Att A - Community Guide Sample

EAPO Assessment - Attachment A Community Guide Sample.pdf

Surveys of State, Tribal, Local and Territorial (STLT) Governmental Health Agencies

Att A - Community Guide Sample

OMB: 0920-0879

⚠️ Notice: This form may be outdated. More recent filings and information on OMB 0920-0879 can be found here:

Document [pdf]

Download: pdf | pdf

Collaborative Care to Improve the
Management of Depressive Disorders
A Community Guide Systematic Review and
Meta-Analysis
Anilkrishna B. Thota, MBBS, MPH, Theresa Ann Sipe, PhD, MPH, CNM, RN,
Guthrie J. Byard, MPH, Carlos S. Zometa, PhD, MSPH, Robert A. Hahn, PhD, MPH,
Lela R. McKnight-Eily, PhD, Daniel P. Chapman, PhD, Ana F. Abraido-Lanza, PhD,
Jane L. Pearson, PhD, Clinton W. Anderson, PhD, Alan J. Gelenberg, MD,
Kevin D. Hennessy, PhD, Farifteh F. Duffy, PhD, Mary E. Vernon-Smiley, MD, MPH,
Donald E. Nease Jr., MD, Samantha P. Williams, PhD,
Community Preventive Services Task Force
Context: To improve the quality of depression management, collaborative care models have been
developed from the Chronic Care Model over the past 20 years. Collaborative care is a multicomponent, healthcare system–level intervention that uses case managers to link primary care providers,
patients, and mental health specialists. In addition to case management support, primary care
providers receive consultation and decision support from mental health specialists (i.e., psychiatrists
and psychologists). This collaboration is designed to (1) improve routine screening and diagnosis of
depressive disorders; (2) increase provider use of evidence-based protocols for the proactive management of diagnosed depressive disorders; and (3) improve clinical and community support for
active client/patient engagement in treatment goal-setting and self-management.
Evidence acquisition: A team of subject matter experts in mental health, representing various
agencies and institutions, conceptualized and conducted a systematic review and meta-analysis on
collaborative care for improving the management of depressive disorders. This team worked under
the guidance of the Community Preventive Services Task Force, a nonfederal, independent, volunteer body of public health and prevention experts. Community Guide systematic review methods
were used to identify, evaluate, and analyze available evidence.
Evidence synthesis: An earlier systematic review with 37 RCTs of collaborative care studies
published through 2004 found evidence of effectiveness of these models in improving depression
outcomes. An additional 32 studies of collaborative care models conducted between 2004 and 2009
were found for this current review and analyzed. The results from the meta-analyses suggest robust
evidence of effectiveness of collaborative care in improving depression symptoms (standardized
mean difference [SMD]⫽0.34); adherence to treatment (OR⫽2.22); response to treatment
(OR⫽1.78); remission of symptoms (OR⫽1.74); recovery from symptoms (OR⫽1.75); quality of
life/functional status (SMD⫽0.12); and satisfaction with care (SMD⫽0.39) for patients diagnosed
with depression (all effect estimates were signifıcant).

Conclusions: Collaborative care models are effective in achieving clinically meaningful improvements in depression outcomes and public health benefıts in a wide range of populations, settings, and
organizations. Collaborative care interventions provide a supportive network of professionals and
peers for patients with depression, especially at the primary care level.
(Am J Prev Med 2012;42(5):525–538) Published by Elsevier Inc. on behalf of American Journal of Preventive
Medicine

Published by Elsevier Inc. on behalf of American Journal of Preventive Medicine

Am J Prev Med 2012;42(5):525–538 525

526

Thota et al / Am J Prev Med 2012;42(5):525–538

Context

epressive disorders are a major contributor to
the burden of disease in high-income countries1
and lead all diseases as a cause for years of life
lived with disability (YLD).1 In the U.S., 14.8 million
adults (6.7% of the population) experience major depressive disorder.2 Further, approximately 1.5% of the adult
U.S. population experience dysthymic disorder every
year—a chronic illness in the depression spectrum that is
less severe than major depressive disorder.2
The prevalence of major depressive disorder is 50%
higher among women than men,3 as is the burden of
disease.1 The prevalence of Major Depressive Episode
(MDE) in 2008 among people aged ⱖ18 years was found
to be highest among those who identifıed themselves as
multiracial, followed by whites, Hispanics, American Indian or Alaska Natives, blacks, and Asians.4 Among
youth aged 12–17 years, the prevalence of MDE was estimated to be 8.3%.4 Direct medical costs, suicide-related
mortality costs, and productivity losses from depression
totaled $83.1 billion in the U.S. in 2000.5 Although 10.6
million adults reported an unmet need for mental health
services in 2008,4 most people with depressive symptoms
seek treatment at the primary care level, where they might
not receive appropriate care.6 Hence, the quality of depression care in the primary care system needs to be
improved.
Various approaches have been employed to improve
the quality of care for chronic diseases. Notable among
these is the Chronic Care Model,7,8 which has improved
the management of chronic diseases such as diabetes,
especially at the primary care level.9,10 The Chronic Care
Model has been adapted to improve the management of
depressive disorders, leading to the development of the
From the Community Guide Branch, Epidemiology and Analysis Program Offıce, Offıce of Surveillance, Epidemiology, and Laboratory Services, (Thota, Sipe, Zometa, Hahn, Byard), Division of Population
Health, (McKnight-Eily, Chapman), and Division of Adolescent and
School Health, (Vernon-Smiley), National Center for Chronic Disease
Prevention and Health Promotion, and the Behavioral Interventions and
Research Branch, National Center For HIV/AIDS, Viral Hepatitis, STD &
TB Prevention (Williams), CDC, Atlanta, Georgia; Mailman School of
Public Health, Columbia University (Abraido-Lanza), New York, New
York; National Institute of Mental Health, (Pearson), Bethesda, Substance
Abuse and Mental Health Services Administration (Hennessy), Rockville,
Maryland; American Psychological Association (Anderson), Washington,
DC; Department of Psychiatry, Penn State Hershey Medical Center (Gelenberg), Hershey, Pennsylvania; American Psychiatric Association (Duffy),
Arlington, Virginia; and American Academy of Family Physicians (Nease
Jr.), Denver, Colorado
The names and affıliations of the Task Force members are listed at
www.thecommunityguide.org/about/task-force-members.html.
Address correspondence to: Anilkrishna B. Thota, MBBS, MPH, Community Guide Branch, CDC, 1600 Clifton Road, Mailstop E69, Atlanta GA
30333. E-mail: [email protected].
0749-3797/$36.00
doi: 10.1016/j.amepre.2012.01.019

