 |
|
Two-Year Effects of Quality Improvement Programs on Medication Management for Depression
Jürgen Unützer, MD, MPH;
Lisa Rubenstein, MD, MSPH;
Wayne J. Katon, MD;
Lingqi Tang, PhD;
Naihua Duan, PhD;
Isabel T. Lagomasino, MD;
Kenneth B. Wells, MD, MPH
Arch Gen Psychiatry. 2001;58:935-942.
ABSTRACT
| |
Background Significant underuse of evidence-based treatments for depression persists
in primary care. We examined the effects of 2 primary care–based quality
improvement (QI) programs on medication management for depression.
Methods A total of 1356 patients with depressive symptoms (60% with depressive
disorders and 40% with subthreshold depression) from 46 primary care practices
in 6 nonacademic managed care organizations were enrolled in a randomized
controlled trial of QI for depression. Clinics were randomized to usual care
or to 1 of 2 QI programs that involved training of local experts who worked
with patients' regular primary care providers (physicians and nurse practitioners)
to improve care for depression. In the QI-medications program, depression
nurse specialists provided patient education and assessment and followed up
patients taking antidepressants for up to 12 months. In the QI-therapy program,
depression nurse specialists provided patient education, assessment, and referral
to study-trained psychotherapists.
Results Participants enrolled in both QI programs had significantly higher rates
of antidepressant use than those in the usual care group during the initial
6 months of the study (52% in the QI-medications group, 40% in the QI-therapy
group, and 33% in the usual care group). Patients in the QI-medications group
had higher rates of antidepressant use and a reduction in long-term use of
minor tranquilizers for up to 2 years, compared with patients in the QI-therapy
or usual care group.
Conclusions Quality improvement programs for depression in which mental health specialists
collaborate with primary care providers can substantially increase rates of
antidepressant treatment. Active follow-up by a depression nurse specialist
in the QI-medications program was associated with longer-term increases in
antidepressant use than in the QI model without such follow-up.
INTRODUCTION
DEPRESSION is a major cause of disability worldwide1
and is common in primary care.2, 3
Despite dissemination of practice guidelines for depression,4, 5, 6
significant underuse of evidence-based treatments for depression persists
in primary care.7, 8 Quality improvement
(QI) efforts in primary care have been shown to increase rates of care and
clinical outcomes for major depression for up to 8 months,9, 10, 11, 12
but improvements in care were not sustained at longer-term follow-ups.13, 14 We describe the impact of 2 QI interventions
on the use of antidepressant medications and minor tranquilizers over 2 years.
This article is based on a randomized quasi experiment of an evidence-based
QI intervention that was conducted in 46 nonacademic managed primary care
practices in 5 states. Clinics were randomized to usual care (UC) or to 1
of 2 QI programs.15, 16 To mirror
the diverse clinical status of patients in community primary care settings,
the study included depressed patients who met the research diagnostic criteria
for major depression or dysthymic disorder and those with subthreshold depression.15
In this article, we follow the Institute of Medicine's formulation of
quality of care17 and examine whether problems
with underuse (lack of use of maintenance antidepressants by patients at high
risk for relapse) or overuse (inappropriate long-term use of minor tranquilizers
for depression) were affected by our interventions. We expected that the QI-medications
(QI-Meds) program would lead to higher rates of antidepressant use than the
QI-therapy program because of additional resources devoted to following up
patients taking antidepressant medications.
PATIENTS AND METHODS
STUDY DESIGN AND STUDY SITES
Partners in Care is a group-level randomized controlled trial carried
out in 6 nonacademic managed care organizations in geographically diverse
areas of the country.15 Forty-six of 48 clinics
belonging to these 6 organizations and 181 of 183 primary care providers (physicians
and nurse practitioners) agreed to participate. Clinics were grouped into
27 clusters that were matched into 9 blocks of 3 clusters each based on patient
demographics, clinician specialty, and distance to specialty mental health
providers (psychiatrists, psychologists, or psychotherapists). Within each
block, clinic clusters were randomly assigned to UC or to 1 of 2 QI programs
(QI-Meds or QI-therapy). Participating organizations included prepaid, staff-model,
mixed fee-for-service/prepaid, and network-model group practices and rural,
managed, public health clinics.16
Study staff screened consecutive patients in the waiting rooms of participating
clinics during a 6-month period between June 1996 and March 1997. Eligible
patients were aged 18 years or older, intended to use the clinic as their
main source of medical care in the coming year, and were determined to have
depression by a 6-item screening instrument that included the "stem" items
for major depression and dysthymic disorder from the 12-month Composite International
Diagnostic Interview (CIDI; edition 2.1), and items assessing the presence
of depressive symptoms in the past month. Based on research diagnoses obtained
by the full affective disorders section of the CIDI, the positive predictive
value of this screener for depressive disorder (major depression or dysthymic
disorder) was 55%.15 Patients who had an immediate
medical emergency, did not speak English or Spanish, or did not have either
insurance or a public-pay arrangement that covered the study interventions
were excluded.
