Content uploaded by John Monahan
Author content
All content in this area was uploaded by John Monahan on Aug 27, 2014
Content may be subject to copyright.
Racial Disparities In
Involuntary Outpatient
Commitment: Are They Real?
Disparitiesinoutpatientcommitmentmustbeunderstoodthrough
the settings where commitment is initially considered.
by Jeffrey Swanson, Marvin Swartz, Richard A. Van Dorn, John Monahan,
Thomas G. McGuire, Henry J. Steadman, and Pamela Clark Robbins
ABSTRACT: In this paper we explore racial disparities in outpatient civil commitment, us-
ing data from Kendra’s Law in New York State. Overall, African Americans are more likely
than whites to be involuntarily committed for outpatient psychiatric care in New York. How-
ever, candidates for outpatient commitment are largely drawn from a population in which
blacks are overrepresented: psychiatric patients with multiple involuntary hospitalizations
in public facilities. Whether this overrepresentation under court-ordered outpatient treat-
ment is unfair depends on one’s view: is it access to treatment and a less restrictive alter-
native to hospitalization, or a coercive deprivation of personal liberty? [Health Affairs 28,
no. 3 (2009): 816–826; 10.1377/hlthaff.28.3.816]
Over the past three decades, arguments over involuntary psychiat-
ric intervention versus the liberty interests of people with mental illnesses
have followed the path of deinstitutionalization into community care.
Whether or not it is legitimate and effective to extend the state’s civil commitment
authority into outpatient treatment—applying it to people who would not meet
the more stringent criteria for compulsory inpatient confinement—has become
one of the most contentious issues in mental health law and policy.
Critics of outpatient commitment allege that historically oppressed racial mi-
nority groups are being singled out disproportionately for court-ordered treat-
ment, and this has become a focal point for the larger debate over the policy’s basic
function—whether it represents a pathway to voluntary treatment and recovery,
or a mechanism of social control masquerading as mental health care.1We explo r e
816 May/June 2009
State Policy
DOI 10.1377/hlthaff.28.3.816 ©2009 Project HOPE–The People-to-People Health Foundation, Inc.
Jeffrey Swanson ( jeffrey.swanson@duke.edu) and Marvin Swartz are professors, Psychiatry and Behavioral
Sciences, at the Duke University School of Medicine in Durham, North Carolina. Richard Van Dorn is an
assistant professor at the Florida Mental Health Institute, University of South Florida, in Tampa. John Monahan
is the John S. Shannon Distinguished Professor at the University of Virginia School of Law in Charlottesville.
Thomas McGuire is a professor, Health Care Policy, at Harvard Medical School in Boston, Massachusetts. Henr y
Steadman is president and Pamela Robbins is vice president of Policy Research Associates in Delmar, New York.
racial differences in rates of outpatient commitment and present empirical data to
help interpret those differences: Do they likely result from discrimination in ap-
plying the law to individuals, or from social and economic factors that affect the
racial composition of the service populations where the law is applied?
nCriteria for outpatient commitment. Virtually all U.S. states permit some
form of outpatient commitment, but its use and enforcement vary widely. The legal
criteria typically include having a severe mental illness that limits a person’s ability
to comply voluntarily with recommended treatment, a history of nonadherence to
treatment resulting in a pattern of multiple hospitalizations, and a likely risk of be-
coming dangerous in the future if treatment is forgone.
New York State’s legal eligibility criteria for outpatient commitment, known as
Assisted Outpatient Treatment (AOT), are as follows.2Apersonmaybeplacedin
AOT only if, after a hearing, the court finds that all of the following have been met:
the person must (1) be age eighteen or older; (2) suffer from a mental illness; (3) be
unlikely to survive safely in the community without supervision, based on a clini-
cal determination; (4) have a history of nonadherence to treatment that has been a
significant factor in his or her being in a hospital, prison, or jail at least twice
within the past thirty-six months or has resulted in one or more acts, attempts, or
threats of serious violent behavior toward self or others within the past forty-
eight months; (5) be unlikely to voluntarily participate in treatment; (6) be, in
view of his or her treatment history and current behavior, in need of AOT to pre-
vent a relapse or deterioration, which would likely result in serious harm to the
person or others; and (7) be likely to benefit from AOT. The court-ordered treat-
ment must be the least restrictive alternative that will allow the person to live
safely in the community.