collaborative care model,11 a multicomponent, healthcare system–level intervention that uses case managers to
link primary care providers, patients, and mental health
specialists. Collaborative care models typically include
case managers, who support primary care providers with
functions such as patient education, patient follow-up to
track depression outcomes and adherence to treatment,
and adjustment of treatment plans for patients who do
not improve.
Primary care providers receive consultation and decision
support from mental health specialists (i.e., psychiatrists and
psychologists). This collaboration is designed to (1) improve
routine screening for and diagnosis of depressive disorders;
(2) increase provider use of evidence-based protocols for the
proactive management of diagnosed depressive disorders;
and (3) improve clinical and community support for active
client/patient engagement in treatment goal-setting and
self-management.
Primary care providers are usually responsible for routine screening and diagnosis of depressive disorders, prescribing antidepressants, and referring patients to mental
health specialists as needed. Mental health specialists provide clinical advice and decision support to primary care
providers. These processes are frequently coordinated
and supported by technology-based resources such as
electronic medical records, telephone support, and provider reminder mechanisms.
Systematic reviews of the literature have found evidence to support the effectiveness of collaborative care
models in improving health outcomes related to depressive disorders.12,13 This review builds on that foundation
and provides current evidence on effectiveness of collaborative care in reducing the burden of depressive disorders, as assessed by an expansive range of depression
outcomes. Using methods developed by the Guide to
Community Preventive Services,14,15 various moderators
of effectiveness that can influence outcomes (e.g., patient
and provider characteristics; geographic location) and be
benefıcial to a community interested in implementing
this intervention were examined. Hence, this review offers an opportunity to assess the state of the evidence on
effectiveness in an active research area as well as the
variables that influence the applicability and generalizability of these collaborative care models to various populations and settings.

Guide to Community Preventive Services
The systematic review in this report was conducted under
the oversight of the independent, nonfederal Community
Preventive Services Task Force (the Task Force). The Task
Force continues to develop, expand, and update the Guide to
Community Preventive Services (the Community Guide)
with the support of DHHS in collaboration with public and
www.ajpmonline.org

Thota et al / Am J Prev Med 2012;42(5):525–538

private partners. The CDC provides staff support to the Task
Force for development of the Community Guide. Previous
topics reviewed, as well as background information on
methods and development of the Community Guide, are
available at www.thecommunityguide.org.

Healthy People 2020 Goals and Objectives
Several Healthy People 2020 goals and objectives16 are
relevant to this review.
●

“Reduce the suicide rate” (Mental Health and Mental
Disorders [MHMD]-1);
● “Reduce the proportion of persons who experience
major depressive episodes” (MHMD-4) among adolescents (MHMD-4.1) and among adults aged 18 years
and older (MHMD-4.2);
● “Increase the proportion of adults aged 18 years and
older with major depressive episodes who receive treatment” (MHMD-9.2);
● “Increase depression screening by primary care providers” (MHMD-11).

Information from Other Advisory Groups
The U.S. Preventive Services Task Force (USPSTF),
which provides recommendations for clinical practice,
recommends screening for depression in adults and adolescents when systems are in place for effıcient diagnosis,
treatment, and follow-up for depressive disorders.17,18
Collaborative care models address all these aspects of
care.
The American Psychiatric Association recently released an update to its practice guideline for the treatment
of major depressive disorders,19 providing evidencebased recommendations on the use of antidepressants
and psychotherapy, somatic, and other forms of therapy.
The guidelines cover a range of situations including
treatment-resistant depression, postpartum depression, and comorbid illnesses.

Evidence Acquisition
Community Guide methods (www.thecommunityguide.org/about/
methods.html) were used to conduct this systematic review and metaanalysis to determine the effectiveness of collaborative care in improving
managementofdepressivedisorders.Thesemethodshavebeendescribed
in detail elsewhere.14,20 Briefly, for this review, a coordination team (“the
team”) was constituted, including subject matter experts on mental health
and mental illness from various agencies, organizations, and academic
institutions together with qualifıed systematic reviewers. The team
worked under the oversight of the Task Force.
For each Community Guide review topic, a team conducts a
review by (1) developing a conceptual approach to identify, organize, group, and select interventions for review; (2) developing an
analytic framework depicting interrelationships among interventions, populations, and outcomes; (3) systematically searching for
and retrieving evidence; (4) assessing and summarizing the quality
May 2012

527

and strength of the body of evidence of effectiveness; (5) translating
evidence of effectiveness into recommendations; (6) summarizing
data about applicability (i.e., the extent to which available effectiveness data might apply to diverse population segments and settings),
economic impact, and barriers to implementation; and (7) identifying and summarizing research gaps.
The Task Force receives the results of the review process, which
include (1) effectiveness and consistency of the intervention in
improving health outcomes and preventing disease; (2) quality of
the body of evidence in terms of design and execution; (3) additional benefıts and potential harms and barriers to implementation; (4) applicability or generalizability of the intervention to a
comprehensive range of populations and settings; and (5) information on economic effıciency. The Task Force also takes into
account the public health importance of the overall effect estimates to reach decisions on making recommendations on using
the intervention for practice and policy.

Conceptual Approach and Analytic Framework
The conceptual approach developed by the team to determine the
effectiveness of collaborative care is represented in the analytic
framework (Figure 1). The team hypothesized that the model
would organize a collaborative arrangement among primary care
providers, case managers, and mental health specialists (i.e., psychiatrists and psychologists). This collaborative arrangement enables processes for primary care providers to improve their screening practices and the quality of care for depressive disorders while
receiving case management support from case managers and clinical decision support and clinical advice from mental health specialists. This arrangement also facilitates the active involvement of
clients/consumers/patients (“patients”) in their own care and treatment plans (i.e., self-care). These systemic changes are expected to
lead to improved results across a wide range of depression-related
outcomes.

