`
`Clinical Practice Guidelines for Bipolar Disorder
`From the Department of Veterans Affairs
`
`Mark S. Bauer, M.D.; Ann M. Callahan, M.D.; Chowdary Jampala, M.D.;
`Frederick Petty, Ph.D., M.D.; Martha Sajatovic, M.D.; Vicky Schaefer, R.N.;
`Byron Wittlin, M.D.; and Barbara J. Powell, Ph.D.
`
`© C
`
`Background: For the last several years, the
`Department of Veterans Affairs (VA) has been
`involved in the development of practice guide-
`lines for major medical, surgical, and mental dis-
`orders. This article describes the development and
`content of the VA-Clinical Practice Guidelines for
`Bipolar Disorder, which are available in their en-
`tirety on the Journal Web site (http://www.
`psychiatrist.com).
`Method: A multidisciplinary work group com-
`posed of content experts in the field of bipolar
`disorder and practitioners in general clinical prac-
`tice was convened by the VA’s Office of Perfor-
`mance and Quality and the Mental Health Strate-
`gic Health Group. The work group was instructed
`in algorithm development and methods of evi-
`dence evaluation. Draft guidelines were devel-
`oped over the course of 6 months of meetings and
`conference calls, and that draft was then sent to
`nationally prominent content experts for final
`critique.
`Results: The Bipolar Guidelines are part of the
`family of the VA Clinical Guidelines for Manage-
`ment of Persons with Psychosis and consist of
`explicit algorithms supplemented by annotations
`that explain the specific decision points and their
`basis in the scientific literature. The guidelines
`are organized into 5 modules: a Core Module for
`diagnosis and assignment to mood state plus 4
`treatment modules (Manic/Hypomanic/Mixed
`Episode, Bipolar Depressive Episode, Rapid Cy-
`cling, and Bipolar Disorder With Psychotic Fea-
`tures). The modules specify particular diagnostic
`and treatment tasks at each step, including both
`somatotherapeutic and psychotherapeutic inter-
`ventions.
`Conclusion: The VA Bipolar Guidelines are
`designed for easy clinical reference in decision
`making with individual patients, as well as for use
`as a scholarly reference tool. They also have util-
`ity in training activities and quality improvement
`programs.
`(J Clin Psychiatry 1999;60:9–21)
`
`opyright 1999 Physicians Postgraduate Press, Inc.
`
`sions were made according to physician experience, tradi-
`tion, and training. However, as economic limitations and
`consumer awareness have increased, better assessment of
`treatments and outcomes in general clinical practice has
`become necessary. Practice guidelines represent one type
`of effort to address this need by articulating parameters
`for optimal clinical practice based on available scientific
`evidence and generally accepted clinical opinion.
`In medicine and surgery, the need for practice guide-
`lines has been apparent for at least 15 years, when major
`unexplained variations in the rates of common surgical
`procedures were reported across neighboring cities.1 The
`implementation of standardized guidelines has been one
`method used to reduce such variability.2,3 By contrast,
`psychiatry has only recently begun to document and ex-
`amine variability in clinical practice. One of the few stud-
`ies on this issue was conducted by Fortney et al.4 in the
`Department of Veterans Affairs (VA), who demonstrated a
`4-fold variation in length of inpatient stay for depression
`
`J Clin Psychiatry 60:1, January 1999
`
`9
`
`Received Jan. 1, 1998; accepted July 28, 1998. From the Department
`of Veterans Affairs (VA) Medical Centers, Providence, R.I. (Dr. Bauer),
`Bronx, N.Y. (Dr. Callahan), Columbus, Ohio (Dr. Jampala), Dallas, Tex.
`(Dr. Petty), Cleveland, Ohio (Dr. Sajatovic), St. Cloud, Minn. (Ms.
`Schaefer), San Francisco, Calif. (Dr. Wittlin), and the Psychology Service,
`Kansas City VA Medical Center, Kansas (Dr. Powell).
`Partial funding for preparation of this manuscript was provided by VA
`Health Services Research and Development grant DEV-97-015 (Dr. Bauer)
`and research support from the VA Medical Research Service and the John
`Schermerhorn Fund (Dr. Petty). No financial support for this endeavor was
`derived from industry or professional guild funds.
`Although the Bipolar Guidelines were developed under the auspices of
`the Department of Veterans Affairs, the opinions expressed by the authors
`in this report are their own and do not necessarily reflect the official
`position of the VA.
