Characteristics of Individuals With
Severe Mental Illness Who Use
Alexander S. Young, M.D., M.S.H.S.
Matthew J. Chinman, Ph.D.
Julie A. Cradock-O’Leary, Ph.D.
Greer Sullivan, M.D., M.S.P.H.
Dennis Murata, M.S.W.
Jim Mintz, Ph.D.
Paul Koegel, Ph.D.
ABSTRACT: Emergency services are both a safety net and a locus for acute treat-
ment. While the population with severe, persistent mental illness uses emergency
services at a high rate, few studies have systematically examined the causes of this
service use. This study examines a random sample of 179 people who were high uti-
Alexander S. Young is affiliated with the Department of Veterans Affairs Desert Pacific Mental
Illness Research, Education and Clinical Center Los Angeles, CA, USA, UCLA Department of
Psychiatry, Los Angeles, CA, USA, and RAND, Santa Monica, CA, USA.
Matthew J. Chinman is affiliated with the Department of Veterans Affairs Desert Pacific Mental
Illness Research, Education and Clinical Center, Los Angeles, CA, USA and RAND, Santa Monica,
Julie A. Cradock-O’Leary is in private practice, in Anchorage, AK, USA.
Green Sullivan is affiliated with RAND, Santa Monica, CA, USA and Department of Veterans
Affairs South Central Mental Illness Research, Education and Clinical Center, North Little Rock,
Dennis Murata is affiliated with Los Angeles County Department of Mental Health, Los
Angeles, CA, USA.
Jim Mintz is affiliated with Department of Veterans Affairs Desert Pacific Mental Illness
Research, Education and Clinical Center, Los Angeles, CA, USA and UCLA Department of
Psychiatry, Los Angeles, CA, USA.
Paul Koegel is affiliated with RAND, Santa Monica, CA, USA.
Address correspondence to Alex Young, M.D., West Los Angeles Veterans Healthcare Center,
11301 Wilshire Blvd. (210A), Los Angeles, CA 90073, USA; Phone: +1-310-478-3711 x42460; fax:
Community Mental Health Journal, Vol. 41, No. 2, April 2005 (? 2005)
? 2005 Springer Science+Business Media, Inc.
lizers of services from the Los Angeles County Department of Mental Health. Inter-
views were conducted and 5 years of service use data were studied. Greater use of
emergency services was associated with male gender, minority race, severe illness,
homelessness, and less family support. Efforts to reduce emergency services need to
improve access to appropriate community services, particularly for people who are
homeless or lack family support.
KEY WORDS: adult; emergency services; psychiatric; mental disorders; schizophrenia; mental
health services; health care utilization; African Americans; needs; health services accessibility;
human; United States.
A variety of community-based treatments are highly effective when
received by people with severe, persistent mental illness (SPMI).
However, mental health funding is limited and provider organizations
do not have enough resources to provide full access to state-of-the-art
care (President’s New Freedom Commission on Mental Health, 2003).
Since a large proportion of mental health dollars continue to be devoted
to emergency and hospital services, policy-makers often seek to reduce
emergency mental health services and redirect funds to ongoing com-
munity-based care (Young, Kapur, & Murata, 2001). Unfortunately,
there have been few systematic studies of the population using emer-
gency services. A better understanding of this population could inform
efforts to reduce the need for emergency care.
In the United States, the only place where people consistently have a
legal right to receive healthcare is the emergency room (Kellermann &
Haley, 2003). As a result, these services function both as a safety net for
populations with limited financial resources and also as a locus for
treatment during periods of crisis. Researchers have studied use of
emergency departments by people with chronic medical disorders such
as asthma and diabetes. People with chronic medical disorders are
more likely to use emergency departments if they are Black, uninsured
or have Medicaid insurance (Oster & Bindman, 2003). Limited access to
primary care contributes substantially to this utilization (Sun, Burstin,
& Brennan 2003).