A total of 44 052 persons were approached in clinic waiting rooms,
but 10 120 were not eligible for screening, mainly because they were
not patients of participating primary care providers. Of the 27 332 patients
completing the screener, 3918 were potentially eligible, but many left the
clinic before completing the multistage enrollment process. Of the 2417 patients
available to confirm insurance eligibility, 241 had ineligible health insurance.
Of those who read the informed consent, 1356 enrolled. The remaining 21% either
refused to participate or left the clinic. The enrolled sample includes 443
patients in the UC group, 424 in the QI-Meds group, and 489 in the QI-therapy
group (Table 1). We controlled
for differences in sex, educational level, and depression diagnosis in the
analyses.
|
|
|
|
Table 1. Characteristics of the 1356 Subjects: Overall and by Intervention
Conditions*
|
|
|
INTERVENTIONS
Our QI approach used a combination of expert intervention design, local
managed care organization involvement, and provider behavior change strategies
that have been described in detail previously.16
Both QI interventions had a common core of physician education and patient
screening, assessment, and education by a depression nurse specialist (DNS).
Physician and nurse education followed a treatment manual adapted from the
Agency for Healthcare Research and Quality (formerly the Agency for Health
Care Policy and Research) treatment guidelines for depression in primary care.18 Providers in both groups were encouraged to initiate
treatment with antidepressants or psychotherapy for patients who met the diagnostic
criteria for major depression or dysthymic disorder. Patients and providers
were free to use either antidepressant medications or psychotherapy, both,
or neither. In addition to this common core, each intervention provided clinics
with a unique set of resources. The QI-Meds program provided participants
in clinics that had been randomized to this intervention access to a DNS who
offered to support ongoing antidepressant treatment by the patients' primary
care providers for either 6 or 12 months (randomly assigned at the patient
level). This follow-up focused on increasing adherence to guidelines for appropriate
use of antidepressant medications.18 In the
QI-therapy program, DNSs assisted all patients in QI-therapy clinics whose
clinicians determined that psychotherapy was appropriate, with a referral
to a study-trained psychotherapist to offer individual or group cognitive-behavioral
therapy at a reduced copayment.16 Prior analyses
showed that the DNSs attempted to contact 96% of eligible patients with depression,
and completed an initial assessment visit with 73% (76% in the QI-Meds group
and 71% in the QI-therapy group). About 55% of eligible participants completed
all visits with the DNS. The mean number of nurse follow-up contacts was 1.8
in the QI-therapy group and 5.1 in the QI-Meds group.16
Usual care clinics were mailed copies of the Agency for Healthcare Research
and Quality provider guidelines for depression in primary care.4
Patients in these clinics had access to all usually available primary care
or specialty mental health treatments, but no extra resources were provided
by the study.
OUTCOMES EXAMINED
Enrolled patients were asked to complete a telephone interview that
included the affective disorders section of the CIDI, information on comorbid
anxiety disorders, and economic information at baseline. Patients also completed
a self-administered mail survey at baseline and at 6, 12, 18, and 24 months.
The surveys assessed the use of psychotherapy or prescription medications
during the prior 6 months. We updated dosage recommendations from the Agency
for Healthcare Research and Quality treatment guidelines4
for newer antidepressants using a consensus panel of 10 academic expert psychiatrists,
and used the low end of the dosage recommendations as "minimum recommended
daily doses" (Table 2).
|
|
|
|
Table 2. Guideline-Level Antidepressant Doses for Patients With Depressive
Disorders
|
|
|
Because the study enrolled a clinically diverse group of patients with
major depression, dysthymic disorder, and subthreshold depression for whom
treatment guidelines are less clear, we used 2 approaches to examine care
for depression. For the entire sample of patients (including those with subthreshold
depression), we describe any use of antidepressant medications (either 1 month
of use during the prior 6 months or any use in the prior 30 days) or any counseling
(at least 1 visit) as indicators of depression treatment.
For the 562 subjects who met the diagnostic criteria for DSM-IV major depression or dysthymic disorder at baseline and who were
at high risk of relapse based on dysthymic disorder or a history of 2 or more
episodes of depression, we examined 3 additional quality indicators: antidepressant
use at minimum daily recommended doses (Table 2) for at least 25 of the past 30 days, for at least 2 of
the past 6 months, and for at least 6 months in the past year (data for this
indicator are not available at 24 months).
We also examined long-term minor tranquilizer use (use for >3 of the
past 6 months). Our intervention materials recommended against such long-term
use in patients with depression who did not have comorbid anxiety disorders
because of the lack of efficacy data for this group of patients and because
of substantial costs and risks associated with the long-term use of minor
tranquilizers.