nTreatment requirements under outpatient commitment. Outpatient com-
mitment orders require compliance with recommended outpatient treatment, but
they stop short of permitting forced medication of legally competent people in out-
patient treatment settings. Typically, these orders are given initially for three to six
months, renewable with a court hearing. When a person under outpatient commit-
ment fails to adhere to the recommended treatment plan, the clinician may request
that the police transport the person to a facility to remedy the nonadherence and po-
tentially be evaluated for inpatient commitment. A recent study of more than 1,000
psychiatric outpatients in public mental health systems in five U.S. cities found that
12–20 percent of these patients had, at some point, been placed in outpatient com-
mitment or a related order for community treatment.3
nCoercion and treatment. Outpatient commitment combines coercion and
treatment; herein lies the key difference between opposing camps. Opponents are
primarily concerned with coercion, seeing outpatient commitment as a deprivation
of liberty and a mechanism of social control. Advocates are primarily concerned
with treatment, viewing outpatient commitment as a form of access to health care
for people who won’t get it on their own.
Outpatient Commitment
HEALTH AFFAIRS ~ Volume 28, Number 3 817
nAllegations of racial bias. An allegation of racial bias in New York State’s out-
patient commitment program—called “Kendra’s Law” in memory of a woman
pushed in front of a subway train by a person with untreated schizophrenia—has
intensified the debate. New York’s program is the largest, best-funded, and most
carefully scrutinized exemplar of outpatient commitment. Between 1999 and 2008,
35 percent of those with outpatient commitment orders have been African Ameri-
cans, who make up 17 percent of the state’s population, while 33 percent of the peo-
ple on outpatient commitment have been whites, who make up 61 percent of the
population.4This has raised difficult questions: Does black New Yorkers’ substan-
tial overrepresentation amount to a true “disparity”? Is outpatient commitment be-
ing applied fairly?
nIOM framework. The most comprehensive review of disproportionality in U.S.
health care remains the 2002 Institute of Medicine (IOM) report, Unequal Treatment.5
One of its most important contributions was to frame health care disparities con-
ceptually and place them in a larger context. In exploring racial disparities and ways
to eliminate them, the IOM framework makes important distinctions among “differ-
ence,” “disparity,” and “discrimination.” The report argues that “disparity” should be
reserved for that portion of the difference in health care quality that is attributable
to (1) systemic, legal, and regulatory factors that treat minorities differently than
their nonminority counterparts; or (2) discrimination, bias, stereotyping, and clini-
cal uncertainty within the system. That portion of the difference that is attributable
to variation in need, clinically appropriate decisions, and patients’ preferences
should be factored out and not be considered a “disparity” (Exhibit 1).6
nDifference versus disparity. Applying the IOM paradigm to outpatient com-
mitment policy, and to Kendra’s Law in New York, is there a disparity? A comparison
of the racial distribution of outpatient commitment cases to the general population
818 May/June 2009
State Policy
EXHIBIT 1
Difference Versus Disparity In The Health Care System
SOURCE: Adapted from T.G. McGuire et al., “Implementing the Institute of Medicine Definition of Disparities: An Application to
Mental Health Care,” Health Services Research 41, no. 5 (2006): 1979–2005 (reprinted with permission).
MinorityNon-minority
Quality of care
Clinical appropriateness and need; patient preference
System factors: structure and operation of health care systems
Discrimination: biases, stereotyping, uncertainty
Difference
Disparity
merely shows a difference. To demonstrate disparity, we would need to consider
whether the underlying clinical need was the same in the two groups; the extent to
which clinical decisions were appropriate in both groups; whether structural or sys-
temic features of outpatient commitment selection (such as its being situated in the
public system of care) might affect groups differently; and how much the groups dif-
fered in subjective preferences for treatment.