Much less is known about the use of emergency mental health ser-
vices. The population using these services is likely to have severe
psychiatric disorders and a high level of distress. Also, emergency
mental health services are delivered in a wide range of venues, ranging
from crisis phone lines and mobile outreach teams, to clinics and tra-
ditional emergency departments. As a result, the population using
these services is exceptionally difficult to study, and very few
160 Community Mental Health Journal
researchers have interviewed a representative sample of individuals
using emergency mental health services. Researchers have found that
the staff of psychiatric emergency rooms believe that clients use their
services to get basic needs met or because of drug use or treatment
noncompliance (Arfken, Zeeman, Yeager, Mischel, & Amirsadri, 2002;
Shaner et al., 1995). Also, reducing access to community care has been
shown to increase use of emergency services (Catalano, McConnell,
Forster, McFarland, & Thornton, 2003).
More is known about risk factors for psychiatric hospitalization.
Individuals are more likely to be hospitalized if they are noncompliant
with medication, dissatisfied with family relationships and social sup-
port, have a history of suicide attempts, are male, or display aberrant or
dangerous behaviors (Doering et al., 1998; Sullivan, Young, & Morgen-
stern, 1997). However, many people who use emergency services are not
services. Risk factors may differ in these two populations.
Access to healthcare services is affected both by characteristics of the
healthcare system and society, and also by characteristics of the indi-
vidual (Andersen & Davidson, 2001). Individual factors have been
categorized as: (a) predisposing factors, such as demographics and
health beliefs, (b) enabling factors, such as family and community re-
sources, and (c) need, including symptoms and functioning. In this
study, we chose to examine individual factors that had proven to be
important in previous studies of psychiatric hospitalization or in
studies of general medical emergency services.
The purpose of this study is to better understand individual char-
acteristics that affect use of emergency mental health services. Most
people with SPMI do not use emergency rooms or a hospital bed in a
given year. However, there is a small group of expensive ‘‘high utiliz-
ers’’ who account for a large proportion of emergency service use,
hospitalization, and overall treatment costs (Young, Sullivan, Murata,
Sturm, & Koegel, 1998). We examine emergency service use and costs
in a random sample of high utilizers from the Los Angeles County
Department of Mental Health (LAC DMH). Participants were inter-
viewed and their service utilization was studied over a 5-year period.
The LAC DMH oversees a geographic area that is larger than the states of Delaware
and Rhode Island combined. In 1993, LAC DMH provided care to more than 60,000
adults with SPMI; and oversaw a treatment network that included two state hospitals,
Alexander S. Young, M.D., M.S.H.S. et al. 161
two county hospitals, 28 directly operated adult programs, and 410 mental health
contracts with 100 community organizations. Participants in this study were drawn
from people who were eligible for the Los Angeles ‘‘PARTNERS’’ program, a novel
assertive community-based treatment program (Young et al., 1998). Individuals were
eligible for PARTNERS if they were between 18 and 64 years old and high utilizers of
mental health services from LAC DMH. High utilization was defined as receiving
treatment from LAC DMH for at least 3 of the years between 1988 and 1993, and
having an average annual treatment cost of $30,000 or greater. The inclusion criteria
resulted in 1241 eligible individuals, of whom 396 were randomly selected for this
survey. This study was approved by the RAND Institutional Review Board.
A survey was developed, pilot tested in people with SPMI, and revised as necessary.
Symptom severity was evaluated using items from the Brief Symptom Inventory
(Derogatis & Melisaratos, 1983) and the Diagnostic Interview Schedule (Robins, Hel-
zar, Croughan, & Ratcliff 1981). Contact with family and friends, and satisfaction with
family, social activities, living situation, daily living, safety, and health were measured
using subscales from the Lehman Quality of Life Interview (QOLI); (Lehman, 1988).