STATISTICAL ANALYSES
Our analyses examine the effects of the 2 QI models (randomly assigned
at the clinic level) compared with UC on the use of medications by patients
enrolled in the respective clinics. We used all 1092 subjects who had baseline
data and at least one follow-up data point and who did not meet the CIDI criteria
for bipolar disorder for the analyses. We performed multilevel analyses, testing
the primary hypotheses that the intervention conditions increased rates of
treatment (as previously specified) significantly more than UC. We tested
secondary hypotheses about the targeting of depression treatments by performing
analyses that interacted the intervention status with baseline disorder status,
treatment preferences, and prior treatment. Because both interventions recommended
against the long-term use of minor tranquilizers in patients without comorbid
anxiety disorders, we also tested the hypothesis that patients in the QI clinics
had lower rates of potentially inappropriate minor tranquilizer use over time.
For each dependent variable, we fitted 3-level mixed-effects logistic
regression models using follow-up data at 6, 12, 18, and 24 months with regression
adjustment for baseline depression treatments in the past 6 months, accounting
for the multilevel data structure with patients nested within clinics and
repeated measurements nested within patients. We treated time as a categorical
variable, and examined the fixed effects for time, intervention condition,
and their interactions. To account for the intraclass correlation expected
in the data, we specified random effects at the clinic and patient level,
including random intercepts and random slopes for the difference among waves.
At the patient level, we specified the covariance structure among the random
effects as the most general unstructured model. At the clinic level, we specified
a more restrictive variance component model because of the limited degrees
of freedom available. Our model specification is analogous to the models used
by others.19, 20, 21
We also included several covariates for additional adjustment: age, squared
age, sex, educational level (less than high school, completed high school,
some college, completed college, or more), chronic medical conditions from
a total of 19 (0, 1, 2, or  3), depressive disorder determined by the CIDI,
an indicator of having at least 2 prior depressive episodes, comorbid anxiety
disorders (CIDI), likely problem drinking determined by the Alcohol Use Disorders
Identification Test, baseline preferences for treatment (antidepressant medications,
counseling, nothing, or wait), study site, and a summary variable of household
wealth modeled after the Health and Retirement Survey.
To test intervention effects at each time point (months 6, 12, 18, and
24), we conducted pairwise 2-sided t tests comparing
QI-Meds vs UC, QI-therapy vs UC, and QI-Meds vs QI-therapy.
To present the effect size of intervention effects, we calculated standardized
predictions for each outcome studied.22 In
deriving these predictions, regression parameters and each individual's actual
covariate values other than intervention status are used to derive 3 predicted
values for each individual, first as a QI-Meds group subject, then as a QI-therapy
group subject, and then as a UC group subject. Predictions under the QI-Meds
scenario are then averaged across the entire analytic sample to obtain an
overall assessment of the QI-Meds outcome; the procedure is then repeated
for the other conditions. This procedure thus standardizes the comparisons
to the characteristics of the full analytic sample.
We used an extended hot deck multiple imputation technique that modifies
the predictive mean matching method23 to impute
missing covariates. (Outcome variables were not imputed.) Instead of filling
in a single value for each missing value, we used the multiple imputation
strategy of Rubin24 to create 5 imputed data
sets. Each of 5 complete data sets was then analyzed using standard complete-data
methods. The predictions across 5 imputed data sets were combined by averaging,
and SEs were derived using the Rubin method to combine within-imputation variability
and between-imputation variability.
RESULTS
TREATMENTS FOR DEPRESSION
There were no significant baseline differences in the use of depression
treatments (antidepressants or psychotherapy) by intervention status (Table 3). At baseline, the rate of any
treatment was somewhat higher in the QI-therapy group than in the UC group,
and we controlled for this in our analyses.
|
|
|
|
Table 3. Use of Depression Treatments by 1092 Depressed Participants*
|
|
|
During the initial 6 months of the study, antidepressant use was greater
in the QI-Meds and QI-therapy groups than in the UC group. Patients in the
QI-Meds group were more likely than patients in the UC group to report antidepressant
use at 6, 12, 18, and 24 months, but the difference was only statistically
significant at 6 and 12 months. Patients in the QI-Meds group had significantly
higher rates of antidepressant use than those in the QI-therapy group at 6,
12, and 24 months. Participants in the QI-Meds group who were randomly assigned
to nurse case management for 12 months did not have higher rates of antidepressant
use than those who were assigned to case management for 6 months (data not
shown).
As a context for our findings on antidepressant use, we also examined
rates of any depression treatment (any use of antidepressants or psychotherapy)
in the prior 6 months. At 6 months, patients in the QI-Meds and QI-therapy
groups reported higher rates of any depression treatment than those in the
UC group. Those in the QI-Meds group had consistently higher rates of any
depression treatment than those in the UC or QI-therapy group, but the difference
was only statistically significant at 6 and 12 months. Further detail on counseling
treatments is available elsewhere.16, 17
We observed significant interaction effects of intervention type with
disorder status (F = 4.09, P = .02) and intervention
type with prior use of counseling (F = 7.49, P<.001),
indicating greater targeting of antidepressants to patients with depressive
disorders or those who underwent prior counseling in the 2 intervention groups
than in the UC group. For example, at the 6-month follow-up, between 26% (UC
group) and 38% (QI-Meds group) of the participants who did not meet the CIDI
criteria for major depression or dysthymic disorder at baseline reported using
antidepressants, compared with 37% (UC group) to 62% (QI-Meds group) of those
who had depressive disorders at baseline.