Outcome Measures Used to Determine
Effectiveness
Consistent with research that describes the course of depression and
treatment,21 the team examined the following widely used depression
outcomes: “depression symptom improvement,” “response to treatment,” “remission,” and “recovery.” Additional primary health outcomes also were examined: “screening and diagnosis,” “adherence to
treatment,” and “health-related quality of life and functional status,”
deemed suitable to facilitate formation of a judgment on intervention
effectiveness. One secondary outcome, “satisfaction with care,” also was
examined. Other outcomes that were directly relevant to depressionrelated morbidity and mortality were likewise eligible for this
review.

Primary Health Outcomes
Depression symptom improvement. Depression symptoms typically are measured with standardized depression scales.
Some examples of scales used in the fıeld include the Structured
Clinical Interview (SCID); the Beck Depression Inventory (BDI);
the Patient Health Questionnaire (PHQ-9); and the Symptom
Checklist (SCL-20 and SCL-90). Decision rules were developed to
determine scale selection when more than one depression scale was
reported.22–25 Scales were selected in the following order: SCID,26
BDI,27 and PHQ-9.28,29

Thota et al / Am J Prev Med 2012;42(5):525–538

528

Collaborative
care

Improved efficiency of
process and tracking

Improved
screening/diagnosis

Enhanced collaborationb
among providersa

Providersa

Improved
depression care
Concordance
Collaboration
Case management
Self-care

Improved support for
client/patient involvement

Increased satisfaction
with care

Increased proportion
of population screened
and diagnosed

Increased adherence
to treatment

Improvement in
depression symptoms
Improved response
Increased remission
Increased recovery

Improved quality of
life and functional status

Legend
Intervention

Secondary
outcomes

Pathways for which
empirical links have
been established

Target
population

Primary
health
outcomes

Pathways for which
empirical links have not
been established

Providers:
Primary care providers (PCP);
Case managers (CM);
Mental health specialists (MHS)

b Collaboration

PCP

Client
CM

MHS

Figure 1. Analytic framework depicting hypothesized collaborative care impact on screening, treatment, and outcomes
of depressive disorders

Response to treatment. Response to treatment was generally
defıned by the commonly accepted convention of reduction in
depressive symptoms of ⱖ50% from baseline.30 However, results
of studies of people with severe depression or treatment-resistant
depression, and those that defıned “response” based on a lower
degree of improvement (e.g., 25% improvement from baseline),
also were included.30

Remission and recovery. To achieve remission, a commonly
accepted criterion states that a “virtual” absence of depressive symptoms must be attained.30 This is defıned by the absence of depression
symptoms or scores below suggested cutoff points on a depression
scale. A patient who is in remission for 4 consecutive weeks is considered to have recovered.30 Because researchers often use these terms
interchangeably and without clearly defıning them, the team defıned
and analyzed three analogous outcomes to remission and recovery,
based on the follow-up period from the beginning of treatment. These
were (1) remission reported at ⬍6 months; (2) remission reported at
6 months; and (3) remission reported at or close to 12 months (considered a proxy for recovery).

Adherence to treatment. Adherence to treatment was defıned as the proportion of patients following an agreed-upon treatment plan, which could include medication and/or other forms of
treatment, such as psychotherapy. Because of the challenges of
assessing adherence, proxies (e.g., evidence of fılled prescriptions
or of receiving or taking a therapeutic dose) were accepted.
Health-related quality of life and functional status. Healthrelated quality of life is “an individual’s or group’s perceived physical and mental health over time.”31 Functional status is “the extent
to which an individual can function to meet basic needs, conduct

his/her regular roles, and preserve health and wellness.”32,33 Reported outcomes that measured both quality of life and functional
status include the Short Form Health Survey–36 or some variant,
EuroQol, and the Functional Assessment of Cancer Therapy–G.
Some examples of measures of functional status alone include
the Health of Nations Outcome Scales 65⫹, the Social Adaptation
Self-evaluation Scale, and Patient Global Impression. If authors
reported more than one subscale for this outcome measure, then
mean effects were calculated. Because these tools have been validated for use in assessing health-related quality of life and functional status, the team pooled the effects reported by individual
studies using these tools to estimate the impact of collaborative care
models on this outcome.

Secondary Outcome
Satisfaction with care. Satisfaction with care is “a patient’s
perception of (1) the quality of healthcare providers, (2) access to
services, (3) communication with providers and administrative staff,
and (4) the success of their treatment.”34,35 This outcome could be
assessed via standardized instruments, such as the Patient Satisfaction
Index or the Client Satisfaction Questionnaire, or by other researcherdeveloped measures of patient satisfaction.

Search for Evidence
Electronic searches were conducted in the following databases: The
Cochrane Library; MEDLINE; Embase; ERIC; NTIS (National
Technical Information Service); PsycINFO; CABI; LILACS;
CINAHL; and Dissertation Abstracts International. Hand-searches
were conducted of fıve journals, published in the 10 years preceding
the review and identifıed by the team as the most relevant to the fıeld
www.ajpmonline.org

Thota et al / Am J Prev Med 2012;42(5):525–538
and this intervention. Also included were unpublished papers, conference proceedings, reports, books, and book chapters identifıed by
team members and other subject matter experts. The initial literature
search was conducted in April 2008 with an updated search in February 2009. Search terms are available at www.thecommunityguide.org/
mentalhealth/SS-collab-care.html.

Criteria for Inclusion
Studies were considered for inclusion in this systematic review if
they
●
●

●

●
●

●

were written in English;
evaluated collaborative care interventions that included at least a
case manager, primary care provider, and mental health specialist with collaboration among these roles;
evaluated interventions targeted to patients with a diagnosis of
major depression, minor depression, or dysthymia, without comorbid psychoses;
were conducted in a high-income nationa;
compared a group of people who had been exposed to the intervention with a group of people who had not been exposed or who
had been less exposed (these comparisons could be concurrent
or in the same group over time); and
measured and reported a primary health outcome of interest as
described above.