History of homelessness, alcohol and drug use, self-mastery, and physical health were
assessed using established scales (Koegel & Burnam, 1988; Perlin & Schooler, 1978;
Ware & Sherbourne, 1992).
Eligible individuals who gave informed consent were interviewed. The survey was
completed by 45% of eligible individuals (n ¼ 179). The most common causes of non-
response were refusal to complete the survey (n ¼ 83), inability to locate the individual
(n ¼ 62), profound cognitive problems (n ¼ 22), and language barriers (n ¼ 23). Data on
emergency utilization and costs were obtained from LAC DMH databases and included
telephone services, mobile outreach, clinic-based emergency services, hospital-based
psychiatric emergency rooms, and family crisis services (Kapur, Young, Murata, Sul-
livan, & Koegel, 1999). The respondents’ usual clinicians assessed diagnoses and
severity of illness using the DSM-III-R and Global Assessment of Functioning (GAF);
(American Psychiatric Association, 1987).
Analyses included two multiple regressions. Both started with the same independent
variables: gender, race, average GAF score over the prior 6 years, contact with family,
history of homelessness, self-mastery, symptoms, physical health, alcohol use, drug
use, and the QOLI scales. The first analysis used logistic regression with the entire
sample to predict any vs. no emergency service use. The second analysis used least
squares regression in the population with any emergency service use to predict the cost
of emergency services. People with no emergency service use were omitted for this
analysis. The dependent variable in this analysis was the total annual cost of emer-
gency services log-transformed. Stepwise backward elimination was used until the
final model contained only variables with effects at p < 0.10. In addition to these two
separate analyses, a two-part analytic model was conducted since the sample contained
people with no emergency service use. This yielded similar results, and only the two
separate analyses are reported.
Respondents were 65% male (n ¼ 116), 53% White (n ¼ 95), 31% Afri-
can-American (n ¼ 55), and 16% other races (n ¼ 33). Twenty-four
percent also stated they were Hispanic (n ¼ 43). Mean age was
39 years (SD ¼ 9) and mean years of education was 12 (SD ¼ 3). Sev-
enty percent had never married (n ¼ 125); 12% were separated, di-
162Community Mental Health Journal
vorced, or widowed (n ¼ 21); and 3% were married (n ¼ 5). Mean
monthly income was $660 (SD ¼ $680) and 86% were receiving SSI
(n ¼ 154). Average GAF score was 35 (SD ¼ 9). Primary diagnosis was
schizophrenia for 57% of respondents (n ¼ 102), bipolar disorder for
15% (n ¼ 27), major depression for 4% (n ¼ 7), dementia for 2% (n ¼ 4),
and an anxiety disorder for 2% (n ¼ 4). Sixty-five percent also had been
diagnosed with a drug or alcohol disorder (n ¼ 116). Twenty-one per-
cent had a history of homelessness (n ¼ 38), with a lifetime mean of
340 days homeless (SD ¼ 920). Forty percent were living in a house,
apartment, or mobile home (n ¼ 72); 41% in a board and care (n ¼ 73);
and 15% in a residential facility (n ¼ 27).
Eighty-five percent of high utilizers used emergency services at least
once (n ¼ 149). Of those, 48% used hospital emergency rooms (n ¼ 72),
37% used clinic-based emergency services (n ¼ 55), 14% used mobile
outreach (n ¼ 21), 2% used telephonic services (n ¼ 3) and 1% used
family services (n ¼ 2). The distribution of annual emergency service
use cost is presented in Figure 1. The mean annual cost was less than
$100 in 23% of individuals (n ¼ 40). Most individuals cost less than
$2000 per year, and the maximum average annual cost was $4110.
The logistic regression predicting any emergency service use was
significant (chi-square ¼ 20.7, df ¼ 6, p ¼ .002). As shown in Table 1,
men were 8.4 times more likely than women to have any service use,
and a 10 point lower GAF score was associated with a 2.1 times greater
likelihood of service use. The regression model predicting emergency
service costs was also significant (F ¼ 4.3, df ¼ 127, p ¼ .001, R2¼ .18).