Figure 1 shows standardized predictions of the proportion of patients
who used antidepressants during the 2-year study period stratified by baseline
antidepressant use. These estimates were derived from the mixed-effects logistic
regression model that included interactions of intervention type with prior
treatment and disorder status.
|
|
|
|
Proportion of patients who used antidepressants in the prior 6 months
stratified by baseline antidepressant use. QI indicates quality improvement.
|
|
|
TYPES OF ANTIDEPRESSANTS USED
At baseline, selective serotonin reuptake inhibitors accounted for about
60% of the antidepressants used in all 3 groups; other newer antidepressants,
such as bupropion hydrochloride, nefazodone hydrochloride, or venlafaxine
hydrochloride, accounted for about 20%; tertiary amine tricyclic antidepressants
accounted for about 15%; and secondary amine tricyclic antidepressants accounted
for about 5%. Monoamine oxidase inhibitors and other antidepressants, such
as amoxapine or maprotiline, represented less than 1% of all antidepressants
used. During the 2 years of follow-up, there was a slight increase in the
use of newer antidepressants other than selective serotonin reuptake inhibitors
in all 3 groups and a corresponding decrease in the use of tricyclic antidepressants.
Between 5% and 9% of the participants used more than 1 antidepressant during
each 6-month period. There were no substantial intervention group differences
in the types of antidepressants used over time.
ADEQUACY OF ANTIDEPRESSANT USE AMONG SUBJECTS WITH DEPRESSION AT HIGH
RISK FOR RELAPSE
Most patients (562 of 674) who met the criteria for major depression
or dysthymic disorder at baseline were at high risk for relapse, as defined
by current dysthymic disorder or a history of 2 or more episodes of depression.4 At baseline, patients in the QI-therapy group at high
risk for relapse reported significantly greater antidepressant use than patients
in the QI-Meds or UC group (Table 4),
and we controlled for these baseline differences in the analyses.
|
|
|
|
Table 4. Antidepressant Use by the 562 Participants With Depression
at High Risk for Relapse*
|
|
|
Rates of guideline-level antidepressant use, as defined by our 3 quality
indicators, were significantly greater in the QI-Meds group than in the UC
group at 6, 12, and 18 months but not at 24 months. Appropriate antidepressant
use was consistently higher in the QI-Meds group than in the QI-therapy group,
but the differences were not always statistically significant.
LONG-TERM USE OF MINOR TRANQUILIZERS
The proportion of subjects using minor tranquilizers for more than 3
months at baseline ranged from 7.1% (QI-therapy group) to 9.2% (UC group).
At the 24-month follow-up, 6.2% of the patients in the QI-Meds group reported
long-term use of minor tranquilizers compared with 10.0% in the QI-therapy
group (t = -2.01, P
= .04) and 10.3% in the UC group (t = -1.87, P = .06). A similar picture emerged when we examined the
use of minor tranquilizers for 3 or more months without a concurrent prescription
of an antidepressant medication. Over 2 years, this rate declined from 4.6%
to 2.5% in the QI-Meds group, but it remained relatively stable in the QI-therapy
(4%-6%) and the UC (4%-7%) groups.
There was no significant interaction between anxiety disorder status
and intervention group, meaning that the interventions did not increase the
targeting of long-term minor tranquilizers to those who had comorbid anxiety
disorders.
COMMENT
Our results suggest that QI interventions that combine key components
from established chronic care models,25, 26
such as clinician education, patient education, case management, and specialist
involvement in primary care, can lead to relatively long-term increases in
antidepressant use in managed primary care practices. Prior studies27, 28 document that less intensive interventions,
such as provider feedback, do not lead to persistent increases in depression
treatment or improvement in clinical outcomes. The necessity for a more intensive
model of care for changing antidepressant use is consistent with the literature29, 30, 31 on health care provider
behavior in general, which indicates that while 1-step behaviors such as ordering
a mammogram have responded to computer feedback, conditions requiring sustained
patient and provider behavior change have not.
While short-term antidepressant use was increased by the QI-Meds and
QI-therapy interventions, longer-term increases in the rates of medication
use were limited to the QI-Meds group. The increased antidepressant use in
the QI-therapy and QI-Meds groups at 6 months implies that intervention components
common to the 2 interventions improved antidepressant use in the short-term.
These include provider education, collaboration of primary care providers
with DNSs, assessment and education of patients by the DNS, and encouragement
of the use of antidepressant medications or psychotherapy for patients with
major depression or dysthymic disorder. The finding of greater antidepressant
use in the QI-Meds group at 12, 18, and 24 months implies that the addition
of at least 6 months of active follow-up by a DNS (available in QI-Meds practices
only) is associated with relatively long-term increases in the use of antidepressants.
