The Expert Consensus Guideline Series: Medication Treatment of Bipolar Disorder 2000.

Abstract

New treatments for bipolar disorder have been reported since we first published survey-based expert consensus guidelines in 1996. The evidence for these treatments varies widely; data are especially limited regarding comparisons between treatments and how to sequence them. We therefore undertook a new survey of expert opinion in order to bridge gaps between the research evidence and key clinical decisions.
Based on a literature review, a written survey was prepared which asked about 1,276 options for psychopharmacologic interventions in 48 specific clinical situations. Most options were scored using a modified version of the RAND Corporation 9-point scale for rating appropriateness of medical decisions. We contacted 65 national experts, 58 of whom (89%) completed the survey. Consensus on each option was defined as a non-random distribution of scores by chi-square test. We assigned a categorical rank (first-line/preferred choice, second-line/alternate choice, third-line/usually inappropriate) to each option based on the confidence interval of its mean rating. Guideline tables indicating preferred treatment strategies were then developed for key clinical situations.
The expert panel reached consensus on many key strategies, including acute and preventive treatment for mania (euphoric, mixed, and dysphoric subtypes), depression, and rapid cycling, and approaches to managing the complications of treatment resistance and comorbidity. Use of a mood stabilizer is recommended in all phases of treatment. Divalproex (especially for mixed or dysphoric subtypes) and lithium are the cornerstone choices among this class for both acute and preventive treatment of mania. Regardless of which is selected first, if monotherapy fails, the next recommended intervention is to use these agents in combination. The combination can then serve as the foundation on which other medications are added, if needed. Carbamazepine is the leading alternative mood stabilizer for mania. Expert opinion regards other new anticonvulsants as second-line options (e.g., if the previously mentioned mood stabilizers fail or are contraindicated). For milder depression, a mood stabilizer, especially lithium, may be used as monotherapy. Divalproex and lamotrigine are other first-line choices. For more severe depression, a standard antidepressant should be combined with lithium or divalproex. Bupropion, selective serotonin reuptake inhibitors (SSRIs), and venlafaxine are preferred antidepressants, and should be tapered 2 to 6 months after remission. Divalproex monotherapy is recommended for initial treatment of either depression or mania with rapid cycling. Antipsychotics are recommended for use with the above regimens for mania or depression with psychosis, and as potential adjuncts in non-psychotic episodes. Atypical antipsychotics, especially olanzapine and risperidone, were generally preferred over conventional antipsychotics. Recommendations are also given concerning the use of electroconvulsive therapy (ECT), clozapine, thyroid hormone, stimulants, and various novel agents for patients with treatment-refractory illness.
The experts reached high levels of consensus on key steps in treating bipolar disorder despite obvious gaps in high-quality data. To evaluate many of the treatment options in this survey, the experts had to extrapolate beyond controlled data; however, their recommendations are generally conservative. Experts reserve strongest support for initial strategies and individual medications for which there are high-quality research data, or for which there are longstanding patterns of clinical usage. Within the limits of expert opinion and with the understanding that new research data may take precedence, these guidelines provide clear pathways for addressing common clinical questions in a manner that can be used to inform clinicians and educate patients regarding the relative merits of a variety of interventions.

Full text

Page 1

APRIL 2000 • A POSTGRADUATE MEDICINE SPECIAL REPORT •
1
The Expert Consensus Guideline Series
Medication Treatment of Bipolar Disorder
2000
Gary S. Sachs, M.D.
Director of Partners Bipolar Treatment Center, Massachusetts General Hospital
Assistant Professor of Psychiatry, Harvard Medical School
David J. Printz, M.D.
Director of the Bipolar Disorder Research Clinic
and Assistant Clinical Professor of Psychiatry, Columbia University
David A. Kahn, M.D.
Associate Clinical Professor of Psychiatry, Columbia University
Daniel Carpenter, Ph.D.
Vice President for Information Systems, Comprehensive NeuroScience, Inc.
John P. Docherty, M.D.
President and Chief Executive Officer, Comprehensive NeuroScience, Inc.
Editing and Design. Ruth Ross, M.A., David Ross, M.A., M.C.E., Ross Editorial
Acknowledgments. The authors thank Jennifer Alzona, of Expert Knowledge Systems, for managing
the data collection, and Aysegul Yildiz, M.D., of Harvard Medical School, for work on the
bibliography. Dr. Kahn was the project manager.
Reprints. An Adobe Acrobat file of this document may be downloaded from the Internet at the Web
site www.psychguides.com. Reprints may be obtained by sending requests with a shipping/ handling
fee of $5.00 per copy to: AdMail, 840 Access Road, Stratford, CT 06615. For pricing on bulk orders
of 50 copies or more, please call Expert Knowledge Systems, L.L.C., at (914) 997-4008. Single or
bulk reprints of the patient-family guide may be obtained from the National Depressive and Manic-
Depressive Association (NDMDA), 800-82-NDMDA (800-826-3632), or from the National
Alliance for the Mentally Ill (NAMI), 800-950-NAMI (800-950-6264).
Funding. The project was supported by unrestricted educational grants to Expert Knowledge
Systems, LLC, and to the Health Knowledge Improvement Foundation, from Abbott Laboratories,
Janssen Pharmaceutica, Eli Lilly and Company, Glaxo Wellcome, Ortho-McNeil Pharmaceutical,
Pfizer, and AstraZeneca. Drs. Sachs, Printz, Kahn, and Docherty have received clinical trials
contracts, speaking fees, or consulting fees from some or all of these companies. Although we did not
inquire, we also believe that many of the individuals completing the survey have similar relationships.