Assessment of Quality and Summarizing the Body
of Evidence on Effectiveness
Two reviewers read and evaluated each study that met the
inclusion criteria using an adaptation of the standardized abstraction form for Community Guide reviews (available
at www.thecommunityguide.org/library/ajpm355_d.pdf), and
disagreements were resolved by consensus between reviewers or
among the entire review team. Reviewers were not blinded to
author or journal names. Each study was assessed for suitability of
study design and threats to validity.14 Based on the number of
threats to validity, studies were characterized as having good (0 –1
limitation); fair (2– 4 limitations); or limited (ⱖ5 limitations) quality of execution.14,15
Studies with limited quality of execution were not included in
the summary of the intervention effect. Studies with good or fair
execution were considered qualifying studies and were included
in the analyses. Bodies of evidence of effectiveness are characterized as strong, suffıcient, or insuffıcient on the basis of the
number of available studies, the suitability of study designs for
evaluating effectiveness, the quality of execution of the studies,
the consistency of the results, and the effect estimates.14
Reviewers abstracted data describing collaborative care intervention elements, participant characteristics, study characteristics, and study results using SRS, version 4.0. Nine study authors
a

World Bank high-income economies are as follows: Andorra, Antigua and
Barbuda, Aruba, Australia, Austria, The Bahamas, Bahrain, Barbados, Belgium, Bermuda, Brunei Darussalam, Canada, Cayman Islands, Channel
Islands, Cyprus, Czech Republic, Denmark, Equatorial Guinea, Estonia,
Faeroe Islands, Finland, France, French Polynesia, Germany, Greece,
Greenland, Guam, Hong Kong (China), Hungary, Iceland, Ireland, Isle of
Man, Israel, Italy, Japan, Republic of Korea, Kuwait, Liechtenstein, Luxembourg, Macao (China), Malta, Monaco, Netherlands, Netherlands Antilles,
New Caledonia, New Zealand, Northern Mariana Islands, Norway, Oman,
Portugal, Puerto Rico, Qatar, San Marino, Saudi Arabia, Singapore, Slovak
Republic, Slovenia, Spain, Sweden, Switzerland, Trinidad and Tobago,
United Arab Emirates, United Kingdom, U.S., Virgin Islands (U.S.).

May 2012

529

were contacted by e-mail when data were missing or when
numbers reported in text did not match data reported in tables
or fıgures. All nine authors responded. Information also was
abstracted on applicability, potential harms, additional benefıts,
and barriers to implementation of collaborative care interventions. Additionally, efforts were undertaken to identify research
gaps and research needs in this fıeld.

Economic Evaluation
Evaluations of economic effıciency are conducted only when
suffıcient or strong evidence of effectiveness of an intervention
has been established. The methods and fındings of the economic
evaluation of collaborative care interventions are described in
an accompanying article.36

Data Analysis Methods
Calculation of estimated effect sizes for each study. Estimated effect sizes for this review were expressed as standardized
mean difference with a correction factor (Hedges’ g37) for those
outcomes that were measured as a mean score, and OR for outcomes reported as proportions. Calculation of 95% CIs and adjustment for baseline data were done for all studies with suffıcient
information reported. When necessary, reported results were
transformed so that Hedges’ g values ⱖ0 and ORs ⬎1.0 indicate
effects in the favorable direction.

Meta-analyses. Meta-analyses were conducted on each outcome variable to assess effectiveness of the collaborative care
model. Estimated effect sizes for individual studies were aggregated using the random effects model to calculate an overall
weighted mean effect estimate (Hedges’ g or OR) with a corresponding 95% CI. The random effects model was chosen a priori
because interventions, populations, and contexts vary substantially in community-based interventions.38 Homogeneity tests
also were conducted using the Q statistic38 and the I2 statistic.39
I2 values can range from 0% to 100%, and in this review values
⬎50% were considered indicative of substantial heterogeneity.39 A combination of Microsoft Excel and Comprehensive
Meta-Analysis, version 2.2050, was used for estimated effect size
calculation and meta-analyses.

Subgroup analyses. Between-study analyses were conducted
for several effect modifıer variables to (1) assess whether they are
associated with increased or decreased intervention effects and
(2) explore potential sources of heterogeneity for each outcome.

Sensitivity Analyses Methods
Publication bias analyses. A funnel plot for each outcome was
inspected visually to examine the data for evidence of publication
bias.40 Next, the Begg and Mazumdar rank–correlation coeffıcient41
was calculated to determine bias through correlation of the effect
estimate and the SE. Last, Orwin’s fail-safe N analysis42 was performed
for the “depression symptom improvement” outcome to assess
whether the estimated effect could be attributable to publication bias.
A Hedges’ g of 0.1 was set as the trivial effect for this analysis, which
determines the number of studies with null or contrary results needed
to overturn the observed overall effect estimate.43,44

Thota et al / Am J Prev Med 2012;42(5):525–538

530

Table 1. Meta-analyses results from Bower et al.12 collaborative care systematic review compared to similar
outcomes from Community Guide systematic review
Bower (2006)12

Community Guide

1966a–2004

2004b–2009

Study arms, n

Effect estimate (95% CI)

Study arms, n

Effect estimate (95% CI)

Depression symptom
improvement

SMD

SMDc

Adherence

Outcome name

0.24 (0.17, 0.32)
OR

0.34 (0.25, 0.43)
10

1.92 (1.54, 2.39)

OR
2.22 (1.67, 2.96)

Search period: 1966 –2004; Earliest collaborative care study was from 1993.
Studies from 2004 not in Bower et al.12
c
Hedges’ g was used as standardized mean difference metric.
SMD, standardized mean difference
b

One-study-removed sensitivity analyses. A one-studyremoved sensitivity analysis was performed on each outcome to
examine how the overall weighted mean effect estimate and CIs
changed when an effect size from any one individual study was
removed and, thus, identify individual studies that overly influence the summary effect estimate.45

Evidence Synthesis
A total of 8354 potentially relevant titles and abstracts obtained from the literature search and review of reference lists
were screened. Of these, full-text versions of 1057 published
articles and reports were obtained and 226 papers and reports relevant to collaborative care were identifıed. These
included papers reporting individual study results as well as
reviews of several collaborative care interventions.
One such review was a systematic review and metaanalysis conducted by Bower et al.12 that identifıed 37 RCTs
on collaborative care, published between 1993 and 2004.
Two outcomes comparable to the current systematic review
were assessed and found to provide robust evidence on
effectiveness of collaborative care for depression (Table 1).
Given the similarity of the current systematic review to the
previous review by Bower et al.,12 only the 33 studies22–25,46–74
published during or after 2004 were included in the current
systematic review. All studies in the current review were of
greatest design suitability, and a majority were RCTs in which
allocation of patients to collaborative care or usual care was
randomized with researchers blinded to this allocation. All but
one study61 had good or fair execution. Thus, 32 studies with 39
study arms qualifıed for analysis.