Contrary to our expectations, patients in the QI-Meds group who were in the
12-month case management group had no greater use of antidepressants than
those who were in the 6-month case management group over time. This could
be because 12-month case management has no additional effect compared with
6 months or because most patients eligible for the 12-month follow-up did
not use it for the full available duration.16
Future studies should examine if even shorter periods of follow-up (ie, 3
months) can achieve similar effects on antidepressant use or if a more fully
implemented longer-term follow-up would lead to even greater long-term increases
in medication use.
Earlier trials9, 10, 11
of collaborative care for depression included patients who had already started
taking antidepressants. We found that patients with and without prior antidepressant
treatment had increases in their rates of care, suggesting that QI interventions
for depression are effective in increasing antidepressant use in both groups.
While treatment rates were between 50% and 75% higher in the QI groups than
in the UC group, it is important to emphasize that only about 50% of patients
who met the diagnostic criteria for major depression ordysthymic disorder
at baseline received appropriate antidepressant management during the first
6 months of the study. There is, thus, substantial room for improvement. Our
interventions were less intensive than most previous evaluations of collaborative
care for depression and were self-applied by usual nonacademic primary care
practices to a diverse population of patients, without direct involvement
of researchers in patient care. We expect that interventions to improve care
for depression will show smaller effects in actual use than in carefully controlled
trials because of the greater diversity of patients and primary care providers
involved. The observed improvement in this study may mirror more closely than
previous studies the impacts that might be achievable in a broad-scale dissemination
of QI for depression. If these results held true in large-scale use, the health
impact of such QI on a population level could be substantial.
Our sample included several patients who did not meet the diagnostic
criteria for depressive disorders, and the literature does not consistently
support the efficacy of antidepressants or psychotherapy for subthreshold
depression. Our educational materials encouraged clinicians to recognize subthreshold
depression and to target antidepressant medications and full doses of cognitive-behavioral
therapy to those with major depression or dysthymic disorder. Patients with
depression who were at risk for relapse had the highest rates of antidepressant
use in all 3 groups, and patients at risk for relapse in the QI-Meds group
had consistently higher rates of medication use than patients in the QI-therapy
or UC group. Our study was not designed to determine the need for or effectiveness
of long-term antidepressant use for different types of patients. Future studies
should, thus, focus increased attention on the need for long-term follow-up
and maintenance therapy for diverse populations of patients with depression.
Our educational materials specifically recommended the use of secondary
amine tricyclic antidepressants or selective serotonin reuptake inhibitors
as "first-line" antidepressant medications and pointed out that both types
of medications are equally efficacious. Despite this recommendation, secondary
amine tricyclic agents accounted for 5% or less of the antidepressants used,
and their use declined during the 2 years of the study. Selective serotonin
reuptake inhibitors and other newer antidepressants made up about 80% of all
antidepressants used, without substantial differences among intervention groups.
The greater use of newer antidepressants that can be easier to titrate and
tolerate32, 33, 34
than older antidepressants represents a substantial change from earlier studies7, 32, 34 of depression management
in primary care.
The availability of newer antidepressants that are better tolerated
and indicated for depression and anxiety disorders may also account for relatively
low rates of long-term minor tranquilizer use compared with earlier studies.7 Minor tranquilizer use was particularly low in participants
without comorbid anxiety disorders, indicating that primary care providers
targeted these medications to patients who were more likely to benefit from
them. We discouraged the long-term use of minor tranquilizers in our educational
materials, and we found substantial reductions in minor tranquilizer use in
the QI-Meds group compared with the QI-therapy or UC group. Thus, it appears
possible for QI interventions to simultaneously correct problems related to
potential underuse of appropriate antidepressant medications and overuse of
inappropriate long-term minor tranquilizers.
While we sampled patients from 6 diverse organizations across the country,15 our sample has limited generalizability to primary
care in general. Our ability to generalize is further limited by a relatively
high nonparticipation rate. We are also limited by our reliance on self-report
data for use of medications and counseling. It is possible that patients underreport
such treatments, but an earlier study by Katon and colleagues9
showed that self-report of antidepressant use during the past 30 days (one
of our quality indicators) was strongly associated with automated pharmacy
data.
Our models may somewhat understate the true intervention effects because,
over time, the need for treatment may have decreased more in the intervention
groups than in the UC group.15 An alternative
method to examine the long-term effects of QI on antidepressant prescribing
would have been to examine antidepressant use in a new cohort of patients
with depression. It is also possible that more patients in the QI-therapy
group had sustained clinical benefits that made it less necessary for them
to undergo long-term antidepressant treatment. However, almost half of our
participants were at high risk for relapse and, thus, candidates for long-term
maintenance antidepressant treatment.18
In conclusion, implementation of QI for depression by diverse managed
primary care practices can substantially increase rates of antidepressant
treatment for depression. The inclusion of at least 6 months of active nurse
care management leads to larger and longer-term increases in antidepressant
use than does the collaborative care model without such follow-up.