Page 2

Expert Consensus Guideline Series
• A POSTGRADUATE MEDICINE SPECIAL REPORT • APRIL 2000
2
The Expert Consensus Panel for Bipolar Disorder
The following participants in the Expert Consensus Survey were identified from several sources: recent research publications and
funded grants, the DSM-IV advisers for mood disorders, the Task Force for the American Psychiatric Association’s Practice
Guidelines for the Treatment of Patients with Bipolar Disorder, and those who have worked on other mood disorder guidelines. Of
the 65 experts to whom we sent the bipolar disorder survey, 58 (89%) replied. The recommendations in the guidelines reflect the
aggregate opinions of the experts and do not necessarily reflect the opinion of each individual on each question.
Michael Allen, M.D.
University of Colorado School of Medicine
Lori Altshuler, M.D.
University of California, Los Angeles
Claudia Baldassano, M.D.
University of Pennsylvania
Ross J. Baldessarini, M.D.
Harvard Medical School, McLean Hospital
James C. Ballenger, M.D.
Medical University of South Carolina
Mark S. Bauer, M.D.
Brown University, VA Medical Center, Providence
Charles L. Bowden, M.D.
U. of Texas Health Sciences Center, San Antonio
Kathleen Brady, M.D.
Medical University of South Carolina
Joseph R. Calabrese, M.D.
Case Western Reserve University
Roy Chengappa, M.D.
Western Psychiatric Institute & Clinic
James C.Y. Chou, M.D.
Bellevue Hospital, New York
William H. Coryell, M.D.
University of Iowa College of Medicine
Jonathan R.T. Davidson, M.D.
Duke University Medical Center
John M. Davis, M.D.
University of Illinois
Lori Davis, M.D.
VA Medical Center, Tuscaloosa
J. Raymond DePaulo, Jr., M.D.
Johns Hopkins University School of Medicine
Steven L. Dubovsky, M.D.
University of Colorado
David L. Dunner, M.D.
University of Washington Medical Center
Rif S. El-Mallakh, M.D.
University of Louisville School of Medicine
Dwight L. Evans, M.D.
University of Pennsylvania
Peter L. Forster, M.D.
University of California, San Francisco
Mark Frye, M.D.
University of California, Los Angeles
Alan J. Gelenberg, M.D.
U. of Arizona Health Sciences Center
Michael Gitlin, M.D.
University of California, Los Angeles
Joseph F. Goldberg , M.D.
Cornell Medical Center
Robert N. Golden, M.D.
U. of North Carolina School of Medicine
Paul J. Goodnick, M.D.
University of Miami School of Medicine
John H. Greist, M.D.
Health Care Technology Systems, Madison, WI
Laszlo Gyulai, M.D.
University of Pennsylvania
Robert M.A. Hirschfeld, M.D.
U. of Texas Medical Branch, Galveston
Philip G. Janicak, M.D.
Psychiatric Institute, U. of Illinois, Chicago
James W. Jefferson, M.D.
Health Care Technology Systems, Madison, WI
Russell Joffe, M.D.
McMaster University
Paul E. Keck Jr., M.D.
University of Cincinnati College of Medicine
Gabor Keitner, M.D.
Brown University, Rhode Island Hospital
Terence A. Ketter, M.D.
Stanford University School of Medicine
Donald F. Klein, M.D.
Columbia University
K. Ranga Rama Krishnan, M.D.
Duke University Medical Center
Justine Lalonde, M.D.
Massachusetts General Hospital, McLean Hospital
Robert H. Lenox, M.D.
University of Pennsylvania
Michael R. Liebowitz, M.D.
Columbia University
Husseini Manji, M.D.
Wayne State University
Lauren Marangell, M.D.
Baylor College of Medicine
Charles B. Nemeroff, M.D.
Emory University School of Medicine
Frederick Petty, M.D., Ph.D.
VA Medical Center, Dallas
Robert M. Post, M.D.
National Institute of Mental Health
S. Craig Risch, M.D.
Medical University of South Carolina
Jerrold F. Rosenbaum, M.D.
Harvard Med. School, Mass. General Hospital
Peter Roy-Byrne, M.D.
Harborview Medical Center, Seattle
Gary S. Sachs, M.D.
Harvard Medical School
David A. Solomon, M.D.
Brown University, Rhode Island Hospital
Andrew L. Stoll, M.D.
Harvard Medical School, McLean Hospital
Trisha Suppes, M.D., Ph.D.
U. of Texas Southwestern Medical Center, Dallas
Alan C. Swann, M.D.
U. of Texas Health Sciences Center, Houston
Michael E. Thase, M.D.
University of Pittsburgh School of Medicine
Peter C. Whybrow, M.D.
UCLA, Neuropsychiatric Institute
John Zajecka, M.D.
Rush-Presbyterian-St. Luke’s Med Center
Carlos Zarate, M.D.
University of Massachusetts

Page 3

MEDICATION TREATMENT OF BIPOLAR DISORDER 2000
APRIL 2000 • A POSTGRADUATE MEDICINE SPECIAL REPORT •
3
Contents
Expert Consensus Panel..............................................................................................................................2
Preface ......................................................................................................................................................4
Introduction: Methods, Summary, and Commentary.................................................................................5
Treatment Selection Algorithms...............................................................................................................14
GUIDELINES
I. TREATMENT OF MANIA
Guideline 1:
Initial Strategy for First Manic Episode.............................................................................16
Guideline 2:
Next Step After Inadequate Response to Initial Strategy for First Manic Episode..............18
Guideline 3:
Maintenance Treatment After a Manic Episode................................................................21
Guideline 4:
Adequate Dose and Duration of Mood Stabilizers.............................................................24
II. TREATMENT OF BIPOLAR DEPRESSION
Guideline 5:
Treatment of First Episode of Bipolar Major Depression...................................................25
Guideline 6:
Inadequate Response to Initial Strategy for Bipolar Depression.........................................30
III. TREATMENT OF RAPID-CYCLING BIPOLAR DISORDER
Guideline 7:
Treatment of Rapid-Cycling Bipolar Disorder ..................................................................36
IV. OTHER TREATMENT ISSUES
Guideline 8:
Selecting Medications for Bipolar Presentations That Resemble Other Disorders..............41
Guideline 9:
Use of Thyroid Hormone in Patients With Bipolar Disorder............................................42
Guideline 10: Managing Special Problems...............................................................................................44
Bibliography.........................................................................................................................................47
SURVEY RESULTS
Expert Survey Results and Guideline References.......................................................................................50
A GUIDE FOR PATIENTS AND FAMILIES ...........................................................................................97

Page 4

Expert Consensus Guideline Series
• A POSTGRADUATE MEDICINE SPECIAL REPORT • APRIL 2000
4
Preface
McGraw-Hill Healthcare Information Programs is pleased to publish the latest revision of The
Expert Consensus Guideline Series: Medication Treatment of Bipolar Disorder 2000. The practice
guidelines described in this publication have employed the latest survey techniques and reflect
only the most current clinical standards. The result is a practical reference tool not only for clini-
cians but also for mental health educators and other healthcare professionals involved in the care
of patients who have bipolar disorder. These guidelines, assembled under the expert direction of
the editors (Gary Sachs, M.D., David J. Prinz, M.D., David A. Kahn, M.D., Daniel Carpenter,
Ph.D., and John P. Docherty, M.D.), are designed to be easy to follow and use.
Treating patients with bipolar disorder is never easy, and the array of pharmacologic interventions
can be difficult to understand and deploy. These guidelines offer a “one stop” reference. They
deal with the initial and long-term management of common scenarios as well as complicated
treatment issues. Interventions for the specific types of bipolar disorder—mania, bipolar depres-
sion, and rapid-cycling bipolar disorder—are outlined in detail. Initial and secondary options are
presented for each type of disorder, along with advice regarding multiple- vs. single-drug therapy,
side effects, and inadequate response to therapy. The section A Guide for Patients and Families
(page 97), which includes information, resource groups, and a reference list, is exceptionally well
done and will be practical for use by both groups. It will also serve as a helpful primer for primary
care physicians.
The printed publication will now become a valuable addition to my reference library. I hope you
find the guidelines to be beneficial in the care of your patients.
William O. Roberts, MD
Editor-in-Chief
McGraw-Hill Healthcare
Information Programs

