –
taken singularly – were related to outcome. That is,
they had little predictive value. However, improvement
at discharge (as defined by mental status ratings) turned
out to be highly predictive of outcome at follow-up.
Improvement begun during hospitalization was inferred
to be adaptive after discharge.
Of special interest was the fact that a follow-up study at
McLean, although involving different adolescents, was
conducted almost a decade later. As of discharge, most
characteristics were the same for the two samples.
From this sample of 65 patients it was found, for
example, that more patients went home after discharge
than in the earlier sample.
Outcome Ecology
Because of its singular importance to the Fort Logan
Pilot study, the current work of Hernandez and Hodges
(1996) is reviewed separately here. The goal of their
System Accountability Project for Children’s Mental
Health at the University of South Florida is “…to better
understand the impact that utilizing measurable
outcomes such as school attendance, juvenile justice
recidivism, and out-of-home placements [emphasis
added] has on the planning and delivery of children’s
mental health services” (Hernandez and Hodges
1996:iii). To this end, Hernandez and Hodges (1996)
recently began a “best practices” investigation tied to
“outcome accountability” in which the service system
and accountability system overlap. Serving as a
foundation to their approach is the premise that service
delivery and outcomes attained are contextual. This
approach emphasizes stakeholder involvement in the
identification of outcomes to be measured, the context
(including cultural) within which information is
developed and delivered, and the transformation of
information – by managers – into appropriate clinical
decisions.
Key Costing Studies
In addition to the study by Chandler et al. (1995)
cited earlier, within the context of community mental
health care, the study by Wolff, Helminiak and
Diamond (1995) in Wisconsin expands upon the
concept of “societal costs of care.” As these costs are
translated into dollars and cents, the authors warn of
errors that might arise in calculating – and inferring
from – categories variously labeled as “charges,”
“payments,” and “accounting costs.” Similarly, the
authors of the present study also found it difficult to
gather cross-system data that would provide cost
analyses of mental health clients within the law
enforcement and correctional systems. In Colorado, of
the approximately 800 adults in a major treatment
outcome study begun during the late 1980s, Shern,
Coen, and Bradley, et al. (1990:75) and Coen, Shockly,
and Bradley (1996) selected 198 for inclusion in a
special costing study. For slightly over half the
participants, schizophrenia was the primary diagnoses
and for the rest of the individuals, affective disorders
were the primary diagnoses. The authors (Shern, Coen,
and Bradley et al. (1990) and Coen, Shockly, and
Bradley (1996) derived cost estimates for seven broad
areas: Mental health, drug and alcohol, law
enforcement, courts, casework services, physical health,
and government benefits. (By comparison, only the
first five are included in the present study). The authors
found that mental health costs comprised about two-
thirds of the total costs and, in turn, inpatient costs
comprised about two-thirds of the mental health costs.
While not identical methodologically, the present
study’s cost analyses were similar to those employed in
the studies by Shern et al. (1990) and Coen et al.
(1996). Like them, and tied to the use of audited unit
cost data, the authors of the present study used units of