Descriptive Results
The characteristics of the study populations are shown
in Table 2. Studies evaluated included a range of populations and contextual factors reflecting the widespread
practice of collaborative care in a variety of settings.

Results from Meta-Analyses
Primary health outcomes. All primary health outcomes
demonstrated improvement, and forest plots were generated to demonstrate results of the meta-analyses using Comprehensive Meta-Analysis® (Figure 2 is the
forest plot for “depression symptom improvement”). The
overall weighted mean effect estimates for depression
symptom improvement, response to treatment, remission at ⬍6 months, remission at 6 months, recovery at
12 months, and adherence to treatment were in the
favorable direction and of suffıcient size to be considered meaningful for improving health (Table 3).
Health-related quality of life/functional status was also in
the favorable direction but had a smaller effect estimate.
Substantial heterogeneity for many of the outcomes was
found, with I2 ⬎50%. The primary source of this heterogeneity appears to be three outlier studies.24,49,54 Sensitivity analyses conducted with these studies removed resulted in slightly smaller effect estimates and a reduction
in heterogeneity (I2⬍50%). No other sources of heterogeneity were identifıed.
Only one study in the review, Reiss-Brennan et al.,64
provided data on the effectiveness of collaborative care
for improving depression screening rates within a
large health system that implemented collaborative
care. The rates of detection of true depression for
adults and children were slightly higher for collaborative care clinics compared to usual care.
Secondary outcomes. The overall weighted mean effect estimate was in the favorable direction and signifıcant for “satisfaction with care” (Table 3), with no indication of heterogeneity issues.
Additional evidence. Reported outcomes from three
studies that could not be combined with the primary or
www.ajpmonline.org

Thota et al / Am J Prev Med 2012;42(5):525–538

Table 2. Baseline demographic characteristics from all
studies that qualified for analysis
Characteristic
Age group (years)

Gender

Category

Study arms, n

Stratified estimate (Hedges’ g)

0.31

Older adult

0.46

Teen

0.25

0.36

Majority male

0.24

Majority white

0.3

Majority African-American

0.56

Majority Latino

0.26

Low

0.27

Majority low

1.02

Majority mid-high

0.09

0.29

Universal

0.46

0.11

Academic

0.38

CBO

0.82

Clinic

0.31

Hospital

0.33

Clinic/hospital

0.37

Clinic/home

0.6

Home

1.01

Worksite

0.06

0.37

Mental health worker

0.08

MD/RN and/or others

0.1

Social worker

0.27

RN/social worker⫹

1.02

0.42

0.29

Physician’s assistant

0.25

Psychiatrist and/or psychologist and others

0.34

Psychiatrist/psychologist

0.36

0.25

Adult

Majority female

MCO

RN⫹

MD and/or RN
Note: Data were not available for all variables for all studies.
*p⬍0.05
CBO, community-based organization; RN, registered nurse; VA, Veterans Affairs

www.ajpmonline.org

Thota et al / Am J Prev Med 2012;42(5):525–538

Table 5. Collaborative care intervention components for
studies that reported the outcome of depression
symptom improvement
Variable

Study arms, n

Hedges’ g

0.48

Yes

0.30

0.53

Yes

0.25

0.31

Yes

0.35

0.39

Yes

0.32

Patient education

Support for self-care*

Provider education

Provider feedback

Provider oversight
No

0.33

Yes

0.34

0.43

Yes

0.30

0.35

Yes

0.33

0.38

Yes

0.25

Evidence-based guidelines

Use of telephone

Use of technology

*p⬍0.05

tance to enroll,23,74 low patient appointment attendance,52,54,55 limited insurance coverage for mental
health care,73 locating organizations in the community
that offer depression care at such nonconventional
points-of-care as the home setting and the worksite,55,57,73 training specialists from other fıelds in collaborative care for patients with depression comorbid
with other chronic illnesses,54 and diffıculties reaching
patients who preferred face-to-face over telephone
contact for counseling and care management.50

Research Issues
An important research need identifıed from this systematic review concerns the essential training and background required of key members of the collaborative care
May 2012

535

team (e.g., requisite skill levels for case managers and
intervention-specifıc training for case managers and primary care providers).24,48,50 Other needs include information on the optimal frequency and intensity of case
management sessions and the utility of additional sessions for patients who do not improve.22,49
Studies are also needed to ensure that collaborative care
models are consistently effective in improving the management and reducing the impact of depressive disorders
among children and adolescents and when targeted to minorities, those of low SES, and those with comorbid conditions. Only one study examined the effect of collaborative
care on improving the quality of screening practices.64 Research studies that focus on improving depression screening
at the primary care level through collaborative care will be
vital to implementers of these models. Gaining more robust
information and knowledge on these aspects will inform the
effective practice of collaborative care in the community.