AUTHOR INFORMATION
Accepted for publication April 19, 2001.
This study was supported by grant R01-HS08349 from the Agency for Healthcare
Research and Quality, Rockville, Md; grant P50 MH54623 from the National Institute
of Mental Health, Rockville; and grant 96-42901A-HE from the John D. and Catherine
T. MacArthur Foundation, Chicago, Ill.
We thank Robert Bell, PhD, Tom Belin, PhD, and Daniel McCaffrey, PhD,
for their statistical advice; Maga Jackson-Triche, MD, MSHS, for her assistance
with the design and implementation of the interventions; and Bernadette Benjamin,
MS, for her meticulous programming support. This study is a sister study to
the National Institute of Mental Health Cooperative Agreement to Test Depression
Practice Guidelines (Lisa Rubenstein, MD, MSHS, Kathryn Rost, PhD, and Daniel
Ford, MD, MPH, principal investigators). We acknowledge the following participating
managed care organizations that provided access to their expertise and patients,
implemented interventions, and provided in-kind resources: Allina Medical
Group, Minneapolis, Minn; Patuxent Medical Group, Columbia, Md; Humana Health
Care Plans, San Antonio, Tex; MedPartners, Los Angeles, Calif; PacifiCare
of Texas, San Antonio; and Valley-Wide Health Services, San Luis Valley, Colo.
We also acknowledge their internal behavioral health organizations and participating
contract behavioral health organizations: Alamo Mental Health Group, San Antonio;
San Luis Valley Mental Health/Colorado Health Networks, San Luis Valley; and
Greenspring Mental Health Services, Columbia. Finally, we acknowledge the
clinicians and patients who contributed their time and efforts to this study.
From the Neuropsychiatric Institute, University of California, Los
Angeles (Drs Unützer, Tang, Duan, and Wells); VA Greater Los Angeles
Healthcare System (Dr Rubenstein) and the Department of Psychiatry, Charles
R. Drew University (Dr Lagomasino), Los Angeles, Calif; RAND, Santa Monica,
Calif (Drs Rubenstein and Wells); and the Department of Psychiatry, University
of Washington, Seattle (Dr Katon).
Corresponding author and reprints: Jürgen Unützer, MD,
MPH, Center for Health Services Research, Neuropsychiatric Institute, University
of California, Los Angeles, 10920 Wilshire Blvd, Suite 300, Los Angeles, CA
90024 (e-mail: unutzer{at}ucla.edu).
REFERENCES
| |
1. Murray CJ, Lopez AD. The Global Burden of Disease: A Comprehensive Assessment
of Mortality and Disability From Disease, Injuries, and Risk Factors in 1990
and Projected to 2020. Boston, Mass: Harvard School of Public Health, on behalf of the World
Health Organization and the World Bank; 1996.
2. Katon W, Schulberg H. Epidemiology of depression in primary care. Gen Hosp Psychiatry. 1992;14:237-247.
FULL TEXT
|
ISI
| PUBMED
3. Spitzer RL, Williams JB, Kroenke K, Linzer M, deGruy III FV, Hahn SR, Brody D, Johnson JG. Utility of a new procedure for diagnosing mental disorders in primary
care: the PRIME MD 1000 study. JAMA. 1994;272:1749-1756.
FREE FULL TEXT
4. Depression Guidelines Panel. Depression in Primary Care: Treatment of Major Depression. Rockville, Md: US Dept of Health and Human Services; 1993. AHRQ (formerly
AHCPR), US Public Health Services, publication 93-0551.
5. Schulberg HC, Katon WJ, Simon GE, Rush AJ. Treating major depression in primary care practice: an update of the
Agency for Health Care Policy and Research Practice Guidelines. Arch Gen Psychiatry. 1998;55:1121-1127.
FREE FULL TEXT
6. Schulberg HC, Katon WJ, Simon GE, Rush AJ. Best clinical practice: guidelines for managing major depression in
primary medical care. J Clin Psychiatry. 1999;60(suppl 7):19-26; discussion, 27-28.
7. Wells KB, Katon W, Rogers B, Camp P. Use of minor tranquilizers and antidepressant medications by depressed
outpatients: results from the Medical Outcomes Study. Am J Psychiatry. 1994;151:694-700.
FREE FULL TEXT
8. Wells KB, Schoenbaum M, Unützer J, Lagomasino I, Rubenstein L. Quality of care for primary care patients with depression in managed
care. Arch Fam Med. 1999;8:529-536.
FREE FULL TEXT
9. Katon W, Von Korff M, Lin E, Walker E, Simon G, Bush T, Robinson P, Russo J. Collaborative management to achieve treatment guidelines: impact on
depression in primary care. JAMA. 1995;273:1026-1031.