`The authors acknowledge the contributions to guideline development
`of the other members of the VA Bipolar Disorders Guidelines Work Group:
`Marcia Esquibel, R.N.; Jennifer Garrett, R.R.A.; Sandra Kaufman,
`L.C.S.W.; Linda Ogle, R.N.; Gillermo Olivos, M.D.; Mark Prohaska,
`Ph.D.; and Rodney Haug, Ph.D.
`The complete text of the Clinical Practice Guidelines for Bipolar
`Disorder From the Department of Veterans Affairs is available via the
`Internet (http://www.psychiatrist.com).
`Reprint requests to: Mark S. Bauer, M.D., Department of Veterans
`Affairs Medical Center-116A, Providence, RI 02908-4799 (e-mail:
`Mark_Bauer@brown.edu).
`
`Ccentury. A generation ago, diagnoses and treatment deci-
`
`linical practice guidelines represent a profound
`paradigm shift as U.S. health care enters the 21st
`
`One personal copy may be printed
`
`1 of 13
`
`Alkermes, Ex. 1069
`
`
`
`Bauer et al.
`
`© C
`
`across VA medical centers. This variability could not be
`explained by either case mix or other patient-related fac-
`tors. Thus, the VA is likely to provide an opportunity to
`study and standardize general clinical practice for com-
`mon mental health problems.
`The VA also provides an ideal, and important, system
`in which to develop and study the impact of mental health
`practice guidelines on general clinical practice. First, VA
`clinicians responsible for making psychiatric treatment
`decisions are trained in a broad spectrum of theoretical
`orientations, thus making it likely that many variations in
`practice patterns such as the above4 are based on individ-
`ual factors.
`Second, the VA serves a large number of seriously men-
`tally ill veterans who, as a group, are consumers of large
`amounts of services, making optimal treatment of this
`population a high priority for the VA system. For instance,
`between 405,000 and 630,000 veterans suffer from serious
`mental illness, and about 326,000 of these veterans use VA
`services each year.5 These seriously mentally ill veterans
`are 5 times more likely to use VA services than veterans in
`the general population. During fiscal year 1993, the VA
`provided 4 million days of inpatient care for these indi-
`viduals at a cost of approximately $1.3 billion, and 4.5 mil-
`lion outpatient visits costing $225 million.6 Further, the
`number of veterans treated in outpatient settings has in-
`creased by nearly 20% between 1990 and 1995.5
`Third, the VA system is centralized and hierarchical
`and maintains an extensive automated data management
`system. These characteristics make it feasible both to
`implement systemwide changes effectively in clinical
`practice and to monitor their results.
`The VA has recognized 3 varieties of clinical guidelines
`as potentially useful: Clinical Practice Guidelines, Clini-
`cal Algorithms, and Clinical Pathways.7 Clinical Practice
`Guidelines are statements that assist both the practitioner
`and patient in making the best decisions about appropriate
`health care in specific circumstances. They take the form
`of explicit recommendations for the performance or exclu-
`sion of specific procedures or services. Clinical Algo-
`rithms, incorporated into Clinical Practice Guidelines, are
`explicit decision tools in the form of flow charts or deci-
`sion trees. They systematically guide the user through a
`series of steps that describe key elements of treatment, e.g.,
`diagnosis, therapeutic interventions, time and/or length of
`treatment. This type of algorithm is the core of the VA Bi-
`polar Guidelines. Clinical Pathways are locally developed
`management tools that are based on systemwide Clinical
`Practice Guidelines and Algorithms. They define key pro-
`cesses and events, which are important to the day-to-day
`management of care in a given environment.
`To date, the VA has developed algorithm-based guide-
`lines for several common health problems of veterans, in-
`cluding heart disease, chronic pulmonary disease, and
`common surgical diagnoses (available through the VA Of-
`
`opyright 1999 Physicians Postgraduate Press, Inc.
`
`fice of Performance and Quality). The first guideline de-
`veloped for a major mental illness was for major depres-
`sive disorder and was completed in 1996.8 Several months
`later, working groups were convened to establish treat-
`ment guidelines for the major psychoses.9 This document
`was divided into 4 individual sections on organic psycho-
`ses, schizophrenia, bipolar disorder, and psychosocial re-
`habilitation. The VA Bipolar Guidelines from this family
`of guidelines are the subject of this review.
`The purpose of this article is to introduce readers to
`the Bipolar Guidelines and to describe their empirically
`based development. The algorithms are presented in their
`entirety, with an overview outlining the most salient or
`controversial decision points. The entire text of the Bi-
`polar Guidelines, comprised of over 50 pages of al-
`gorithms and annotations, is available on the Journal
`Web site (http://www.psychiatrist.com). Comparison with
`other major guidelines for bipolar disorder is found in the
`Discussion section of this article.