service that were “inherently logical and specifically
tied to costs or charges” (Coen, Shockly, and Bradley
1996). Services within hospital or residential settings
were counted as “days” or “months” and those that
were encounter-based were counted as “contacts,”
while others were counted as “cases” or (in the case of
non-hospital prescribed medications) “items.” While
not identical conceptually, the present study’s cost-
analytic “frame” is similar to that of Wolff et al. (1995).
Financial costs are emphasized, while social costs are
implied. To this end, maintenance, institutional, and
“other” cost categories (e.g. law enforcement and
correctional) are featured. Unlike Wolff et al. (1995),
the authors of the present study did not calculate burden
or burden-sharing factors.
Children’s Mental Health Graphical Tracking
Charts
The two mental health graphical tracking charts,
which follow, are examples of how the data complied
by the authors of the present study can be compacted
into single charts. In the upper left corner, some major
life events, presenting problems at Fort Logan,
medications, and costs (pre-and post-Fort Logan only),
and other interventions and costs of the child are listed.
In the upper right corner, the child’s basic
demographics and his or her total cost to the child-
service system are listed. In the actual chart, the vertical
axis displays the incremental cost expenditures
(monthly service costs), while the horizontal axis shows
the time frame of the expenditures, including the cost of
various residential placements, treatments, and related
services the child encountered during the course of this
study.
Conclusion
While this study was somewhat limited in size, it
does demonstrate that data from existing sources can be
compiled and used to calculate system costs for a
special population. The method is applicable cross-
culturally when the focus is on system-wide cost
analysis for any population. That is, by employing a
financial cost analysis, any government’s expenses to
institutions can be measured and a given population can
be tracked through the system.
Tracking as conducted in the current study has
enormous cost. Including outcome tracking in an
integrated mental health system database may decrease
the cost. State mental health data systems are already
conducting outcome tracking in a limited sense, since
discharge data is routinely collected. However, after
discharge from the state hospital, for example, the
client may have other placements that will not be
included in the state mental health data base, until the
next admission to the hospital. Accurate outcome
tracking data, i.e., where the client has been treated
since discharge from the hospital, and outcomes from
the treatment, can inform this process. The outcome
tracking system would likely function as a type of
“registry,” in which all mental health client services and
outcomes would be recorded.
While overall system costs may be described using
this method, alternative placements and costs can be
analyzed as well. In some cases, the costs for specific
services or products, for example, medications are very
difficult to measure since those costs may be subsumed
within other costs such as residential placement.
Moreover, both Graphing Tracking Charts 7 and 12
demonstrate the relatively high cost of private
psychiatric hospitalization, which reflects the degree of
treatment and security compared to other alternative
forms of residential and non-residential treatment.
Generally, costs tended to increase prior to
hospitalization to Fort Logan stabilized, then tailed off
as the individual reintegrated into the larger
community.
The advantages of the graphical tracking charts are
that they provide a historical context and can be read
quickly and they provide a good overview of the child’s
encounters with the child-service system. Moreover,
this method could provide analyses of system-wide
accountability of child placements, child-service system
best practices analyses, and of course money
expenditures.
Notes
1. Michael H. Moynihan and George Kerin work as
independent consultants, in California and Colorado,
respectively. Moynihan also teaches at the College of
San Mateo and Skyline College. Peter W. Van Arsdale
and Eric Curton work for the Colorado Mental Health
Institute – Ft. Logan. Van Arsdale also teaches at the
University of Denver. William Madura works for
Colorado’s Division of Child Welfare Services, and
also teaches at Metropolitan State College. Among
them, the disciplines of social psychology, cultural
anthropology, social work, clinical psychology, and
educational research are represented. Inquiries can be
addressed to Van Arsdale at CMHI-Ft. Logan, 3520 W.
Oxford Avenue, Denver, Colorado 80236.