Costs were significantly higher for individuals who were male (55%
greater, p ¼ .03), minority (35% greater, p ¼ .03), severely ill (18%
greater per10 GAFpoints,
p ¼ .03),
(59% greater, p ¼ .02) or had less contact with family (18% greater per
QOLI point, p ¼ .04). No significant effects were found for age, marital
status, income, symptom severity, alcohol or drug use, quality of life
domains, or physical health.
In high utilizers with SPMI, individual characteristics significantly
affect use of emergency mental health services. Predisposing factors
and need influenced whether an individual would use any emergency
services. Men were more likely to use any services than women, and
use was more likely for individuals with greater severity of psychiatric
Alexander S. Young, M.D., M.S.H.S. et al.163
illness. Although service use was not associated with severe symptoms,
it was strongly associated with severity of disability and aberrant
behaviors as measured by the GAF. Predisposing factors and need also
affected the amount of services used. However, the amount of services
used was also affected by enabling factors. Both a history of home-
lessness and lack of family contact were associated with greater use of
Family involvement in treatment has been shown to improve out-
comes and prevent rehospitalization (Salokangas, 1997). Homelessness
predisposes individuals to emergency service use by increasing the risk
of victimization and trauma (Padgett, Struening, Andrews, & Pittman,
1995), and by making compliance more difficult. Both housing and
family stabilize individuals in the community, and facilitate access to
outpatient services that avert the need for emergency care.
There clearly is overlap between the population using emergency
mental health services and the population using psychiatric hospitals.
Studies of hospitalization have also found greater use among Blacks
than Whites (Strakowski et al., 1995). Disruptive behaviors increase
(average over 5 years)
$2,500 $3,000$3,500 $4,000
Number of Individuals
Distribution of the Cost of Annual Emergency Service Use for
164Community Mental Health Journal
the risk for hospitalization (Sullivan et al., 1997). These behaviors
would also be expected to lead to emergency service use, and could
explain why this study finds greater emergency service use among men
and severely ill individuals.
In people with SPMI, substance abuse increases rehospitalization
through an association with medication noncompliance (Haywood
et al., 1995). This study does not find an association between substance
abuse and emergency service use. One possible explanation is that the
measures of substance abuse focused on the past 30 days, while utili-
zation was examined over a 5 year period. If the respondents had re-
cently changed their level of substance use, this could have accounted
for the lack of a significant association. In any case, while substance
abuse did not have an independent association with emergency service
use, it was present in more than two-thirds of individuals. It is a pre-
valent and critical problem in users of emergency services.
As is the case with chronic general medical disorders, individuals
using emergency services appear to have worse access to outpatient
treatment. The risk factors for use of emergency services are remark-
ably similar to previously identified risk factors for poor quality
The Effect of Individual Characteristics on Use of Emergency
Services and Emergency Service Costs
Any emergency service
use (n = 179)
Cost of emergency services
for people with any usea
(n = 149)
Odds ratiop value Beta coefficientp value
Notes: The effect of the following variables were not significant at p < .10 and were not included in
the final model: alcohol use; drug use; physical health; symptom severity; marital status; age; total
income; and satisfaction with social activities, family, friends, living situation, daily living, safety
aCosts have been logarithmically transformed.
bGAF was divided by 10, and averaged over the prior 6 years.
Alexander S. Young, M.D., M.S.H.S. et al.165
outpatient care. In the United States, individuals with serious mood
and anxiety disorders have less access to care if they are male or Black
(Surgeon General (US), 2001). Individuals with schizophrenia are less
likely to receive appropriate treatment if they are Black or severely ill
(Young, Sullivan, & Duan 1999; Sue, Fujino, Hu, Takeuchi, & Zone,
One limitation of this study is that it is not able to fully disentangle
individual and contextual factors. It is not possible to tell, for instance,
whether minorities are more likely to use emergency services because
they are individually predisposed to do so, or whether they use these
services because the system of care in minority neighborhoods is
inadequate. Another limitation is that, despite intensive efforts, only
about half of the eligible individuals were interviewed. Utilizers of
emergency services have high rates of homelessness and untreated
psychosis, and these lowered the response rate. While this could affect
the generalizability of the findings, this problem will often be
unavoidable in this population.