Page 5

MEDICATION TREATMENT OF BIPOLAR DISORDER 2000
APRIL 2000 • A POSTGRADUATE MEDICINE SPECIAL REPORT •
5
Introduction: Methods, Summary, and Commentary
Gary S. Sachs, M.D., David J. Printz, M.D., David A. Kahn, M.D., Daniel Carpenter, Ph.D., John P. Docherty, M.D.
ABSTRACT
Objectives. New treatments for bipolar disorder have been
reported since we first published survey-based expert
consensus guidelines in 1996. The evidence for these
treatments varies widely; data are especially limited regard-
ing comparisons between treatments and how to sequence
them. We therefore undertook a new survey of expert
opinion in order to bridge gaps between the research
evidence and key clinical decisions.
Method. Based on a literature review, a written survey was
prepared which asked about 1,276 options for psycho-
pharmacologic interventions in 48 specific clinical situa-
tions. Most options were scored using a modified version of
the RAND Corporation 9-point scale for rating appropri-
ateness of medical decisions. We contacted 65 national
experts, 58 of whom (89%) completed the survey. Consen-
sus on each option was defined as a non-random distribu-
tion of scores by chi-square test. We assigned a categorical
rank (first-line/preferred choice, second-line/alternate
choice, third-line/usually inappropriate) to each option
based on the confidence interval of its mean rating. Guide-
line tables indicating preferred treatment strategies were
then developed for key clinical situations.
Results. The expert panel reached consensus on many key
strategies, including acute and preventive treatment for
mania (euphoric, mixed, and dysphoric subtypes), depres-
sion, and rapid cycling, and approaches to managing the
complications of treatment resistance and comorbidity.
Use of a mood stabilizer is recommended in all phases
of treatment. Divalproex (especially for mixed or dysphoric
subtypes) and lithium are the cornerstone choices among
this class for both acute and preventive treatment of mania.
Regardless of which is selected first, if monotherapy fails,
the next recommended intervention is to use these agents in
combination. The combination can then serve as the foun-
dation on which other medications are added, if needed.
Carbamazepine is the leading alternative mood stabilizer
for mania. Expert opinion regards other new anticonvul-
sants as second-line options (e.g., if the previously men-
tioned mood stabilizers fail or are contraindicated).
For milder depression, a mood stabilizer, especially
lithium, may be used as monotherapy. Divalproex and
lamotrigine are other first-line choices. For more severe
depression, a standard antidepressant should be combined
with lithium or divalproex. Bupropion, selective serotonin
reuptake inhibitors (SSRIs), and venlafaxine are preferred
antidepressants, and should be tapered 2 to 6 months after
remission. Divalproex monotherapy is recommended for
initial treatment of either depression or mania with rapid
cycling.
Antipsychotics are recommended for use with the
above regimens for mania or depression with psychosis, and
as potential adjuncts in non-psychotic episodes. Atypical
antipsychotics, especially olanzapine and risperidone, were
generally preferred over conventional antipsychotics.
Recommendations are also given concerning the use of
electroconvulsive therapy (ECT), clozapine, thyroid hor-
mone, stimulants, and various novel agents for patients
with treatment-refractory illness.
Conclusions. The experts reached high levels of consensus
on key steps in treating bipolar disorder despite obvious
gaps in high-quality data. To evaluate many of the treat-
ment options in this survey, the experts had to extrapolate
beyond controlled data; however, their recommendations
are generally conservative. Experts reserve strongest
support for initial strategies and individual medications
for which there are high-quality research data, or for
which there are longstanding patterns of clinical usage.
Within the limits of expert opinion and with the under-
standing that new research data may take precedence,
these guidelines provide clear pathways for addressing
common clinical questions in a manner that can be used
to inform clinicians and educate patients regarding the
relative merits of a variety of interventions. (Postgrad Med
Special Report. 2000(April):1–104)
WHY A REVISION?
When we published the first Expert Consensus Guidelines
for the Treatment of Bipolar Disorder
surveys completed in 1995), we were aware that new
research and the introduction of new treatments might
soon require us to revise them. The 2000 Guidelines are
our first update. We based them on a survey of 58 leading
experts on the medication treatment of bipolar disorder.
Because the sheer number of potentially useful medications
has made clinical decisions ever more complex, we elected
to focus on medications and not to review options for
psychosocial treatment. Readers may still refer to the
earlier edition of the Guidelines
chosocial issues.
The contribution of expert consensus to practice
guideline development continues to evolve throughout
medicine, alongside the “gold standard” of meta-analysis of
clinical trials and other experimental data. Developers of
guidelines throughout medicine continue to struggle with
the problem that the number of possible combinations and
sequences of available treatments for many diseases makes
1 in 1996 (based on
1 for information on psy-

Page 6

Expert Consensus Guideline Series
• A POSTGRADUATE MEDICINE SPECIAL REPORT • APRIL 2000
6
it difficult to establish practical guidelines based entirely
on scientific data.
describing expert opinion in a quantitative, reliable manner
to help fill some of the gaps in evidence-based guidelines.
This method has been applied to a variety of psychiatric
disorders.
2, 3 Our group has developed a method for
1, 4–8
METHOD OF DEVELOPING
EXPERT CONSENSUS GUIDELINES
Creating the Surveys
We first created a skeleton algorithm based on a literature
review. We sought to identify key decision points in the
medication treatment of bipolar disorder as well as all the
feasible treatment options. We highlighted important
clinical questions that had not yet been adequately ad-
dressed or definitely answered.
was then developed covering 48 specific clinical situations,
divided into 166 subsections based on contingencies (e.g.,
subtypes, treatment history, comorbidity) with a total of
1,276 options for intervention. We began with questions
concerning broad strategies, such as classes of medication,
and then delved into tactics, such as specific medication
selection and dosing. The survey took 2 or more hours to
complete.
9 A written questionnaire
The Rating Scale
For 1,065 of the options in the survey, we asked raters to
evaluate appropriateness by means of a 9-point scale
slightly modified from a format developed by the RAND
Corporation for ascertaining expert consensus.
other options asked raters to fill in a blank, such as dosage
or duration of treatment.) We explicitly asked the raters to
consider both personal experience and research evidence
(we did not provide a literature review) in making their
ratings, but not to consider financial cost. We presented
the rating scale to the experts with the anchors shown in
figure 1.
10 (The 211
Figure 1. The Rating Scale
Extremely 1 2 3 4 5 6 7 8 9
Inappropriate
Extremely
Appropriate
9 = Extremely appropriate: this is your treatment of choice
7–8 = Usually appropriate: a first-line treatment you would
often use
4–6 = Equivocal: a second-line you would sometimes use
(e.g., patient/family preference or if first-line
treatment is ineffective, unavailable, or unsuitable)
2–3 = Usually inappropriate: a treatment you would rarely use
1 = Extremely inappropriate: a treatment you would never
use
Figure 2 shows an excerpt from Survey Question 1 as
an example of our question format.
Figure 2. Sample Survey Question
1. Treatment of mania: first episode, initial strategy. A physically
healthy person in his or her 20s presents with a first manic epi-
sode severe enough to warrant hospital admission, or a first hy-
pomanic episode severe enough to pose a likely eventual threat
to functioning if unchecked. Based on the dominant symptom
pictures shown below, please rate each of the following overall
strategies as an initial intervention, assuming the patient is
willing to take oral medication. (Subsequent questions will ask
you about specific medications within the broad classes.)
Euphoric Mania
Mood stabilizer alone1 2 3 4 5 6 7 8 9
Antipsychotic alone1 2 3 4 5 6 7 8 9
Mood stabilizer + antipsychotic 1 2 3 4 5 6 7 8 9
Benzodiazepine alone1 2 3 4 5 6 7 8 9
Benzodiazepine + mood stabilizer1 2 3 4 5 6 7 8 9
Benzodiazepine + mood stabilizer +
antipsychotic
1 2 3 4 5 6 7 8 9
Selecting the Expert Panel
We identified 65 leading American experts in bipolar disor-
der through the following sources: authors of important
publications in the past 5 years, recipients of research grants
from government or industry, and members of American
Psychiatric Association task forces for bipolar disorder
practice guidelines and the affective disorders section of the
Diagnostic and Statistical Manual of Mental Disorders, Fourth
Edition (DSM-IV). We excluded individuals who had
previously declined to complete surveys for us. We sought to
include both new and established clinical investigators.
Data Analysis for Options Scored on the Rating Scale
For each option, we first defined the presence or absence of
consensus as a distribution unlikely to occur by chance by
performing a chi-square test (P<0.05) of the distribution of
scores across the 3 ranges of appropriateness (1–3, 4–6, 7–
9). Next we calculated the mean and 95% confidence
interval (CI). A categorical rating of first-, second-, or
third-line was designated based on the lowest category in
which the CI fell, with boundaries of 6.5 or greater for
first-line, and 3.5 or greater for second-line. Within first-
line, we designated an item as “treatment of choice” if at
least 50% of the experts rated it as 9.