172
References Cited
Achenbach, T.
1991 Manaul for the Teacher’s Report Form and
1991 Profile. University of Vermont,
Department of Psychiatry. Burlington,
Vermont.
Barton, T., V. Pillai, and T. Dietz
1996 Program evaluation using event history
analysis. In Evaluation Practice, 17:7-17.
Buchard, J., M. Gendebien, and B. Hinden, et al.
1992 The Vermont system for tracking client
progress. Poster presented at the 5th
Annual
Conference of the Research and Training
Center for Children’s Mental Health. Tampa,
Florida.
Chandler, D., T. Hu, and J. Meisel, et al.
1995 Mental health costs, other public costs, and
family burden among mental health clients in
capitated integrated service agencies.
Unpublished government manuscript. Lewin-
VHI/California, Department of Mental Health.
Chandler, D., J. Meisel, and T. Hu, et al.
1996 Client outcomes in a three-year controlled
study of an integrated service model. In
Psychiatric Services, 47:1337-1343.
Coen, A., H. Shockley, and L. Bradley
1996 FY 1991 estimated costs and expenditures
associated with a cohort of individuals with
serious mental disabilities: An update of FY
1987 costs and expenditures for two Colorado
communities. Unpublished government
manuscript. Colorado Mental Health Services,
Denver: Colorado.
Curry, J.
1991 Outcome research on residential treatment:
Implications and suggested directions. In
American Journal of Orthopsychiatry, 61:348-
357.
Ellis, R., N. Wilson, and F. Foster
1985 State-wide treatment outcome assessment in
Colorado: The Colorado Client Assessment
Record (CCAR). Unpublished government
manuscrip. Colorado Division of Mental
Health, Denver: Colorado.
Forward, J., M. Moynihan, and N. Wilson
1990 Tracking systems research for severely
disturbed children and adolescents in
Colorado 1990-1992. Unpublished
government manuscript. Colorado Division of
Mental Health, Denver: Colorado.
Gendebien, M., B. Hinden, and N. Pandina, et al.
1992 The Vermont system for tracking client
progress. Paper presented at the 5th
Annual
Conference of the Research and Training
Center for Children’s Mental Health, Tampa:
Florida.
Goodrick, W.
1994 Research issues in adolescent inpatient
psychiatry. In Handbook of Adolescent
Inpatient Psychiatric Treatment. H. Ghuman
and R. Sarles, editors. New York:
Brunner/Mazel.
Gossett, L., J. Lewis, and F. Barnhart.
1983 To Find a Way: The Outcome of Hospital
Treatment of Disturbed Adolescents. New
York: Brunner/Mazel.
Grob, M., and J. Singer
1974 Adolescent Patients in Transition: Impact of
Outcome of Psychiatric Hospitalization. New
York: Behavioral Publications.
Hawkins, R., M. Almeida, and B. Fabry, et al.
1991 Restrictiveness of living environments for
troubled children and youths: A sample
measure for program evaluation, policy
planning, and placement decisions. In
Hospital and Community Psychiatry, 43:54-
57.
Heller, L., and A. Coen
1996 Supporting the socio-emotional development
of young children and their families in
Colorado: A review of early intervention
issues. Unpublished government manuscript,
Mental Health Services, Denver: Colorado.
Hernandez, M., and S. Hodges
1996 The ecology of outcomes. In The System
Accountability Project for Children’s Mental
Health. University of South Florida, Tampa:
Florida.

Imse, A.
2000 Hannah’s dark journey. Denver Rocky
Mountain News, Denver: Colorad. March 12.
Kutash, K., and V. Rivera
1996 What Works in Children’s Mental Health
Services? Uncovering Answers to Critical
Questions. Baltimore: MD.
Piers, E.
1984 Piers-Harris Children’s Selt-Concept Scale,
Revised Manual 1984. Los Angeles: Western
Psychological Services.
Shern, D., A. Coen, and L. Bradley, et al.
1990 The comprehensive costs of chronic mental
illness: Estimates from two Colorado
communities. In Proceedings of the
NASMHPD Research Institute Naitonal
Conference on State Mental Health System
Research. Arlington: VA.
Van Arsdale, P.
1996 Ethnography, the life course, and the use of
event calenders: A research note. Cultural
Anthropology Methods Journal, 8:11-12.
Wolff, N., T. Helminiak, and R. Diamond.
1995 Estimated societal costs of assertive
community mental health care. Psychiatric
Services, 46:898-906.

174

A Method For Tracking Severely Emotionally Disturbed Children And Adolescents Through A State Mental Health System

More Related Content

Similar to A Method For Tracking Severely Emotionally Disturbed Children And Adolescents Through A State Mental Health System (20)

More from Andrea Porter (20)

Recently uploaded (20)

A Method For Tracking Severely Emotionally Disturbed Children And Adolescents Through A State Mental Health System