FREE FULL TEXT
10. Katon W, Robinson P, Von Korff M, Lin E, Bush T, Ludman E, Simon G, Walker E. A multifaceted intervention to improve treatment of depression in primary
care. Arch Gen Psychiatry. 1996;53:924-932.
FREE FULL TEXT
11. Katon W, Von Korff M, Lin E, Simon G, Walker E, Unützer J, Bush T, Russo J, Ludman E. Stepped collaborative care for primary care patients with persistent
symptoms of depression: a randomized trial. Arch Gen Psychiatry. 1999;56:1109-1115.
FREE FULL TEXT
12. Schulberg HC, Block MR, Madonia MJ, Scott CP, Rodriguez E, Imber SD, Perel T, Lave J, Houck PR, Coulehan JL. Treating major depression in primary care practice: eight-month clinical
outcomes. Arch Gen Psychiatry. 1996;53:913-919.
FREE FULL TEXT
13. Lin EH, Katon WJ, Simon GE, Von Korff M, Bush TM, Rutter CM, Saunders KW, Walker EA. Achieving guidelines for the treatment of depression in primary care:
is physician education enough? Med Care. 1997;35:831-842.
FULL TEXT
|
ISI
| PUBMED
14. Lin E, Simon GE, Katon WJ, Russo JE, Von Korff M, Bush TM, Ludman EJ, Walker E. Can enhanced acute phase treatment of depression improve long-term
outcomes? a report of randomized trials in primary care. Am J Psychiatry. 1999;156:643-645.
FREE FULL TEXT
15. Wells KB, Sherbourne C, Schoenbaum M, Duan N, Meredith L, Unützer J. Impact of disseminating quality improvement programs for depression
in managed primary care: a randomized controlled trial. JAMA. 2000;283:212-220.
FREE FULL TEXT
16. Rubenstein LV, Jackson-Triche M, Unützer J, Miranda J, Minnium K, Pearson ML, Wells KB. Evidence-based care for depression in managed primary care practices. Health Aff (Millwood). 1999;18:89-105.
ABSTRACT
17. Chassin MR, Galvin RW. The urgent need to improve health care quality: Institute of Medicine
National Roundtable on Health Care Quality. JAMA. 1998;280:1000-1005.
FREE FULL TEXT
18. Rubenstein LV, Unützer J, Miranda J, Katon W, Wieland M, Jackson-Triche M, Minnium K, Mulrow C, Wells K. Clinician Guide to Depression and Management in Primary
Care Settings. Santa Monica, Calif: RAND; 1996.
19. Gibbons RD, Bock RD. Trend in correlated proportions. Psychometrika. 1987;52:113-124.
FULL TEXT
|
ISI
20. Gibbons RD, Hedeker D. Application of random-effects probit regression models. J Consult Clin Psychol. 1994;62:285-296.
FULL TEXT
|
ISI
| PUBMED
21. Hedeker D, Gibbons RD. A random-effects ordinal regression model for multilevel analysis. Biometrics. 1994;50:933-944.
FULL TEXT
|
ISI
| PUBMED
22. Graubard BI, Korn EL. Predictive margins with survey data. Biometrics. 1999;55:652-659.
FULL TEXT
|
ISI
| PUBMED
23. Little RJ. Missing-data adjustments in large surveys. J Business Econ Stat. 1988;6:287-301.
24. Rubin DB. Multiple Imputation for Nonresponse in Surveys. New York, NY: John Wiley & Sons Inc; 1987.
25. Wagner EH, Austin BT, Von Korff M. Improving outcomes in chronic illness. Manag Care Q. 1996;4:12-25.
PUBMED
26. Katon W, Von Korff M, Lin E, Unützer J, Simon G, Walker E, Ludman E, Bush T. Population-based care for depression: effective disease management
strategies to decrease prevalence. Gen Hosp Psychiatry. 1997;19:169-178.
FULL TEXT
|
ISI
| PUBMED
27. Simon GE, Von Korff M, Rutter C, Wagner E. Randomized trial of monitoring, feedback, and management of care by
telephone to improve treatment of depression in primary care. BMJ. 2000;320:550-554.
FREE FULL TEXT
28. Kroenke K, Taylor-Vaisey A, Dietrich AJ, Oxman TE. Intervention to improve provider diagnosis and treatment of mental
disorders in primary care. Psychosomatics. 2000;41:39-52.
FREE FULL TEXT
29. Davis DA, Thomson MA, Oxman AD, Haynes RB. Changing physician performance: a systematic review of the effect of
continuing medical education strategies. JAMA. 1995;274:700-705.
FREE FULL TEXT
30. Rubenstein LV, Mittman BS, Yano EM, Mulrow CD. From understanding health care provider behavior to improving health
care: the QUERI framework for quality improvement: Quality Enhancement Research
Initiative. Med Care. 2000;38(suppl 1):I129-I141.
31. Oxman AD, Thomson MA, Davis DA, Haynes RB. No magic bullets: a systematic review of 102 trials of interventions
to improve professional practice. CMAJ. 1995;153:1423-1431.