In the case of outpatient commitment, however, any finding of disparity would
also raise this question: is disparity a disadvantage? Do the benefits of outpatient
commitment outweigh its drawbacks to recipients—or is the treatment received
worth the coercion required to get it?7
nEffectiveness of outpatient commitment. Empirical evidence for the effec-
tiveness of outpatient commitment is arguably mixed, and certainly contested. Pro-
ponents point to a randomized clinical trial in North Carolina that found that it re-
duces hospitalizations.8Opponents find the evidence inconclusive.9
Secondary analyses from the same North Carolina study showed that outpa-
tient commitment lasting six months or longer was associated with more frequent
use of mental health services and that when such services were combined with
outpatient commitment, medication adherence improved, risk of violent behavior
and victimization declined, and subjective quality of life increased. Outpatient
commitment also tended to increase perceived coercion.10 However, studies of
stakeholders’opinionsaboutoutpatientcommitmenthaveshownthatmanypeo-
ple in the target populations consider its benefits to exceed its costs.11
If one were to assume that outpatient commitment equates to receiving com-
munity-based mental health care that is needed, beneficial, of high quality, and
even preferred by its recipients over the long run, then the initial premise of dis-
parity would be undermined. Under such an assumption, it is whites who might
be considered to be the disadvantaged group relative to African Americans, insofar
as whites are underrepresented among recipients of outpatient commitment and
thus receive proportionately fewer of whatever resources and benefits such pro-
grams might provide.
In contrast, to understand how differences in rates of outpatient commitment
could disadvantage African Americans, we need to go beyond the “more is better”
perspective that underlies much of the health care disparities literature. Rather,
we should see these differences against the historical backdrop of long-term insti-
tutional confinement of people with mental illnesses—blacks in greater propor-
tions than whites—and the subsequent “revolving door” syndrome of involuntary
hospital readmissions. We should also consider the segregating effects of a de
facto two-tier system of care: a private system of care for people with employer-
based insurance; and a public system of care for people who are poor, unemployed,
and uninsured or who have public, entitlement-based insurance.12
nOur task here. In this paper we examine potential racial differences and dis-
parities in outpatient commitment, using data from the implementation of Kendra’s
Outpatient Commitment
HEALTH AFFAIRS ~ Volume 28, Number 3 819
Law in New York. Our analysis compares rates of outpatient commitment for Afri-
can Americans and whites both from a total-population perspective and within the
narrower context of the relevant clinical populations from which outpatient com-
mitment cases are drawn.
Starting from the total population perspective of the IOM framework and as-
suming that outpatient commitment represents undesirable coercion, large racial
disparities may indeed exist. Such disparities could result from social, economic,
and systemic factors that interact to produce higher base rates of involuntary hos-
pitalization among blacks than among whites. However, starting from the nar-
rowerperspectiveofthemostrelevanttargetpopulationsforoutpatientcommit
-
ment—and taking the mental health system as it is—there may be minimal
differences between black and white patients in rates of applying outpatient com-
mitment; without basic differences, the question of disparities becomes moot.
Moreover, if one assumes that outpatient commitment might provide beneficial
and ultimately desirable treatment—a less restrictive alternative to, or a remedy
for, involuntary hospitalization—it may have minimal disadvantages; without ba-
sic disadvantages, the question of difference itself becomes moot. We discuss the
implications of our empirical findings in light of our conceptual framing of outpa-
tient commitment as a policy that may provide both desirable resources and unde-
sirable coercion.
Data And Methods
To examine potential racial differences and disparities in the use of outpatient
commitment, we carried out two sets of statistical analyses. First, we estimated
and compared rates of outpatient commitment in New York for blacks and whites,
using several alterative denominators. These denominators can be thought of as a
series of concentric circles encompassing relevant target populations, from the
broadest to the narrowest definitions of who is “at risk” for receiving outpatient
commitment: the general population, those with severe mental illnesses in the
community, those with severe mental illnesses receiving mental health services,
the public mental health system’s adult services recipient population, those with
severe mental illnesses who have been hospitalized during a given year, and those
who have been involuntarily committed to inpatient facilities more than once in
the previous year. Second, we present a multivariable analysis of the association
between race and outpatient commitment at the county level, to see whether the
relationship may be accounted for by other factors that may vary along with race
and outpatient commitment.