Page 7

MEDICATION TREATMENT OF BIPOLAR DISORDER 2000
APRIL 2000 • A POSTGRADUATE MEDICINE SPECIAL REPORT •
7
Figure 3. Results of the Euphoric Mania Section of Survey Question 1
9 5 % C O N F I D E N C E I N T E R V A L S
Third LineSecond Line
Tr of 1st
Chc Line Line Line
2nd 3rd
First Line Avg(SD)
Euphoric Mania
Mood stabilizer alone8.0(1.4)
57
88130
Mood stabilizer + benzodiazepine7.3(1.8)
36
75205
Mood stabilizer + antipsychotic6.4(2.2)
15
582616
Mood stabilizer + benzodiazepine + antipsychotic5.4(2.3)
6
40 3327
Antipsychotic alone4.0(2.0)
2
13 38 49
Benzodiazepine alone2.5(1.3)
0
%
0
%
13
%
88
%123456789
Displaying the Survey Results
The results of Question 1 (figure 2) are presented graphi-
cally in figure 3. The confidence intervals (CIs) for each
treatment option are shown as horizontal bars and the
numerical values are given in the table on the right. (The
display of all of the results can be found in the survey
results section, pages 50–96.)
The Ratings
First-line treatments
out on top when the experts’ responses to the survey were
statistically aggregated. These are options that the panel
feels are usually appropriate as initial treatment for a given
situation. Treatment of choice,
especially strong first-line recommendation (having been
rated as “9” by at least half the experts). In choosing
between several first-line recommendations, or deciding
whether to use a first-line treatment at all, clinicians should
consider the overall clinical situation, including the pa-
tient’s prior response to treatment, side effects, general
medical problems, and patient preferences.
are those strategies that came
when it appears, is an
Second-line treatments
patients who cannot tolerate or do not respond to the first-
line choices. Alternatively, a second-line choice might be
used for initial treatment if the first-line options are
deemed unsuitable for a particular patient (e.g., because of
poor previous response, inconvenient dosing regimen,
particularly annoying side effects, general medical contra-
indication, potential drug-drug interaction, or if the
experts don’t agree on a first-line treatment).
For some questions, second-line ratings dominated,
especially when the experts did not reach any consensus on
first-line options. In such cases, to differentiate within the
pack, we label those items whose CIs overlap with the first-
line category as “high second-line.”
are reasonable choices for
Third-line treatments
only when preferred alternatives have not been effective.
are usually inappropriate or used
No consensus.
chi-square test to determine whether the experts’ responses
were randomly distributed across the 3 categories, which
suggests a lack of consensus. These items are indicated by
an unshaded bar in the survey results.
For each item in the survey, we used a
Statistical differences between treatments. While we did
not perform tests of significances for most treatments, the
reader can perform an “eyeball” test by looking to see
whether the CIs overlap (indicating no significant differ-
ence between options by t-test). The wider the gap be-
tween the CIs, the smaller the P value would be (i.e., the
more significant the difference). In some questions there
are striking and important differences within levels, which
we occasionally point out. Often, however, the differences
within levels are not significant from a statistical perspec-
tive. Also, there are sometimes no statistical differences
between choices rated at the bottom of first-line and those
at the top of second-line.
From Survey Results to Guidelines
After the survey results were analyzed and ratings assigned,
the next step was to turn these recommendations into user-
friendly guidelines. We distinguish 2 levels, preferred op-
tions and alternate options, that generally correspond to
first- and higher second-line ratings. Whenever the guide-
line gives more than 1 treatment in a rating level, we list
them in the order of their mean scores. As an example, the
full results of the question presented above are shown on
page 50 and are used in Guideline 1: Initial Strategy for a
First Manic Episode (page 16). A mood stabilizer as mono-
therapy is the preferred initial approach for most types of
mania, while a mood stabilizer plus an antipsychotic was
clearly preferred for mania with psychosis. As mentioned in
*
*

Page 8

Expert Consensus Guideline Series
• A POSTGRADUATE MEDICINE SPECIAL REPORT • APRIL 2000
8
the legend, bold italics indicate treatment of choice rating,
an especially strong opinion. The clinician might try these
approaches, but move to the combinations with adjunctive
medications if the patient could not be managed with
monotherapy. Note that the choice of adjunct—benzodia-
zepine or antipsychotic—differs with the subtype of mania.
Summary of Results
The complete set of data from the survey is presented on
pages 50–96. The guidelines derived by the editors from
the data are presented on pages 16–46. Graphic treatment
algorithms summarizing the expert recommendations are
provided on pages 14–15. We summarize the highlights
here.
Who were the panelists? Of the 65 national experts we
contacted, 58 (89%) completed the survey. Of the 58
completers, 35 (60%) had also completed the 1996 survey.
The average panelist was 49 years old (standard deviation,
9.3). Most of the panelists (80%) carry out their research
and practice activities in non-Veterans Administration,
non-governmental academic medical centers, typically
spending about 25% of their time actually seeing patients;
94% of the panel participated in clinical research on
bipolar disorder in the past 5 years. Most of the panelists
had seen between 20 and 100 bipolar patients in clinical
trials in the past year and had treated additional bipolar
patients outside clinical trials.
What was the degree of consensus? Consensus was reached
on 950 (89%) of the 1,065 options that used the rating
scale. At least 1 first-line option was identified in 146
(88%) of the 166 subsections. Those areas in which no
first-line options were identified all involved complex
comorbidities or treatment-refractory illness.
What are the key recommendations? In terms of clinical
practice, the single most important recommendation is to
use a mood stabilizer in all phases of treatment. Divalproex
and lithium are the cornerstone choices among this class
for both acute-phase and preventive treatment. They
should be tried first when monotherapy is desired, in
combination when either has failed, and as the bedrock
upon which other medications may be layered. The leading
alternative mood stabilizers are carbamazepine, especially
for mania, and the newer agent lamotrigine, especially for
depression. The next major finding is that when an anti-
psychotic is needed, atypicals are generally preferred over
conventionals for initial treatment. The only presentation
where conventionals join atypicals as a first-line option is
mania with psychosis. A third important finding is that
mild depression should be treated with mood stabilizer
monotherapy initially, while severe depression should be
treated from the start with an antidepressant plus a mood
stabilizer. However, after resolution of a first episode of
bipolar depression, antidepressants should be tapered in 2 to
6 months, a much shorter continuation period than is
generally advised for non-bipolar depression. A fourth
finding is that either mania or depression with rapid cycling
should be treated initially with a mood stabilizer alone,
preferably divalproex for either phase. The top-rated choices
for initial medication treatment are shown in table 1.
Table 1. Top-Rated Choices for Initial Medication
(Assumes no contraindications; adjunctive medications
added subsequently if indicated)
Euphoric mania or
hypomania
Lithium or divalproex
Mixed or dysphoric maniaDivalproex
Mania with psychosis Divalproex or lithium with
antipsychotic (atypical or
conventional)
Milder depression Lithium
More severe depressionLithium or divalproex
with antidepressant (plus
atypical antipsychotic if
delusional)
Mania or depression with
recent rapid cycling
Divalproex
Key comparisons with the last survey. Support for the use of
divalproex has increased. Ratings for lithium and carbamaz-
epine remain stable. Lamotrigine, which was included in this
survey for the first time, received positive ratings for the
treatment of bipolar depression. In an important switch,
atypical antipsychotics are now generally rated ahead of
conventionals. Finally, venlafaxine received significantly
stronger ratings in this survey and joined bupropion and
SSRIs in the group of first-line antidepressants.
Key comparisons with recent literature. While panelists
were not asked to review the literature in order to answer
the survey, we have informally evaluated the degree to
which their recommendations are supported by evidence.
(It is interesting to note that a comparison of the 1996
Guidelines with evidence-based guidelines from the Ameri-
can Psychiatric Association and other sources revealed no
contradictions, but differences in emphasis and in degree
of specificity.
which high-quality data, such as methodologically sound,
placebo-controlled trials, are available. They showed
intermediate or less support for treatments for which only
less rigorous studies and case reports are available. Even so,
there are many situations for which there are no well-
controlled data, such as key drug-drug comparisons or the
management of illness that is refractory to first-line treat-
ments. In these situations, the panel did not display strong
11) The experts mostly favored treatments for