`
`METHOD
`
`Overview of the Developmental Process
`for VA Mental Health Guidelines
`The VA Office of Performance and Quality and the
`Mental Health Strategic Health Care Group coordinated
`the development of Major Depressive Disorder8 and Psy-
`choses Guidelines,9 with the Bipolar Guidelines a subset
`of the latter. The principles for development of each of the
`guidelines were identical. With support from the VA’s Ex-
`ternal Peer Review program, multidisciplinary work
`groups were created to work on each of the guidelines.
`Each group consisted of facilitators who were experi-
`enced in algorithm development and decision-making
`processes, content experts, and professionals in general
`clinical practice in VA, university, and/or private practice
`venues. The consulting group conducted an extensive lit-
`erature search using bipolar affective disorder, schizoaf-
`fective disorder, and related terms, and recent articles
`were provided to team members for use in the guideline
`development. Consumer input was solicited from clients
`and family members by conducting focus groups at 5
`medical centers across the nation.
`The working groups first met in November 1996 for a
`2-day orientation and education session. All members re-
`ceived instruction in formal algorithm methodology and
`group decision-making methods (e.g., nominal group pro-
`cess, delphi method). The group was also instructed in the
`U.S. Agency for Health Care Policy and Research
`(AHCPR)10 and American College of Cardiologists and
`American Heart Association (ACC/AHA)11 methods for
`evidence evaluation, as summarized in Table 1. The
`groups were oriented to the framework for the final prod-
`uct, which was to consist of a set of freestanding algo-
`rithms supplemented by a series of text annotations that
`
`One personal copy may be printed
`
`10
`
`J Clin Psychiatry 60:1, January 1999
`
`2 of 13
`
`Alkermes, Ex. 1069
`
`
`
`Clinical Practice Guidelines for Bipolar Disorder
`
`Given the complexity of bipolar disorder, each content
`expert was given responsibility for each of several key
`areas, which were to be developed into separate but
`linked algorithms. In addition to the core diagnostic mod-
`ule, which was developed by the entire group, the 4 key
`areas designated for individual modules were Manic/
`Hypomanic/Mixed Episode, Bipolar Depressive Episode,
`Rapid Cycling, and Bipolar with Psychotic Features (in-
`cluding schizoaffective disorder). The content expert so-
`licited assistance from other members, such as perform-
`ing literature searches, critiquing, editing, and revising.
`In addition to the 2 face-to-face meetings, approximately
`16 hours of conference calls were devoted to these activi-
`ties. In addition, group members communicated with
`each other as needed via e-mail, fax, and personal tele-
`phone calls.
`The resultant Bipolar Guidelines draft was then
`sent to 10 content experts (predominantly non-VA), who
`provided written or verbal critiques. Version 1.0 was re-
`leased to the field in September 1997 as part of the Clini-
`cal Guidelines for Management of Persons with Psycho-
`ses,9 which also included the other 3 guidelines noted
`above. Minor text and algorithm corrections and clarifi-
`cations were then incorporated in the subsequent several
`months, with Version 1.1 (the version summarized in this
`article) released in early 1998.
`The results section of this article serves several func-
`tions. First, it provides an overview of the structure and
`use of the Bipolar Guidelines; these are similar to the
`other VA guidelines for mental illnesses. Second, the con-
`tent of the Bipolar Guidelines is summarized. Third, any
`particularly controversial or important point is noted and
`briefly reviewed. A more extensive review of these issues
`can be found in the annotations of the guidelines them-
`selves, located in their entirety on the Journal Web site;
`reference to specific annotations in the text of this article
`points the reader to the appropriate section of the appro-
`priate module of the guidelines for further review.
`
`RESULTS
`
`Core Diagnostic Module
`The Core Module (Module D) is intended to guide cli-
`nicians in assessing a patient’s current mood state and
`episode history so that individuals with suspected bipolar
`disorder can be routed into the appropriate algorithm for
`future assessment and treatment. It is assumed that indi-
`viduals entering the Core Module have been screened in
`the base module of the overall Psychoses Guidelines
`for (1) the presence of a mood disturbance and (2) the
`absence of secondary medical or substance abuse/
`dependence that might account for the mood disturbance.