Limitations
Care-seeking behavior for mental illness is frequently hindered by societal and cultural stigmas, which often present
the greatest obstacle to any mental health intervention. Although collaborative care models provide motivation and
support to depressed patients who have entered the healthcare system, it is unclear how these interventions can
motivate untreated people with depression to initiate
care-seeking. It might well be outside the purview of
collaborative care interventions to influence this initial
care-seeking behavior for depression at the community
level.
The potential for selection bias when interpreting the
results from studies in the present review must be considered. Researchers might recruit only patients with “major
depression” or “severe depression” into studies, which increases the possible amount of improvement in depression
symptoms. Alternatively, implementers might recruit patients with minor symptoms, increasing the chances of remission or recovery. The interventions represented in the
present review included patients with the entire spectrum of
depressive disorders, from dysthymia to major depression,
reflecting the real-world picture of patients seeking care for
different levels and types of depressive disorders and symptoms. None of the studies provided information on the existence of “double depression,” that is, a major depressive
episode complicating underlying dysthymia, in participants.
Further, sensitivity analyses did not reveal differential effects
by severity of depression.
Sources for other biases, including attrition bias and
referral bias, were identifıed to the extent possible in the
quality scoring process and were not found to be factors
skewing the results from studies within this evidence
base. According to Community Guide methods,14,15 all

536

Thota et al / Am J Prev Med 2012;42(5):525–538

but one study had good or fair quality of execution.
Hence, excluding the one study with limited quality of
execution from analysis was unlikely to have affected the
generalizability of fındings.
Other potential limitations could include the use of the
existing review,12 which included only RCTs, and comparing it with evidence identifıed by the search in the update
interval, which included both RCTs and other study designs
with concurrent comparison groups; restricting studies to
those written in English; and ending the search in 2009.
Given the large number of studies identifıed by both reviews
(i.e., 37 and 33 studies, respectively), the robust effect estimates reported by both reviews, and results from Orwin’s
fail-safe N calculation for the depression symptom improvement outcome, it is highly unlikely that the generalizability
of the fındings is affected by this update approach and by
ending the search in 2009. Further, excluding non-English
studies is reported to have little impact on overall effect
estimates in systematic reviews.76

Conclusion
This systematic review and meta-analysis found that robust
evidence is available and demonstrates the effectiveness of
collaborative care models in the treatment of depressive
disorders. These interventions are applicable in most primary care settings and for most populations to improve a
range of depression outcomes. Organizational changes at
the healthcare-system level are necessary for the successful
implementation of these models so that a coordinated team,
consisting of primary care providers, case managers, and
mental health specialists, can be utilized to improve quality
of depression care. Few variables that substantially moderated the effectiveness of this type of intervention were found,
suggesting that although collaborative care models are composed of several moving parts working simultaneously, it
remains diffıcult to identify and estimate the individual contributions of specifıc components to overall effectiveness.
Collaborative care models also provide a supportive network that encourages patients with depression to take an
active role in their own care, thus constituting a vital resource of social support as these patients seek to initiate and
maintain treatment for depression.
Points of view are those of the authors and do not necessarily
reflect those of the CDC.
AJG is a major stock owner of Healthcare Technology Stystems, Inc.; he consults to Dey Pharma, PGxHealth, Myriad
Genetics, and Zynx Health; and is the principal investigator on
an investigator-initiated grant from Pfızer Pharmaceuticals to
Penn State. No other fınancial disclosures were reported by the
authors of this paper.

References
1. Lopez, AD, Mathers CD, Ezzati M, Jamison DT, Murray CJ. Global and
regional burden of disease and risk factors, 2001: systematic analysis of
population health data. Lancet 2006;367(9524):1747–57.
2. Kessler RC, Chiu WT, Demler O, Merikangas KR, Walters EE. Prevalence, severity, and comorbidity of twelve-month DSM-IV disorders in
the National Comorbidity Survey Replication (NCS-R). Arch Gen
Psychiatry 2005;62(6):617–27.
3. Strine TW, Mokdad AH, Balluz LS, et al. Depression and anxiety in the
U.S: fındings from the 2006 Behavioral Risk Factor Surveillance System. Psychiatr Serv 2008;59(12):1383–90.
4. Substance Abuse and Mental Health Services Administration. Results
from the 2008 National Survey on Drug Use and Health: national
fındings. Rockville MD: Offıce of Applied Studies, 2009. NSDUH Series H-36, HHS Publication No. SMA 09-4434.
5. Greenberg P, Kessler R, Birnbaum H, et al. The economic burden of
depression in the U.S.: how did it change between 1990 and 2000? J Clin
Psychiatry 2003;64(12):1465–75.
6. Young AS, Klap R, Sherbourne CD, Wells KB. The quality of care for
depressive and anxiety disorders in the U.S. Arch Gen Psychiatry
2001;58(1):55– 61.
7. Wagner EH. Chronic disease management: what will it take to improve
care for chronic illness? Eff Clin Pract 1998;1(1):2– 4.
8. Wagner EH, Austin BT, Von Korff M. Organizing care for patients
with chronic illness. Milbank Q 1996;74(4):511– 44.
9. Coleman K, Austin BT, Brach C, Wagner EH. Evidence on the chronic
care model in the new millennium. Health Aff 2009;28(1):75– 85.
10. Wagner EH, Grothaus LC, Sandhu N, et al. Chronic care clinics for
diabetes in primary care. Diabetes Care 2001;24(4):695–700.
11. Katon W, Von Korff M, Lin E, Simon G. Rethinking practitioner roles
in chronic illness: the specialist, primary care physician, and the practice nurse. Gen Hosp Psychiatry 2001;23(3):138 – 44.
12. Bower P, Gilbody S, Richards D, Fletcher J, Sutton A. Collaborative
care for depression in primary care: making sense of a complex intervention: systematic review and meta-regression. Br J Psychiatry
2006;189(6):484 –93.
13. Gilbody S, Bower P, Fletcher J, Richards D, Sutton AJ. Collaborative
care for depression: a cumulative meta-analysis and review of longerterm outcomes. Arch Intern Med 2006;166(21):2314 –21.
14. Briss PA, Zaza S, Pappaioanou M, et al. Developing an evidence-based
Guide to Community Preventive Services—methods. The Task Force on
Community Preventive Services. Am J Prev Med 2000;18(1S):35– 43.
15. Zaza S, Wright-deAguero LK, Briss PA, et al. Data collection instrument and procedure for systematic reviews in the Guide to Community
Preventive Services. Task Force on Community Preventive Services.
Am J Prev Med 2000;18(1S):44 –74.
16. Healthy People 2020. Mental health and mental disorders.
www.healthypeople.gov/hp2020/Objectives/TopicArea.aspx?id⫽34&
TopicArea⫽Mental%20Health%20and%20Mental%20Disorders.
17. U.S. Preventive Services Task Force. Screening for depression in adults:
U.S. Preventive Services Task Force recommendation statement. Ann
Intern Med 2009;151(11):784 –92.
18. U.S. Preventive Services Task Force. Screening and treatment for major
depressive disorder in children and adolescents: U.S. Preventive Services
Task Force recommendation statement. Pediatrics 2009;123(4):1223– 8.
19. American Psychiatric Association. Practice guideline for the treatment
of patients with major depressive disorder, third edition. Am J Psychiatry 2010;167(Oct Suppl).
20. Task Force on Community Preventive Services. The Guide to Community Preventive Services: what works to promote health? New York NY:
Oxford University Press, 2005.
21. Kupfer DJ. Long-term treatment of depression. J Clin Psychiatry
1991;52(S):28 –34.
22. Dobscha SK, Corson K, Hickam DH, Perrin NA, Kraemer DF, Gerrity MS.
Depression decision support in primary care: a cluster randomized trial [see