ABSTRACT
32. Simon GE, Von Korff M, Wagner EH, Barlow W. Patterns of antidepressant use in community practice. Gen Hosp Psychiatry. 1993;15:399-408.
FULL TEXT
|
ISI
| PUBMED
33. Simon GE, Heiligenstein J, Revicki D, Von Korff M, Katon WJ, Ludman E, Grothaus L, Wagner E. Long-term outcomes of initial antidepressant drug choice in a "real
world" randomized trial. Arch Fam Med. 1999;8:319-325.
FREE FULL TEXT
34. Katon W, Von Korff M, Lin E, Bush T, Ormel J. Adequacy and duration of antidepressant treatment in primary care. Med Care. 1992;30:67-76.
FULL TEXT
|
ISI
| PUBMED
 CiteULike  Connotea  Del.icio.us  Digg  Reddit  Technorati  Twitter
What's this?
THIS ARTICLE HAS BEEN CITED BY OTHER ARTICLES
|
Trial of Computerized Screening for Adolescent Behavioral Concerns
Stevens et al.
Pediatrics 2008;121:1099-1105.
ABSTRACT
| FULL TEXT
Care Management for Depression in Primary Care Practice: Findings From the RESPECT-Depression Trial
Nutting et al.
Ann Fam Med 2008;6:30-37.
ABSTRACT
| FULL TEXT
Guidelines for Adolescent Depression in Primary Care (GLAD-PC): I. Identification, Assessment, and Initial Management
Zuckerbrot et al.
Pediatrics 2007;120:e1299-e1312.
ABSTRACT
| FULL TEXT
Cost-Effectiveness of Quality Improvement Programs for Patients With Subthreshold Depression or Depressive Disorder
Wells et al.
Psychiatr. Serv. 2007;58:1269-1278.
ABSTRACT
| FULL TEXT
Predicting Outcomes of Primary Care Patients With Major Depression: Development of a Depression Prognosis Index
Rubenstein et al.
Psychiatr. Serv. 2007;58:1049-1056.
ABSTRACT
| FULL TEXT
Cost Shifting Under Managed Behavioral Health Care
Zuvekas et al.
Psychiatr. Serv. 2007;58:100-108.
ABSTRACT
| FULL TEXT
Collaborative care for depression in primary care: Making sense of a complex intervention: systematic review and meta-regression
Bower et al.
Br. J. Psychiatry 2006;189:484-493.
ABSTRACT
| FULL TEXT
Collaborative Care for Bipolar Disorder: Part II. Impact on Clinical Outcome, Function, and Costs
Bauer et al.
Psychiatr. Serv. 2006;57:937-945.
ABSTRACT
| FULL TEXT
Quality Improvement for Depression in Primary Care: Do Patients With Subthreshold Depression Benefit in the Long Run?
Wells et al.
Am. J. Psychiatry 2005;162:1149-1157.
ABSTRACT
| FULL TEXT
Influence of Patients' Requests for Direct-to-Consumer Advertised Antidepressants: A Randomized Controlled Trial
Kravitz et al.
JAMA 2005;293:1995-2002.
ABSTRACT
| FULL TEXT
Clinical Results for Patients With Major Depressive Disorder in the Texas Medication Algorithm Project
Trivedi et al.
Arch Gen Psychiatry 2004;61:669-680.
ABSTRACT
| FULL TEXT
The Partners in Care Approach to Ethics Outcomes in Quality Improvement Programs for Depression
Halpern et al.
Psychiatr. Serv. 2004;55:532-539.
ABSTRACT
| FULL TEXT
Five-Year Impact of Quality Improvement for Depression: Results of a Group-Level Randomized Controlled Trial
Wells et al.
Arch Gen Psychiatry 2004;61:378-386.
ABSTRACT
| FULL TEXT
Major Depression, Physical Illness, and Suicidal Ideation in Primary Care
Goodwin et al.
Psychosom. Med. 2003;65:501-505.
ABSTRACT
| FULL TEXT
Educational and Organizational Interventions to Improve the Management of Depression in Primary Care: A Systematic Review
Gilbody et al.
JAMA 2003;289:3145-3151.
ABSTRACT
| FULL TEXT
Aligning Incentives in the Treatment of Depression in Primary Care With Evidence-Based Practice
Frank et al.
Psychiatr. Serv. 2003;54:682-687.
ABSTRACT
| FULL TEXT
Improving the detection and management of depression in primary care
Gilbody et al.
Qual Saf Health Care 2003;12:149-155.
ABSTRACT
| FULL TEXT
Collaborative Care Management of Late-Life Depression in the Primary Care Setting: A Randomized Controlled Trial
Unutzer et al.
JAMA 2002;288:2836-2845.
ABSTRACT
| FULL TEXT
Other Articles Noted
Evid. Based Ment. Health 2002;5:8-8.
FULL TEXT
|