Page 9

MEDICATION TREATMENT OF BIPOLAR DISORDER 2000
APRIL 2000 • A POSTGRADUATE MEDICINE SPECIAL REPORT •
9
preferences—rather they supported the reasonableness of
trying those options for which there was some evidence. For
the treatment of refractory illness, they prefer combining or
switching to established treatments (including ECT and
clozapine) before experimenting with less established medi-
cations.
COMMENTARY
What do the new survey results tell us about the state of
optimum practice in treating bipolar disorder? In this
section, we discuss similarities and changes since our last
survey in 1995 and consider the relationship between
opinion and evidence in the experts’ key decisions.
Mood Stabilizers for Mania
Just as in 1995, divalproex and lithium are still the highest
rated first-line treatments for all subtypes of mania. Lith-
ium’s numerical scores were unchanged, while scores for
divalproex increased: scores for divalproex are now nearly
indistinguishable from those for lithium in mania with
euphoric mood, and divalproex is preferred as the treatment
of choice for both mixed and dysphoric mania. These
findings are consistent with the results of a large-scale
prospective clinical trial comparing divalproex, lithium, and
placebo in acute mania
subtype.
most favored alternative mood stabilizer, a result that is
consistent with it being the only other non-antipsychotic
medication that has been shown to be effective for mania in
well-designed studies.
second-line ratings as an initial treatment, reflecting the
preliminary nature of the evidence for its efficacy
perhaps concern about the need for slow titration of dosage
to minimize the risk of rashes. Gabapentin was rated mostly
a third-line choice, which is consistent with the lack of high-
quality research.
panic disorder, and as an add-on therapy to other mood
stabilizers. Controlled data for gabapentin are unlikely to be
published, given the absence of positive findings. Further
studies are unlikely, perhaps owing to the imminent expira-
tion of its patent and consequent lack of commercial poten-
tial. Its apparent popularity among clinicians may be more
for bipolar II disorder (the survey focused more on bipolar I
disorder). A related anticonvulsant, pregabalin, which is not
yet available, may receive more rigorous research. Among
treatments that have been anecdotally reported to be helpful,
topiramate and calcium channel blockers were rated as
appropriate for patients who are resistant to other treat-
ments, although not for initial therapy. There was some
support for the omega-3 fatty acids but no overall consensus
on their usefulness, while tiagabine was viewed as generally
not helpful. Antipsychotic monotherapy was not highly
recommended except for possible use in psychotic mania; we
12 and post hoc analyses of response by
13 Carbamazepine, as in the last survey, remains the
14 Lamotrigine received, at best, low
15 and
16 Its strongest ratings are for comorbid
will discuss the use of antipsychotics in more detail in a later
section of this introduction.
Bipolar Depression
The lack of data on major depression in bipolar disorder
makes it a difficult topic to address in evidence-based guide-
lines, but an interesting one on which to examine the consen-
sus of opinion. The experts’ recommendations concerning the
treatment of depression without rapid cycling have shifted
somewhat since 1995. The experts now lean more toward the
use of a mood stabilizer alone for milder initial episodes. Just
as in 1995, however, they recommend adding an antidepres-
sant from the start for severe depression and would also add
an antipsychotic or give ECT for psychotic depression.
The first choice of mood stabilizer for monotherapy for
depression is lithium. However, in contrast with the 1995
survey results, divalproex (not well studied for this use) and
lamotrigine were both also rated first-line, although below
lithium. It should be noted that lamotrigine is the only
medication that has been shown to be effective for bipolar
depression in a large, randomized, controlled clinical trial,
except for studies from decades earlier that showed benefit
from lithium.
It is interesting that the choice of mood stabilizer
changes when it is to be combined with an antidepressant:
lamotrigine drops from the first-line group, and divalproex
rises nearly to the level of lithium. This suggests an emphasis
on using tried-and-true mood stabilizers to guard against
antidepressant-induced switches to mania. Only the top
choice, lithium, has been well tested in this situation, and
was found to protect against mania when combined with
imipramine in a long-term maintenance study.
A new topic we asked about in the current survey was
breakthrough depression (i.e., an episode of depression that
occurs while a patient is on maintenance treatment with
lithium or divalproex). In this situation, the experts favor
first maximizing doses of lithium or divalproex. Divalproex,
lamotrigine, or antidepressants are equally ranked as add-ons
if lithium monotherapy fails, another clear indication of the
growing role for lamotrigine.
Concerning the choice of specific antidepressants, there
are few data on which the panel could base their opinions.
As in the last survey, bupropion and SSRIs are among the
first-line choices for initial therapy, and bupropion is espe-
cially favored for more moderate depression. (The available
SSRIs all received roughly comparable ratings.) It should be
noted that venlafaxine is now rated as a first-line option,
especially for more severe depression; its rating has moved
up considerably since 1995 when it had just been intro-
duced. Monoamine oxidase inhibitors are still a favored
backup and received equal ratings with nefazodone and
mirtazapine. As before, tricyclics are the least favored anti-
depressants but received a few strong votes for more severe
depression.
17
18