www.ajpmonline.org

Thota et al / Am J Prev Med 2012;42(5):525–538

23.

24.

25.
26.

27.
28.
29.

30.
31.
32.
33.

34.

35.

36.

37.
38.
39.
40.
41.

42.
43.
44.

45.

46.

47.

comment; summary for patients in Ann Intern Med 2006;145(7):I10; PMID:
17015861]. Ann Intern Med 2006;145(7):477–87.
McMahon L, Foran KM, Forrest SD, et al. Graduate mental health
worker case management of depression in UK primary care: a pilot
study. Br J Gen Pract 2007;57(544):880 –5.
Sharpe M, Strong V, Allen K, et al. Management of major depression in
outpatients attending a cancer centre: a preliminary evaluation of a
multicomponent cancer nurse-delivered intervention. Br J Cancer
2004;90(2):310 –3.
Williams LS, Kroenke K, Bakas T, et al. Care management of poststroke
depression: a randomized, controlled trial. Stroke 2007;38(3):998 –1003.
Spitzer RL, Williams JB, Gibbon M, First MB. The Structured Clinical
Interview for DSM-III-R (SCID): history, validation, and description.
Arch Gen Psychiatry 1992;49:624 –9.
Rush A, Pincus A, First M, et al. Handbook of psychiatric measures.
Washington DC: American Psychiatric Association, 2000.
Lee PW. Concordance between the PHQ-9 and the HSCL-20 in depressed primary care patients. J Affect Disord 2007;99(1-3):139 – 45.
Lowe B, Unutzer J, Callahan CM, Perkins AJ, Kroenke K. Monitoring depression treatment outcomes with the Patient Health Questionnaire-9.
Med Care 2004;42(12):1194–201.
Rush AJ. Report by the ACNP Task Force on response and remission in major
depressive disorder. Neuropsychopharmacology 2006;31(9):1841–53.
National Center for Chronic Disease Prevention and Health Promotion. Health-related quality of life. www.cdc.gov/hrqol/.
Leidy N. Functional status and the forward progress of merry-go-rounds:
toward a coherent analytical framework. Nurs Res 1994;43(4):196 –202.
Wilson IB, Cleary PD. Linking clinical variables with health-related
quality of life: a conceptual model of patient outcomes. JAMA
1995;273(1):59 – 65.
Rapkin B, Weiss E, Chhabra R, et al. Beyond satisfaction: using the
Dynamics of Care assessment to better understand patients’ experiences in care. Health Qual Life Outcomes 2008;6:20.
Cherin DA, Huba GJ, Steinberg J, et al. Satisfaction with services in
innovative managed care programs for groups of traditionally underserved individuals with HIV/AIDS: empirical models. Home Health
Care Serv Q 2001;19(1–2):103–25.
Jacob V, Chattopadhyay SK, Sipe TA, et al. Economics of collaborative
care for management of depressive disorders: a Community Guide
systematic review. Am J Prev Med 2012;42(5):539 – 49.
Hedges LV. Distribution theory for Glass’s estimator of effect size and
related estimators. J Educ Behav Stat 1981;6(2):107–28.
Cooper HM, Hedges LV. The handbook of research synthesis. New
York: Russell Sage Foundation Publications, 1994.
Higgins J, Thompson S. Quantifying heterogeneity in a meta-analysis.
Stat Med 2002;21(11):1539 –58.
Sutton A, Duval S, Tweedie R, Abrams K, Jones D. Empirical assessment of
effect of publication on meta-analyses. BMJ 2000;320(7249):1574 –7.
Borenstein M. Software for publication bias. In: Rothstein H, Sutton A,
Borenstein M, eds. Publication bias in meta-analysis—prevention, assessment and adjustments. New York: John Wiley and Sons, Ltd., 2005.
Orwin RG. A fail-safe N for effect size in meta-analysis. J Educ Behav
Stat 1983;8(2):157–9.
Lipsey M, Wilson D. Practical meta-analysis. Thousand Oaks CA: Sage
Publications, 2001.
Becker B. Combining signifıcance levels. In: Cooper HM, Hedges LV,
eds. The handbook of research synthesis. New York: Russell Sage
Foundation Publications, 1994:215–30.
Greenhouse J, Iyengar S. Sensitivity analysis and diagnostics. In: Cooper HM, Hedges LV, eds. The handbook of research synthesis. New
York: Russell Sage Foundation Publications, 1994:383–98.
Asarnow JR, Jaycox LH, Duan N, et al. Effectiveness of a quality improvement intervention for adolescent depression in primary care clinics: a
randomized controlled trial [see comment]. JAMA 2005;293(3):311–9.
Baldwin R, Pratt H, Goring H, Marriott A, Roberts C. Does a nurse-led
mental health liaison service for older people reduce psychiatric mor-