nData sources. New York State Office of Mental Health (OMH) administrative and clini-
cal records on people receiving court-ordered treatment under Kendra’s Law. The state main-
tains an electronic data system that records and monitors the status of all outpatient
commitment orders, including dates of initiation, expiration, and renewal. Informa-
tion is also collected on the personal characteristics of those receiving outpatient
820 May/June 2009
State Policy
commitment, including their race and ethnicity.13
New York State OMH data on patient characteristics and hospital admissions. The state
conducts a survey every two years to collect information on the population being
served in the state’s mental health care system and what kinds of services they are
receiving. The OMH also tracks hospital admissions and the numbers of people
being involuntarily committed to state psychiatric hospitals each year.14
U.S. census online database. We used this database to provide estimates of county
population by race and poverty status.15
County estimates of the pre valence of severe mental illness. Estimates were available that
apply epidemiological survey data to the demographic profile of each county.16
These estimates were obtained for the total number of blacks and whites with se-
vere mental illnesses in each county, whether or not they were in treatment.
nOutpatient commitment rates. Using these data, we calculated rates of out-
patient commitment per 10,000 population in 2003 for blacks and whites. To cap-
ture regional variation in the relationship between race and outpatient commit-
ment, we calculated race-specific rates of commitment within New York regions—
Western, Central, Hudson River, New York City, and Long Island—and also in six
representative index counties—Albany, Erie, Monroe, Nassau, New York, and
Queens. Next, we divided each of the rates for blacks by the corresponding rate for
whites to obtain a ratio of comparison, the “parity index.” If the commitment case
rate was the same in both groups, the parity index was equal to 1. If the rate for
blacks was higher than that for whites, the parity index was higher than 1.
nOutpatient commitment and race. For the second analysis, we conducted a
multivariable, longitudinal regression analysis to assess the relationship between
outpatient commitment and race at the county level and to investigate whether that
relationship might result from the prior correlation of race with other factors linked
to commitment. Initially, we calculated the simple association between two vari-
ables: the counties’ outpatient commitment rate and the racial composition of the
counties’ severely mentally ill population. Then we calculated an adjusted associa-
tion between outpatient commitment and race, controlling for several county char-
acteristics that could be linked to both outpatient commitment and race: the coun-
ties’ poverty rate, total severe mental illness prevalence, the percentage of people
with severe mental illnesses who were in treatment, the involuntary and voluntary
hospitalization rate, and the rate of outpatient mental health services use.
Study Results
Rates of outpatient commitment per 10,000 were higher for blacks than for
whites at every level (Exhibit 2). A broad application of the IOM framework
would identify and further explore these as racial disparities in health and health
care access affecting the system in which outpatient commitment is used. How-
ever, the rates converge—that is, the relative difference between blacks’ and
whites’ outpatient commitment rates diminishes—when moving from the total
Outpatient Commitment
HEALTH AFFAIRS ~ Volume 28, Number 3 821
population level to increasingly restricted definitions of the target population.
nOutpatient commitment rates at the county and state levels. Outpatient
commitment indices for six New York counties and the state show that when con-
sidered for the total population, outpatient commitment affects African Americans
three to eight times more frequently than it affects whites—about five times more
frequently,onaverage,statewide(Exhibit3).Putsimply,ablackNewYorkerchosen
at random from the community would have about a five times greater chance of be-
ingplacedinoutpatientcommitmentthanawhiteNewYorkerchosenatrandom.
However, the analysis also shows that these differences are dependent on con-
text; when considered for the most relevant target populations, the parity index
moves closer to 1. The ratio is greatly reduced when the denominators used are the
numbers of blacks and whites who are estimated to have a severe mental illness.