Page 10

Expert Consensus Guideline Series
• A POSTGRADUATE MEDICINE SPECIAL REPORT • APRIL 2000
10
In a new question, we asked about strategies for patients
who had not responded to an SSRI. There was strong
consensus to switch to bupropion (treatment of choice) or
alternatively to venlafaxine (first-line), rather than to try
another SSRI. Bupropion was voted the least likely to cause
a switch to mania both in first episode patients and in those
with a history of switching to mania while taking other
antidepressants. This result is consistent both with the last
survey and with its being the only antidepressant that has
been studied prospectively with this question in mind.
For depression that does not respond to an initial
treatment regimen of a mood stabilizer plus an antidepres-
sant, the panel’s confidence in lithium is underscored by the
recommendation that it be added to any regimen that does
not include it—especially to augment a partial response. As
in 1995, the panelists did not have any clear-cut recommen-
dations concerning other augmentation strategies, although
they appear to suggest this sequence:
19
● try an anticonvulsant if the patient is not taking one
(divalproex tied with lamotrigine, carbamazepine next,
then gabapentin)
or
● try another antidepressant
● add thyroid hormone (T3 preferred over T4)
● add a stimulant
● add an atypical antipsychotic, possibly clozapine
or
for seasonal depression, use phototherapy.
As in mania, a host of unproved treatments (dopamine
agonists, omega-3 fatty acids, sleep deprivation, topiramate,
nimodipine, inositol, buspirone), for which there are only
anecdotal reports, received equivocal ratings, with the least
support for phenytoin and St. John’s wort.
Rapid Cycling
For the manic phase of rapid cycling, divalproex, which was
rated treatment of choice in 1995, received still higher
ratings in the current survey. Lithium moved up to a first-
line alternative for non-dysphoric cases, while carbamaz-
epine remained the other major first-line choice. In the
depressed phase of rapid cycling, which we did not explicitly
ask about in 1995, divalproex monotherapy was rated
treatment of choice, and lamotrigine and lithium were tied
as lower first-line choices. Antidepressants should be added
only if mood stabilizers fail. The strong consensus concern-
ing treatments for rapid cycling is generally consistent with
the results of open studies and retrospective analyses. These
show that divalproex and carbamazepine are superior to
lithium,
beneficial in depression than in severe mania.
received modest support for rapid cycling. Atypical antipsy-
chotics were a favored add-on for treatment-refractory
20, 21 especially in mania, and that lamotrigine is more
22 Gabapentin
mania or depression with rapid cycling, an approach that is
supported only by case reports.
Long-Term Preventive Treatment
The panel supported using either lithium or divalproex or a
combination of both, whatever worked during the acute
phase of treatment, for long-term prevention after a manic
episode. Lithium’s efficacy for prevention is well established
based on data from the 1970s (although more recent data
also indicate substantial failure rates). The confidence in
divalproex reflects clinical experience; it has been harder to
obtain controlled evidence for divalproex in maintenance
treatment due to the difficulty of enrolling suitable patients
in randomized, placebo-controlled studies.
clinical reports and open, prospective studies show that
divalproex, in some cases added to lithium or even used in
triple therapy in combination with lithium and carbamaz-
epine, may succeed in patients with treatment-refractory
illness, including rapid cyclers.
about the use of carbamazepine as preventive treatment,
small-scale studies have shown benefits for its use alone
in combination with lithium.
23 Nonetheless,
24, 25 Although we did not ask
26 and
27
Atypical Antipsychotics
This discussion would not be complete without mention-
ing the growing role of atypical antipsychotics in the
treatment of bipolar disorder, especially in light of the
expected Food and Drug Administration approval of
olanzapine for mania and earlier approvals of conventional
antipsychotics for manic psychosis. In the current survey,
atypical antipsychotics other than clozapine are now rated
as first-line agents for adjunctive treatment of mania (as are
benzodiazepines in non-psychotic cases). Atypical antipsy-
chotics are also rated as first-line agents for combined
treatment of psychotic depression, in contrast to the
panel’s preference for conventional antipsychotics in 1995.
The current panel also strongly preferred atypicals when an
antipsychotic is needed for long-term maintenance. Atypi-
cals were also highly rated backups in any phase of rapid
cycling. Conventional antipsychotics received first-line
ratings only for psychotic mania or for patients with mania
who have failed to respond to 1 or 2 atypical antipsychot-
ics. Depot conventional antipsychotics received strong
second-line support for the treatment of nonadherent
patients. The results of the current survey are compatible
with available data, including a pivotal study showing that
olanzapine is more effective than placebo as monotherapy
for mania,
and preliminary experience with quetiapine.
evidence also allays earlier fears of frequent “activation” by
atypicals (although isolated instances may occur). These
data, taken together with the lower risk of extrapyramidal
side effects, support the panel’s conclusions.
28 growing evidence in support of risperidone,
29, 30
31 The recent

Page 11

MEDICATION TREATMENT OF BIPOLAR DISORDER 2000
APRIL 2000 • A POSTGRADUATE MEDICINE SPECIAL REPORT •
11
How should atypicals be sequenced in treatment? First,
although the experts strongly support the adjunctive use of
atypical antipsychotics, they still hesitate to recommend
them over traditional mood stabilizers for monotherapy in
mania. Second, for an adjunct to add to a mood stabilizer in
hypomania, the panel prefers a benzodiazepine over an
antipsychotic. The panel gives equal ratings to adjunctive
benzodiazepines or antipsychotics in more severe mania
without psychosis, and endorses antipsychotics as essential in
psychotic mania. This recommendation—to reserve antipsy-
chotics for more severe cases—is the same as in the last
survey and is consistent with clinical tradition rather than
clear-cut data.
As in the last survey, clozapine is well rated for any
phase of treatment-resistant illness or for rapid cycling.
These uses are supported by well-designed research.
sidone, although unapproved for use at the time of this
survey, was favorably viewed by experts who had used it in
clinical trials for various disorders. It was also mentioned as a
good alternative for patients who have gained weight on
other medications.
The role of atypical antipsychotics as add-on treatments
and as primary mood stabilizers in different phases of bipo-
lar disorder is an important current research area.
32 Zipra-
Dosing and Duration of Treatment
We did not ask about dosing in our previous survey. In the
current survey, we found standard deviations of about 30%
to 40% in recommendations for acute doses of carbamaz-
epine, divalproex, and lithium and about 50% for long-term
maintenance. The variances were even wider for lamotrigine,
gabapentin, topiramate, and tiagabine. We were not sur-
prised by the wide variances given the tremendous variability
between individuals in metabolism, dose-response require-
ments, and tolerance for side effects. It is worth noting that
there was greater agreement on doses of the more established
agents, which is consistent with the availability of far more
experimental data. As in the last survey, experts look for a
quick response to a mood stabilizer in mania: if there is no
response within a week, they would consider using another
medication; if a partial response has reached a plateau after 2
weeks, the experts would begin to use another medication.
When initiating treatment of mania with divalproex,
most experts suggest starting with a full therapeutic dose
(e.g., 20 mg/kg/day), often referred to as a “loading dose.”
Consistent with this approach, the panel also believed that
one might see results about a day or 2 earlier with dival-
proex than with other mood stabilizers (generally advising
a wait of at least 1 to 2 weeks before adding another mood
stabilizer). Adjunctive antipsychotics in mania were given
less time to prove their mettle—a week or less before the
experts recommend trying a second antipsychotic. The
experts also were willing to combine atypical and conven-
tional antipsychotics in treatment-resistant illness, consis-
tent with a common, though experimentally untested,
clinical practice.
We did not repeat a 1995 question on acute-phase
doses of antidepressants. The recommendation in the earlier
survey was to start low and go slow, but to aim eventually
for the same maximum doses as in non-bipolar depression.
For continuation antidepressant treatment, the experts
reiterated their earlier advice to taper after 2 to 6 months of
remission, compared with the 6 to 12 months commonly
recommended for non-bipolar depression. There are few
research data on dosing and duration of antidepressant
treatment in bipolar disorder.
Electroconvulsive Therapy
The experts’ enthusiasm for ECT has not diminished since
1995, despite the plethora of new medications. The panelists
favor the use of ECT early in psychotic depression, prefer-
ring it to lithium augmentation. They would press for the
use of ECT in nonpsychotic depression if the patient has
had an inadequate response to treatment with 2 mood
stabilizers, including a trial of lithium, plus 2 antidepres-
sants. For a patient having a manic episode who has failed to
respond to lithium and divalproex plus antipsychotics, the
experts gave equal ratings to ECT or adding a third mood
stabilizer. In several questions about treatment resistance or
rapid cycling, many experts actually wrote in ECT as a
preference for acute-phase or maintenance treatment even
when we did not offer it as an option. ECT is well sup-
ported by research studies in mania
33 and depression.
CONCLUSIONS
Limitations and Advantages of
Expert Consensus Guidelines
These guidelines can be viewed as an expert consultation, to
be weighed in conjunction with other information and in
the context of each individual patient-physician relationship.
The recommendations do not replace clinical judgment,
which must be tailored to the particular needs of each
clinical situation. We describe groups of patients and make
suggestions intended to apply to the average patient in each
group. However, individual patients will differ greatly in
their treatment preferences and capacities, history of response
to previous treatments, family history of treatment response,
and tolerance for different side effects. Therefore, the experts’
first-line recommendations certainly will not be appropriate
in all circumstances.
We remind readers of several other limitations of these
guidelines:
1. The guidelines are based on a synthesis of the opinions
of a large group of experts. From question to question,
some of the individual experts would differ with the con-
sensus view.