Individuals who use emergency mental health services are a par-
ticularly vulnerable population. They are likely to be severely ill,
homeless, and lack support from their family. As with other chronic
medical disorders, people with SPMI who use emergency services are
also at risk for poor access to routine, effective outpatient treatments.
When developing interventions to reduce the need for emergency
mental health services, attention should be paid both to improved ac-
cess to care and also to the importance of enabling factors such as
homelessness and lack of family support. Although Assertive Commu-
nity Treatment has been proven to improve access and community
support, its cost has limited its use to a small proportion of clients.
Other approaches, such as peer support, also show promise. Mental
health authorities should consider a range of strategies for making
community resources available to vulnerable individuals with SPMI.
This research was supported by the Ernst Van Loben Sels Charitable
Foundation, the Zellerbach Family Fund, the NIMH UCLA-RAND
Center for Research on Quality in Managed Care (grant P50 MH-
068639), and the Department of Veterans Affairs. Any opinions ex-
pressed are only the authors’, and do not necessarily represent the
views of any affiliated institutions.
166Community Mental Health Journal
American Psychiatric Association (1987). Diagnostic and Statistical Manual of Mental Disorders,
3rd edition, Revised. Washington, DC: American Psychiatric Association.
Andersen, R. M. & Davidson, P. L. (2001). Improving access to care in America: Individual and
contextual indicators. In R. M. Andersen, T. H. Rice, & G. F. Kominski, (Eds), Changing the
US Health Care System (pp. 3–30). San Francisco, CA: Jossey-Bass.
Arfken, C. L., Zeman, L. L., Yeager, L., Mischel, E. & Amirsadri, A. (2002). Frequent visitors to
psychiatric emergency services: Staff attitudes and temporal patterns. Journal of Behavioral
Health Services and Research, 29 (4), 490–496.
Catalano, R., McConnell, W., Forster, P., McFarland, B., & Thornton, D. (2003). Psychiatric
Emergency Services and the System of Care. Psychiatric Services, 54, 351–355.
Derogatis, L. R., & Melisaratos, N. (1983). The Brief Symptom Inventory: An introductory report.
Psychological Medicine, 13 (3), 595–605.
Doering, S., Muller, E., Kopcke, W., Pietzcker, A., Gaebel, W., Linden, M., Muller, P., Muller-
Spahn, F., Tegeler, J., & Schussler, G. (1998). Predictors of relapse and rehospitalization in
schizophrenia and schizoaffective disorder. Schizophrenia Bulletin, 24 (1), 87–98.
Haywood, T. W., Kravitz, H. M., Grossman, L. S., Cavanaugh, J. L., Davis, J. M. & Lewis, D. A.
(1995). Predicting the ‘‘revolving door’’ phenomenon among patients with schizophrenic,
schizoaffective, and affective disorders. American Journal of Psychiatry, 152 (6), 856–861.
Kapur, K., Young, A.S., Murata, D., Sullivan, G. & Koegel, P. (1999). The economic impact of
capitated care for high utilizers of public mental health services: The Los Angeles PARTNERS
program experience. Journal of Behavioral Health Services and Research, 26 (4), 416–429.
Kellermann, A. L. & Haley, L., Jr. (2003). Hospital emergency departments: where the doctor is
always ‘‘in’’. Medical Care, 41 (2), 195–197.
Koegel, P. & Burnam, M. A. (1988). Alcoholism among homeless adults in the inner city of Los
Angeles. Archives of General Psychiatry, 45 (11), 1011–1018.