`On the basis of the current episode, individuals are
`triaged through a series of specific algorithm steps into
`1 of the 4 diagnosis-specific modules. Individuals with
`
`One personal copy may be printed
`
`Table 1. Classification of Evidence and Recommendations
`According to the AHCPR and ACC/AHA Systemsa
`AHCPR10 Classification of Strength of Evidence
`Class A: Randomized controlled trials
`Class B: Well-designed clinical studies
`Class C: Panel consensus
`ACC/AHA11 Classification of Strength of Recommendations
`Class I: Usually indicated, always acceptable, and considered useful
`and effective
`Class II: Acceptable, of uncertain efficacy, and may be controversial
`IIa: Weight of evidence in favor of usefulness/efficacy
`IIb: Not well established by evidence, can be helpful and
`probably not harmful
`Class III: Not indicated and may be harmful
`aAbbreviations: AHCPR = U.S. Agency for Health Care Policy and
`Research, ACC/AHA = American College of Cardiologists and
`American Heart Association.
`
`© C
`
`would include expansion of the recommendations and
`scholarly reviews of evidence. Thus, the content is similar
`to that of the AHCPR guidelines, but the algorithms and
`text were to be separated for ease of use. During the initial
`meeting, the work groups responsible for developing the 4
`psychosis guidelines also met separately to formulate
`plans and strategies for how to best accomplish their task
`of having a draft algorithm completed by February 1997.
`The second and final face-to-face meeting of all partici-
`pants took place in March 1997. During this 2-day meet-
`ing, the individual draft guidelines were reviewed and cri-
`tiqued by all of the groups working on the psychosis
`guidelines in order to identify and reconcile interface and
`coordination issues among the guidelines.
`During the entire process of algorithm development,
`the empirical basis for their construction was recorded in
`a series of text annotations that were associated with the
`relevant algorithm steps. These annotations were used to
`expand on instructions presented in skeletal form in the
`algorithm itself, to provide references for further informa-
`tion, and, importantly, to present the scientific basis for
`each specific algorithm step. In this last endeavor, the
`work groups recorded their evaluation of the scientific
`evidence based on AHCPR standards and indicated the
`confidence of the resulting recommendation based on
`ACC/AHA standards. The primary source references that
`served as the basis for the recommendations were typi-
`cally summarized in the form of evidence tables for easy
`reference by the users.
`
`opyright 1999 Physicians Postgraduate Press, Inc.
`
`Specific Developmental Process
`for the Bipolar Guidelines
`The Bipolar Guidelines work group consisted of both
`content experts and practitioners in general clinical prac-
`tice. Individuals were by design drawn from several disci-
`plines (7 M.D.s, 3 Ph.D.s, 3 R.N.s, and 1 L.C.S.W.) and
`was led by M.D. and Ph.D. cochairs. The majority of par-
`ticipants were not acquainted and/or had not worked to-
`gether prior to the initial meeting.
`
`J Clin Psychiatry 60:1, January 1999
`
`11
`
`3 of 13
`
`Alkermes, Ex. 1069
`
`
`
`Bauer et al.
`
`Figure 1. Bipolar Disorder Core Module (Module D)
`1
`
`Persons with
`signs, symptoms,
`or history
`of mood disorders
`
`4
`
`Y
`
`Meets
`DSM-IV
`criteria for
`current manic,
`hypomanic,
`or mixed
`episode?
`[D]
`
`5
`
`Y
`
`Person with
`Rapid Cycling?
`[E]
`
`Y
`
`Go to
`Rapid Cycling
`Module G
`
`N
`
`N
`
`11
`
`Y
`
`Treat for
`cyclothymia
`[C]
`
`6
`
`Meets DSM-IV
`criteria for past
`manic, hypomanic,
`or mixed episode?
`[D]
`
`Y
`
`opyright 1999 Physicians Postgraduate Press, Inc.
`
`Evaluate for
`concurrent
`substance
`abuse problem
`and institute
`Module C
`
`© C
`
`Meets
`DSM-IV
`criteria for
`current or past
`major
`depressive
`episode?
`[A]
`
`N
`
`Meets
`DSM-IV
`criteria for
`cyclothymia?
`[B]
`
`N
`
`2
`
`3
`
`10
`
`12
`
`12
`
`Person with
`Rapid Cycling?
`[E]
`
`Y
`
`N
`
`Go to
`Current Bipolar
`Depressive
`Episode
`Module F
`
`Go to
`Rapid Cycling
`Module G
`
`One personal copy may be printed
`
`Go to
`Current
`Manic/Hypomanic/
`Mixed Episode
`Module E
`
`8
`
`Y
`
`Meets DSM-IV
`criteria for current
`major depressive
`episode?