Page 12

Expert Consensus Guideline Series
• A POSTGRADUATE MEDICINE SPECIAL REPORT • APRIL 2000
12
2. We have relied on expert opinion precisely because we
are asking crucial questions that are not yet well an-
swered by the literature. One thing that the history of
medicine teaches us is that expert opinion at any given
time can be very wrong. Accumulating research will ul-
timately reveal better and clearer answers. Clinicians
should therefore stay abreast of the literature for devel-
opments that would make at least some of our recom-
mendations obsolete. We hope to revise the guidelines
periodically based on new research information and on
reassessment of expert opinion to keep them up-to-date.
3. The guidelines are financially sponsored by the pharma-
ceutical industry, which could possibly introduce biases.
Because of this, we have made every step in guideline de-
velopment transparent, reported all results, and taken
little or no editorial liberty.
4. These guidelines are comprehensive but not exhaustive;
because of the nature of our method, we omit some in-
teresting topics on which we did not query the expert
panel.
Despite the limitations, these guidelines represent a
significant advance because of their specificity, ease of use,
and the credibility that comes from achieving a very high
response rate from a large sample of the leading experts in
the field.
Guidelines Research
It is easy to get experts to agree on key steps, but how
do we know they are right? Two major research projects
illustrate the power of consensus guidelines and test their
ability to improve care.
The first is the Texas Medication Algorithm Project
(TMAP), a controlled, 2-year study of whether patients
treated according to guidelines have better outcomes than
patients receiving “treatment as usual” care.
edition of The Guidelines
bipolar disorder module.
The second major project is the Systematic Treatment
Enhancement Program for Bipolar Disorder (STEP-BD)
sponsored by the National Institute of Mental Health. This
project is based on the recognition that, at most of the
decision points in this survey, more than 1 option is rated
first-line, and that together these options represent a menu
of reasonable options for clinical care. It is likely that this
multiplicity of first-line options indicates a state of clinical
uncertainty, which can be resolved only by high-quality
empirical data. Over the next 5 years, STEP-BD will report
outcomes for 5,000 bipolar patients in treatment centers
around the United States which have agreed to implement a
common intervention model based on various first-line
expert recommendations.
Advances in public health do not always require tech-
nological breakthroughs or long periods of waiting for new
34 The previous
1 was the starting point for the
35
data. Immediate gains can be made by increasing the fre-
quency at which the best available known treatments are
implemented. Guidelines offer a rapid means for communi-
cating a distillate of expert opinion. When reaching a clinical
decision point, practitioners and patients can use guidelines
to generate a menu of reasonable choices and then select the
option that is judged best for each individual. This process
drives the next round of expert opinion and the next round
of empirical studies.
REFERENCES
1. Kahn DA, Carpenter D, Docherty JP, et al. The expert
consensus guideline series: treatment of bipolar disorder. J Clin
Psychiatry 1996;57(Suppl 12a):1–88
2. Djulbegovic B, Hadley T. Evaluating the quality of clinical
guidelines: linking decisions to medical evidence. Oncology
1998;12 (11A):310–4
3. Shekelle PG, Kahan JP, Bernstein SJ, et al. The reproducibil-
ity of a method to identify the overuse and underuse of medi-
cal procedures. N Engl J Med 1998;338(26):1888–95
4. McEvoy JP, Weiden PJ, Smith TE, et al. The expert consen-
sus guideline series: treatment of schizophrenia. J Clin Psy-
chiatry 1996;57(Suppl 12b):1–58
5. March JS, Frances A, Carpenter D, et al. The expert consen-
sus guideline series: treatment of obsessive-compulsive disorder.
J Clin Psychiatry 1997;58(Suppl 4):1–72
6. Alexopoulos GS, Silver JM, Kahn DA, et al. The expert
consensus guideline series: treatment of agitation in older per-
sons with dementia. Postgrad Med Special Report
1998;April:1–88
7. McEvoy JP, Scheifler PL, Frances A. The expert consensus
guideline series: treatment of schizophrenia 1999. J Clin Psy-
chiatry 1999;60(Suppl 11):1–80
8. Foa EB, Davidson JRT, Frances A. The expert consensus
guideline series: treatment of posttraumatic stress disorder. J
Clin Psychiatry 1999;60(Suppl 16):1–76
9. Kahn DA, Docherty JP, Carpenter D, et al. Consensus
methods in practice guideline development: a review and de-
scription of a new method. Psychopharmacol Bull 1997;33(4):
631–9
10. Brook RH, Chassin MR, Fink A, et al. A method for the
detailed assessment of the appropriateness of medical technolo-
gies. Int J Tech Assess Health Care 1986;2:53–63