822 May/June 2009
State Policy
EXHIBIT 2
Outpatient Commitment (OPC) Rates, By Region And Race, New York State, Using
Alternative Population Denominators
Region/race
No. of people
with OPC
orders in
2003
County
population
in 2003 (OPC
rate/10,000)
Estimated
no. of SMI
casesa(OPC
rate/10,000)
OMH service
recipients
(OPC rate/
10,000)
Hospitalized
OMH service
recipients
(OPC rate/
10,000)
OMH service
recipients with
>2 involuntary
admissions in
yearb(OPC
rate/10,000)
New York City
White
Black
192
416
2,345,564
(0.82)
1,393,859
(2.98)
91,670
(20.94)
72,345
(57.50)
28,046
(68.46)
27,994
(148.60)
1,962
(978.59)
2,628
(1,582.95)
611
(3,143.99)
1,140
(3,649.59)
Long Island
White
Black
94
43
1,607,625
(0.58)
153,532
(2.80)
51,600
(18.22)
5,966
(72.08)
10,541
(89.18)
2,492
(172.55)
923
(1,018.42)
267
(1,610.49)
432
(2,175.71)
138
(3,105.94)
Hudson River
White
Black
63
24
1,931,196
(0.33)
196,197
(1.22)
74,851
(8.42)
11,722
(20.47)
16,415
(38.38)
4,592
(52.26)
1,200
(525.00)
370
(648.65)
375
(1,681.94)
96
(2,501.93)
Central
White
Black
22
3
1,204,855
(0.18)
58,004
(0.52)
59,391
(3.70)
5,669
(5.29)
10,615
(20.73)
1,216
(24.67)
791
(278.13)
99
(303.03)
390
(564.44)
31
(959.60)
Western
White
Black
18
18
1,817,497
(0.10)
172,625
(1.04)
79,607
(2.26)
12,333
(14.59)
16,741
(10.75)
4,679
(38.47)
957
(188.09)
338
(532.54)
405
(444.01)
138
(1,301.78)
SOURCE: Original data analysis by authors.
NOTES: SMI is serious mental illness. OMH is Office of Mental Health (New York State).
aEstimates from epidemiological survey data applied to local county demographic characteristics.
bEstimates based on extrapolation from involuntary admission rates in OMH-licensed facilities.
These county illness estimates incorporate poverty status, which is statistically
associated with both severe mental illness and African American race.17 The parity
index declines even further when public-sector service recipients are used as the
denominator. Finally, there is no difference in blacks’ and whites’ rates of outpatient
commitment among those who have been involuntarily hospitalized at least twice.
nIs the racial difference in outpatient commitment really about race? We
addressed this question with a county-level analysis in two stages. Exhibit 4 dis-
plays the initial results, showing that counties with a higher proportion of African
Americans among people with severe mental illnesses in the public mental health
system also tend to have markedly higher rates of outpatient commitment.
We then examined the significance of this association in a multivariable, longi-
tudinal regression analysis. Without adjusting for other county-level factors,
county-years with a high proportion of black service recipients who had severe
mental illnesses were more than nine times as likely to have a high rate of outpa-
tient commitment, compared to county-years with a lower proportion of such re-
cipients. However, the net association between race and outpatient commitment
Outpatient Commitment
HEALTH AFFAIRS ~ Volume 28, Number 3 823
EXHIBIT 3
Outpatient Commitment (OPC) Racial Parity Indices In Six Representative New York
Counties And Statewide: Black-To-White Ratios Of OPC Case Rates In 2003 Using
Alternative Denominators
SOURCE: Original data analysis by the authors.
NOTES: Period prevalnce of outpatient commitment cases active in 2003, by selected denominators. SMI is serious mental
illness. Parity line ratio = 1.00. OMH is Office of Mental Health (New York State).
7.00
6.00
5.00
4.00
Ratio, OPC rate for blacks to OPC rate for whites
3.00
County
population
County SMI
population
Erie
Albany
8.00
Nassau
Monroe
2.00
1.00
0.00
County SMI
population in
OMH services
OPC case rate denominator
County SMI popula-
tion in OMH services
and hospitalized
during year
County SMI
population with >1
involuntary hospital-
ization during year
New York
Weighted state mean
Queens
rate was not statistically significant when other factors including poverty, the
prevalence of mentally ill people in the community, and rates of hospitalization
and outpatient mental health services use were controlled for (Exhibit 5). This
analysis implies that the outpatient commitment rate is influenced by a number of
“upstream” social and systemic variables that may correlate with race.
Discussion
Involuntary outpatient commitment is one of the most controversial issues in
mental health law today. We propose that disparities in outpatient commitment
must be understood through the social, clinical, and institutional settings where it
is initially considered. In addition to the formal legal criteria for applying outpa-
tient commitment in any given case, there are upstream factors—such as poverty
and the organization and financing of public mental health care—that bring some
people into target populations.