Lehman, A. F. (1988). A quality of life interview for the chronically mentally ill. Evaluation and
Program Planning, 11, 51–62.
Oster, A., & Bindman, A. B. (2003). Emergency department visits for ambulatory care sensitive
conditions: insights into preventable hospitalizations. Medical Care, 41 (2), 198–207.
Padgett, D. K., Struening, E. L., Andrews, H. & Pittman, J. (1995). Predictors of emergency room
use by homeless adults in New York City: the influence of predisposing, enabling and need
factors. Social Science & Medicine, 41 (4), 547–556.
Perlin, L., & Schooler, C. (1978). The structure of coping. Journal of Health and Social Behavior,
President’s New Freedom Commission on Mental Health (2003). Achieving the Promise: Trans-
forming Mental Health Care in America. Final Report. http://www.mentalhealthcommis-
sion.gov/reports/reports.htm. Accessed January 21, 2004.
Robins, L. N., Helzer, J. E., Croughan, J., & Ratcliff, K. S. (1981). National Institute of Mental
Health Diagnostic Interview Schedule. Its history, characteristics, and validity. Archives of
General Psychiatry, 38 (4), 381–389.
Salokangas, R. K. (1997). Living situation, social network and outcome in schizophrenia: a five-
year prospective follow-up study. Acta Psychiatrica Scandinavica, 96 (6), 459–468.
Shaner, A., Eckman, T.A., Roberts, L.J., Wilkins, J.N., Tucker, D.E., Tsuang, J.W. & Mintz, J.
(1995). Disability income, cocaine use, and repeated hospitalization among schizophrenic
cocaine abusers–a government-sponsored revolving door? New England Journal of Medicine,
Strakowski, S. M., Lonczak, H. S., Sax, K. W., West, S. A., Crist, A., Mehta, R. & Thienhaus, O. J.
(1995). The effects of race on diagnosis and disposition from a psychiatric emergency service.
Journal of Clinical Psychiatry, 56 (3), 101–107.
Sue, S., Fujino, D. C., Hu, L. T., Takeuchi, D. T., & Zane, N. W. (1991). Community mental health
services for ethnic minority groups: A test of the cultural responsiveness hypothesis. Journal
of Consulting and Clinical Psychology, 59 (4), 533–540.
Sullivan, G., Young, A. S., & Morgenstern, H. (1997). Behaviors as risk factors for rehospital-
ization: Implications for predicting and preventing admissions among the seriously mentally
ill. Social Psychiatry and Psychiatric Epidemiology, 32 (4), 185–190.
Alexander S. Young, M.D., M.S.H.S. et al.167
Sun, B. C., Burstin, H. R., & Brennan, T. A. (2003). Predictors and outcomes of frequent emer- Download full-text
gency department users. Academic Emergency Medicine, 10, 320–328.
Surgeon General (US) (2001). Mental Health: Culture, Race, and Ethnicity, A Supplement to
Mental Health: A Report of the Surgeon General. http://www.surgeongeneral.gov/library/
mentalhealthservices. Accessed January 21, 2004.
Ware, J. E., Jr., & Sherbourne, C. D. (1992). The MOS 36-item short-form health survey (SF-36). I.
Conceptual framework and item selection. Medical Care, 30 (6), 473–483.
Young, A. S., Kapur, K., & Murata, D. (2001). The time course of treatment costs among patients
with severe mental illness. Psychiatric Services, 52 (1), 21.
Young, A. S., Sullivan, G., & Duan, N. (1999). Patient, provider, and treatment factors associated
with poor-quality care for schizophrenia. Mental Health Services Research, 1 (4), 201–211.
Young, A. S., Sullivan, G., Murata, D., Sturm, R., & Koegel, P. (1998). Implementing publicly
funded risk contracts with community mental health organizations. Psychiatric Services, 49
168Community Mental Health Journal