Page 13

MEDICATION TREATMENT OF BIPOLAR DISORDER 2000
APRIL 2000 • A POSTGRADUATE MEDICINE SPECIAL REPORT •
13
11. Suppes T, Habermacher E, Potter W. Bipolar disorder. In:
Fawcett J, Stein DJ, Jobson KO, eds. Textbook of Treatment
Algorithms in Psychopharmacology. New York: John Wiley &
Sons, 1999;59–66
12. Bowden CL, Brugger AM, Swann AC, et al. Efficacy of
divalproex vs lithium and placebo in the treatment of mania.
The Depakote Mania Study Group. JAMA 1994;271(12):
918–24
13. Swann AC, Bowden CL, Morris D, et al. Depression during
mania: treatment response to lithium or divalproex. Arch Gen
Psychiatry 1997;54(1):37–42
14. Emilien G, Maloteaux JM, Seghers A, et al. Lithium com-
pared to valproic acid and carbamazepine in the treatment of
mania: a statistical meta-analysis. Eur Neuropsychopharmacol
1996; 6(3):245–52
15. Calabrese JR, Bowden CL, McElroy SL, et al. Spectrum of
activity of lamotrigine in treatment-refractory bipolar disorder.
Am J Psychiatry 1999;156(7):1019–23
16. Letterman L, Markowitz JS. Gabapentin: a review of pub-
lished experience in the treatment of bipolar disorder and other
psychiatric conditions. Pharmacotherapy 1999;19(5):565–72
17. Calabrese JR, Bowden CL, Sachs GS, et al. A double-blind
placebo-controlled study of lamotrigine monotherapy in out-
patients with bipolar I depression. Lamictal 602 Study Group.
J Clin Psychiatry 1999;60(2):79–88
18. Prien RF, Kupfer DJ, Mansky PA, et al. Drug therapy in the
prevention of recurrences in unipolar and bipolar affective dis-
orders. Report of the NIMH Collaborative Study Group com-
paring lithium carbonate, imipramine, and a lithium
carbonate–imipramine combination. Arch Gen Psychiatry
1984;41(11):1096–104
19. Sachs GS, Lafer B, Stoll AL, et al. A double-blind trial of
bupropion versus desipramine for bipolar depression. J Clin
Psychiatry 1994;55(9):391–3
20. Calabrese JR, Markovitz PJ, Kimmel SE, et al. Spectrum of
efficacy of valproate in 78 rapid-cycling bipolar patients. J Clin
Psychopharmacol 1992;12(Suppl 1):S53–6
21. Calabrese JR, Fatemi SH, Kujawa M, et al. Predictors of
response to mood stabilizers. Clin Psychopharmacol 1996;16
(2 Suppl 1): S24–31
22. Bowden CL, Calabrese JR, McElroy SI, et al. The efficacy of
lamotrigine in rapid cycling and non-rapid cycling patients
with bipolar disorder. Biol Psychiatry 1999;45(8):953–8
23. Bowden CL, Swann AC, Calabrese JR, et al. Maintenance
clinical trials in bipolar disorder: design implications of the
divalproex-lithium-placebo study. Psychopharmacol Bull 1997;
33(4):693–9
24. Denicoff KD, Smith-Jackson EE, Bryan AL, et al. Valproate
prophylaxis in a prospective clinical trial of refractory bipolar
disorder. Am J Psychiatry 1997;154(10):1456–8
25. Solomon DA, Keitner GI, Ryan CE, et al. Lithium plus
valproate as maintenance polypharmacy for patients with bi-
polar I disorder: a review. Clin Psychopharmacol 1998;18(1):
38–49
26. Denicoff KD, Smith-Jackson EE, Disney ER, et al. Com-
parative prophylactic efficacy of lithium, carbamazepine, and
the combination in bipolar disorder. J Clin Psychiatry
1997;58(11):470–8
27. Greil W, Kleindienst N. The comparative prophylactic
efficacy of lithium and carbamazepine in patients with bipolar
I disorder. Int Clin Psychopharmacol 1999;14(5):277–81
28. Tohen M, Sanger TM, McElroy SL, et al. Olanzapine versus
placebo in the treatment of acute mania. Olanzapine HGEH
Study Group. Am J Psychiatry 1999;156(5):702–9
29. Tohen M, Zarate CA Jr, Centorrino F, et al. Risperidone in
the treatment of mania. J Clin Psychiatry 1996;57(6):249–53
30. Segal J, Berk M, Brook S. Risperidone compared with both
lithium and haloperidol in mania: a double-blind randomized
controlled trial. Clin Neuropharmacol 1998;21(3):176–80
31. Ghaemi SN, Katzow JJ. The use of quetiapine for treatment-
resistant bipolar disorder: a case series. Ann Clin Psychiatry
1999;11(3):137–40
32. Suppes T, Webb A, Paul B, et al. Clinical outcome in a
randomized 1-year trial of clozapine versus treatment as usual
for patients with treatment-resistant illness and a history of
mania. Am J Psychiatry 1999;156(8):1164–9
33. Mukherjee S, Sackeim HA, Scnurr DB. Electroconvulsive
therapy of acute manic episodes: a review. Am J Psychiatry
1994;151:249–276
34. Rush AJ, Rago WV, Crismon ML, et al. Medication treat-
ment for the severely and persistently mentally ill: the Texas
Medication Algorithm Project. J Clin Psychiatry 1999;60(5):
284–91
35. Dennehy EB, Suppes T. Medication algorithms for bipolar
disorder. J Pract Psychiatry Behav Health 1999;5(3):142–52

Page 14

• A POSTGRADUATE MEDICINE SPECIAL REPORT • APRIL 2000
14
Expert Consensus Guideline Series
Rapid cycling,
currently manic
ACUTE TREATMENT OF MANIA
Psychotic Mixed/dysphoric
Severe
residual symptoms
Euphoric†
Mild/moderate
residual symptoms
*
Dvp
Li or Dvp Dvp
Dvp or Li
+
AAp or CAp
change Ap
or add Bzd
Dvp + Li
change Ap
or combine
AAp + CAp
offer ECT
Dvp + Li + Cbz
or add Clz*
*
Gbp or Tpr
add Bzd
add AAp
Gbp or Tpr
*
*
*
Gbp
*
Lmg
*
Clz
offer ECT
Clz or Gbp
or Tpr
Dvp + Li
change AAp
Dvp + Li + Cbz
Add Li or Cbz
Add AAp
Dvp + Li + Cbz
offer ECT
Tpr or T4 or
calcium channel
blocker
Legend
AAp atypical antipsychotic
Ad
antidepressant
Ap
antipsychotic
Bzd
benzodiazepine
CAp conventional antipsychotic
Cbz
carbamazepine
Clz
clozapine
*When adding second-line medications, the clinician may wish to discontinue 1 or
more of the previous medications. Clinicians should avoid combining carbamazepine
and clozapine.
Dvp
ECT electroconvulsive therapy
Gbp gabapentin
Lmg lamotrigine
Li
lithium
T3
triiodothyronine
T4
L-thyroxine
Tpr
topiramate
divalproex
†Recommendations for hypomania are essentially the same as for euphoric mania, but with
less support for the use of antipsychotics.

Page 15

APRIL 2000 • A POSTGRADUATE MEDICINE SPECIAL REPORT •
15
MEDICATION TREATMENT OF BIPOLAR DISORDER 2000
Not rapid cycling, not psychotic,‡
history of at least 1 manic episode
Rapid cycling,
currently depressed
ACUTE TREATMENT OF BIPOLAR DEPRESSION
Breakthrough episode
On Li
Cycles
more
No
response
On Dvp
Stays
depressed
maximize
Dvp
maximize
Li
change Ad
add Dvp
or Lmg
add Dvp
or Lmg
change Ad
T3
*
*
*
other mood stabilizers
AAp
Clz, stimulant,
or phototherapy
offer ECT if severe
add Dvp
or Lmg
or AD
add Li
Dvp
add Li
or Cbz
or Lmg
stop Ad
or change
class of Ad
*
change Ad
*
T3 or T4
or AAp
*
Gbp or Clz
or photo-
therapy
T4 or AAp
Gbp or Clz
*
add Ad
offer ECT
add Ad
or Lmg
add or change Ad
First episode of depression,
currently on no medication
MildSevere
Partial
response
Li or Dvp
+ Ad
add Li
if on Dvp
Li
Li + Ad
‡For psychotic depression, include an AAp in initial
treatment; offer ECT at any point.