May 2012

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

537

bidity in acute general medical wards? A randomised controlled trial.
Age Ageing 2004;33(5):472– 8.
Bogner HR, de Vries HF. Integration of depression and hypertension
treatment: a pilot, randomized controlled trial. Ann Fam Med
2008;6(4):295–301.
Ciechanowski P, Wagner E, Schmaling K, et al. Community-integrated
home-based depression treatment in older adults: a randomized controlled trial. JAMA 2004;(13):1569 –77.
Chew-Graham CA, Lovell K, Roberts C, et al. A randomised controlled
trial to test the feasibility of a collaborative care model for the management
of depression in older people. Br J Gen Pract 2007;57(538):364 –70.
Cole MG, McCusker J, Elie M, Dendukuri N, Latimer E, Belzile E.
Systematic detection and multidisciplinary care of depression in older
medical inpatients: a randomized trial. CMAJ 2006;174(1):38 – 44.
Cullum S, Tucker S, Todd C, Brayne C. Effectiveness of liaison psychiatric nursing in older medical inpatients with depression: a randomised
controlled trial. Age Ageing 2007;36(4):436 – 42.
Dietrich AJ, Oxman TE, Williams JW Jr, et al. Re-engineering systems
for the treatment of depression in primary care: cluster randomised
controlled trial [references]. BMJ 2004;329(7466):602.
Dwight-Johnson M, Ell K, Lee P-J. Can collaborative care address
the needs of low-income Latinas with comorbid depression and
cancer? Results from a randomized pilot study. Psychosomatics
2005;46(3):224 –32.
Ell K, Unutzer J, Aranda M, Gibbs NE, Lee P, Xie B. Managing depression in home health care: a randomized clinical trial. Home Health
Care Serv Q 2007;26(3):81–104.
Ell K, Xie B, Quon B, Quinn DI, Dwight-Johnson M, Lee PJ. Randomized controlled trial of collaborative care management of depression among low-income patients with cancer. J Clin Oncol
2008;26(27):4488 –96.
Fortney JC, Pyne JM, Edlund MJ, et al. A randomized trial of
telemedicine-based collaborative care for depression. J Gen Intern
Med 2007;22(8):1086 –93.
Gallo JJ, Bogner HR, Morales KH, Post EP, Lin JY, Bruce ML. The effect
of a primary care practice-based depression intervention on mortality
in older adults: a randomized trial. Ann Intern Med
2007;146(10):689 –98.
Gensichen J, von Korff M, Peitz M, et al.; PRoMPT (PRimary care
Monitoring for depressive Patients Trial). Case management for depression by health care assistants in small primary care practices. Ann
Intern Med 2009;151(6):369 –78.
Joubert J, Joubert L, Reid C, et al. The positive effect of integrated care
on depressive symptoms in stroke survivors. Cerebrovasc Dis
2008;26(2):199 –205.
Knight MM, Houseman EA. A collaborative model for the treatment of
depression in homebound elders. Issues Ment Health Nurs
2008;29(9):974 –91.
Ludman EJ, Simon GE, Grothaus LC, Luce C, Markley DK, Schaefer J.
A pilot study of telephone care management and structured disease
self-management groups for chronic depression. Psychiatr Serv
2007;58(8):1065–72.
Oslin DW, Thompson R, Kallan MJ, et al. Treatment effects from
UPBEAT: a randomized trial of care management for behavioral health
problems in hospitalized elderly patients. J Geriatr Psychiatry Neurol
2004;17(2):99 –106.
Reiss-Brennan B, Briot P, Daumit G, Ford D. Evaluation of “depression
in primary care” innovations [references]. Adm Policy Ment Health
2006;33(1):86 –91.
Richards DA, Lovell K, Gilbody S, et al. Collaborative care for depression in UK primary care: a randomized controlled trial. Psychol Med
2008;38(2):279 – 87.
Rollman BL, Belnap BH, LeMenager MS, et al. Telephone-delivered
collaborative care for treating post-CABG depression: a randomized
controlled trial. JAMA 2009;302(19):2095–103.

538

Thota et al / Am J Prev Med 2012;42(5):525–538

67. Schrader G, Cheok F, Hordacre AL, Marker J, Wade V. Effect of
psychiatry liaison with general practitioners on depression severity in
recently hospitalised cardiac patients: a randomised controlled trial.
Med J Aust 2005;182(6):272– 6.
68. Simon GE, Katon WJ, Lin EH, et al. Cost-effectiveness of systematic
depression treatment among people with diabetes mellitus [see comment]. Arch Gen Psychiatry 2007;64(1):65–72.
69. Simon GE, Ludman EJ, Operskalski BH. Randomized trial of a telephone care management program for outpatients starting antidepressant treatment. Psychiatr Serv 2006;57(10):1441–5.
70. Smit A, Kluiter H, Conradi HJ, et al. Short-term effects of enhanced
treatment for depression in primary care: results from a randomized
controlled trial. Psychol Med 2006;36(1):15–26.
71. Stiefel F, Zdrojewski C, Bel Hadj F, et al. Effects of a multifaceted
psychiatric intervention targeted for the complex medically ill: a
randomized controlled trial. Psychother Psychosom 2008;77(4):
247–56.

72. Strong V, Waters R, Hibberd C, et al. Management of depression for
people with cancer (SMaRT oncology 1): a randomised trial. Lancet
2008;372(9632):40 – 8.
73. Wang PS, Simon GE, Avorn J, et al. Telephone screening, outreach, and
care management for depressed workers and impact on clinical and
work productivity outcomes: a randomized controlled trial. JAMA
2007;298(12):1401–11.
74. Wells KB, Tang L, Miranda J, Benjamin B, Duan N, Sherbourne CD.
The effects of quality improvement for depression in primary care at
nine years: results from a randomized, controlled group-level trial.
Health Serv Res 2008;43(6):1952–74.
75. Hedrick SC, Chaney EF, Felker B, et al. Effectiveness of collaborative
care depression treatment in Veterans’ Affairs primary care. J Gen
Intern Med 2003;18(1):9 –16.
76. Juni P, Holenstein F, Sterne J, Bartlett C, Egger M. Direction and
impact of language bias in meta-analyses of controlled trials: empirical
study. Int J Epidemiol 2002;31(1):115–23.

www.ajpmonline.org

File Type	application/pdf
File Modified	2012-04-16
File Created	2012-04-05