Our analysis allowed us to address empirically the questions bearing on the
fairness of outpatient commitment’s application. We found no evidence of racial
bias. Defining the target population as public-system clients with multiple hospi-
talizations, the rate of application to white and black clients approaches parity.
As the literature on disparities emphasizes, and we underline, absence of racial
bias does not equate to absence of disparity. Our data show disproportionality in
the application of outpatient commitment to the black population overall. But the
underlying facts are that in comparison to whites in the same county, blacks are
more likely to suffer from severe mental illnesses and, conditional on illness, are
824 May/June 2009
State Policy
EXHIBIT 4
County Outpatient Commitment (OPC) Rate Distribution, By Percentage African
Americans With Severe Mental Illnesses In The Public Mental Health System
SOURCE: Original data analysis by the authors.
40
30
20
10
Percent of county-year observations
0
Low (0–5%) Medium (6–20)
Percent of county SMI service population that is African American
High (>20%)
Medium (1–5 per 100,000 in OPC)
Low (no OPC orders)
50
High (>5 per 100,000 in OPC)
60
more likely to be served in the public system. This could reflect both a health dis-
parity between whites and blacks and a disparity in where people get care. What
also deserves emphasis is that these differences are “upstream” from outpatient
commitment itself. The factors that lead to higher rates of serious mental illness
among blacks are various and poorly understood, but they certainly do not include
outpatient commitment. Likewise, the determinants of social position that im-
pede blacks’ access to private health insurance and private services do not impli-
cate outpatient commitment. Thus, our data are consistent with an outpatient
commitment program that treats clients equally with respect to race.
Outpatient commitment is a huge natural experiment, and
the jury is still out on whether its potential benefits outweigh its social
and human costs for people with serious mental illnesses. However, we
found no evidence that would suggest that the observed correlation between race
and outpatient commitment in New York State results from bias on the part of
outpatient commitment petitioners and legal decisionmakers; rather, within the
narrow population from which candidates come, rates of outpatient commitment
for blacks and whites were very similar. Insofar as outpatient commitment by stat-
ute targets a “revolving door” population of involuntarily hospitalized patients
who are concentrated in the public mental health system, it will inevitably select a
greater proportion of African Americans than their share in the general popula-
tion, because that is the racial distribution of the target population—for historical
reasons unrelated to outpatient commitment. Whether that is good or bad, on bal-
ance, for the population disproportionately affected remains to be seen.
Outpatient Commitment
HEALTH AFFAIRS ~ Volume 28, Number 3 825
EXHIBIT 5
Logistic Regression Time-Series Analysis Of High County Outpatient Commitment
(OPC) Rate By Proportion Of African Americans In The Severe Mental Illness (SMI)
Service Recipient Population, Unadjusted And Adjusted For Relevant Covariates
Odds ratio for high OPC rates (>5 per 100,000)
Unadjusted Adjusteda
Independent variables OR
95% confidence
interval OR
95% confidence
interval
African American SMI population
Low (0–5%) [reference]
Medium (6–20%)
High (>20%)
[1.00]
2.08
9.64
(0.83–5.21)
(3.33–27.89)****
0.85
1.62
(0.34–2.14)
(0.41–6.36)
SOURCE: Original data analysis by authors.
NOTE: Analytic N = 305 county-year observations for 62 counties in 2000–2006.
aControlling for (county-level) year, population, poverty rate, SMI prevalence, SMI in treatment, involuntary and voluntary
hospitalization rate, Assertive Community Treatment (ACT) rate, and intensive case management (ICM) rate.
****p< 0.0001
Portions of the study results were presented to a scientific audience at the annual meeting of the American
Psychology–La w Society Meeting in Jacksonville, Florida, 7 March 2008. The authors are conducting a
legislatively mandated, independent evaluation of the impact of Kendra’s Law in Ne w York, otherwise known as
Assisted Outpatient Treatment (AOT). The study is funded by the New York State Office of Mental Health through
a competitive bid, with additional support from the John D. and Catherine T. MacArthur Foundation through its
Research Network on Mandated Community Treatment.
NOTES
1. New York Lawyers for the Public Interest, “Implementation of ‘Kendra’s Law’ Is Severely Biased,” 7 April
2007, http://www.nylpi.org/images/FE/chain234siteType8/site203/client/DLC%20-%20Report%20on%
20Kendra’s%20Law.pdf (accessed 20 February 2009); and M. Cooper, “Racial Disproportion Seen in Ap-
plying ‘Kendra’s Law’,” New York Times, 7 April 2005.
2. New York State Office of Mental Health, “Eligibility Criteria for AOT,” http:// bi.omh.state.ny.us/aot/about?
p=eligibility (accessed 19 February 2009).
3. J.Monahanetal.,“UseofLeveragetoImproveAdherencetoPsychiatricTreatmentintheCommunity,”
Psychiatric Ser vices 56, no. 1 (2005): 37–44.
4. New York State OMH, “Assisted Outpatient Treatment Reports: Characteristics of Recipients,” http://bi
.omh.state.ny.us/aot/characteristics?p=demographics-race (accessed 20 February 2009).
5. B.D. Smedley, A.Y. Stith, and A.R. Nelson, eds., Unequal Treatment: Confronting Racial and Ethnic Disparities in
Health Care (Washington: National Academies Press, 2002).
6. T.G. McGuire et al., “Implementing the Institute of Medicine Definition of Disparities: An Application to
Mental Health Care,” Health Services Research 41, no. 5 (2006): 1979–2005.
7. J.W. Swanson et al., “Effects of Involuntary Outpatient Commitment on Subjective Quality of Life in Per-
sons with Severe Mental Illness,” Behavioral Sciences and the Law 21, no. 4 (2003): 473–491.
8. M.S. Swartz et al., “Can Involuntary Outpatient Commitment Reduce Hospital Recidivism? Findings
from a Randomized Trial with Severely Mentally Ill Individuals,” American Journal of Psychiatry 156, no. 12
(1999): 1968–1975.
9. S. Kisely et al., “Randomized and Non-randomized Evidence for the Effect of Compulsory Community
and Involuntary Out-Patient Treatment on Health Service Use: Systematic Review and Meta-Analysis,”
Psychological Medicine 37, no. 1 (2007): 3–14.
10. M.S. Swartz et al., “A Randomized Controlled Trial of Outpatient Commitment in North Carolina,” Psychi-
atric Services 52, no. 3 (2001): 325–329.
11. M.S. Swartz et al., “Assessment of Four Stakeholder Groups’ Preferences concerning Outpatient Commit-
ment for Persons with Schizophrenia,” American Journal of Psychiatry 160, no. 6 (2003): 1139–1146.
12. M.S. Swartz et al., “Administrative Update, Utilization of Services, I. Comparing Use of Public and Private
Mental Health Services: The Enduring Barriers of Race and Age,” Community Mental Health Journal 34, no. 2
(1998): 133–144.
13. New York State OMH, “About AOT,” http://bi.omh.state.ny.us/aot/about (accessed 19 February 2009).
14. New York State OMH, Patient Characteristics Survey, June 2008, http://www.omh.state.ny.us/omhweb/pcs/
pcsmain.htm (accessed 19 February 2009).
15. U.S. Census Bureau, “Population Estimates,” http://www.census.gov/popest/estimates.php (accessed 26
January 2009).
16. C.E. Holzer III, Estimation of Need for Mental Health Services, http://psy.utmb.edu/estimation/estimation.htm
(accessed 20 February 2009); and R.C. Kessler et al., “The U.S. National Comorbidity Survey Replication
(NCS-R): Design and Field Procedures,” International Journal of Methods in Psychiatric Research 13, no. 2 (2004):
69–92.
17. E. Silver, E.P. Mulvey, and J.W. Swanson, “Neighborhood Structural Characteristics and Mental Disorder:
Faris and Dunham Revisited,” Social Science and Medicine 55, no. 8 (2002): 1457–1470; and C.E. Holzer et al.,
“Ethnicity, Social Status, and Psychiatric Disorder: Evidence from the Epidemiologic Catchment Area Sur-
vey,” in Social Psychiatry Across Perspectives, ed.R.Price,B.Shea,andH.Mookherjee(NewYork:Plenum,
1995), 93–104.
826 May/June 2009
State Policy