`
`7
`
`9
`
`N
`
`N
`
`Go to
`Guideline for
`Major Depressive
`Disorder
`
`Institute
`prophylaxis and
`consider
`psychosocial
`rehabilitation
`[F]
`
`Meets
`DSM-IV
`criteria for current
`or past manic,
`hypomanic, or
`mixed episode?
`[D]
`
`13
`
`Y
`
`Meets
`DSM-IV
`criteria for
`currentmanic,
`hypomanic, or
`mixed episode?
`[D]
`
`14
`
`Y
`
`Person with
`Rapid Cycling?
`[E]
`
`Y
`
`N
`
`15
`
`Go to
`Guideline for
`Major Depressive
`Disorder to rule
`out Dysthymia
`
`N
`
`Institute
`prophylaxis
`and consider
`psychosocial
`rehabilitation
`[F]
`
`N
`
`Go to
`Current Manic/
`Hypomanic/Mixed
`Episode
`Module
`E
`
`Go to
`Rapid Cycling
`Module G
`
`J Clin Psychiatry 60:1, January 1999
`
`4 of 13
`
`Alkermes, Ex. 1069
`
`
`
`Clinical Practice Guidelines for Bipolar Disorder
`
`antidepressants, these medications should be discontin-
`ued. If there is a history of response to a previous mood-
`stabilizing regimen that has been stopped, that regimen
`should be restarted; if there has been no previous treat-
`ment with a mood stabilizer, one should be initiated (an-
`notation J). If, after 3 weeks of treatment, there is no re-
`sponse to the optimal dose of the initial mood stabilizer,
`or if there is a clear history of nonresponse to the current
`mood stabilizer, the guidelines recommend starting a dif-
`ferent mood stabilizer and tapering off the initial one (an-
`notation K). If there is only a partial response, or if none
`of the mood stabilizers prove to be efficacious, a combi-
`nation of different mood stabilizers (preferably lithium
`plus one of the anticonvulsants) is recommended treat-
`ment. In the event that mood stabilizers, either singly or in
`combination, do not control the acute manic symptoms,
`other agents with possible antimanic properties (e.g.,
`clozapine, lamotrigine, or gabapentin) should be tried (an-
`notation K). Once the acute manic symptoms are under
`control, prophylactic treatments and psychoeducation
`should be initiated, along with psychosocial rehabilitation
`if indicated (annotation F).
`One of the more controversial aspects of the guidelines
`is their assessment of the relative strength of evidence for
`the available mood stabilizers—lithium, valproate, and
`carbamazepine—as antimanic agents. Based on the
`strength of evidence review of the literature, lithium is
`recommended as the first-line agent for both acute
`antimanic and prophylactic use for treating manic and
`mixed episodes, although some recent evidence indicates
`that valproate may be more effective than lithium in
`mixed episodes (annotation J). Also of relevance is the
`fact that lithium is the only agent to date for which effi-
`cacy has been established as a prophylactic agent for
`management after the acute episode has resolved, adding
`to the strength of recommendation that lithium should be
`the first-line antimanic agent.
`While there is currently considerable enthusiasm for
`using the anticonvulsant valproate as a first-line acute
`treatment, only a relatively small number of controlled tri-
`als exist compared with the more extensive data on lithi-
`um. Those data that do exist indicate that its overall effi-
`cacy is comparable with that of lithium.13,14 Valproate may
`be particularly useful in treating individuals with mania
`who fail to respond to lithium14,15 or individuals with
`mania with concurrent depressive features (mixed
`manics).15,16 Evidence for the efficacy of carbamazepine in
`treating acute mania is less extensive than that for lithium.
`Electroconvulsive therapy (ECT) may also be efficacious
`as a treatment for acute mania, have a role in the treatment
`of selected individuals, and be used as a maintenance
`treatment if there are compelling reasons for not using the
`mood-stabilizing medications. Clearly, though, additional
`controlled studies are in progress, and this issue will have
`to be revisited in later revisions of the guidelines.
`
`opyright 1999 Physicians Postgraduate Press, Inc.
`
`© C
`
`suspected bipolar disorder who are found to have major
`depressive disorder or dysthymia are screened out and re-
`ferred to the VA Major Depression Guidelines.8
`The Core Module algorithm (Figure 1) serves as a pro-
`totype for the algorithms for the other 4 modules; thus, it
`is explained here in somewhat greater detail. The starting
`place for the algorithm is an oval called the “clinical state
`box,” which describes the presenting problem. The algo-
`rithm then guides the reader through a series of yes/no
`decision-making steps (hexagons). Steps that require
`some clinical action for all individuals are denoted as “do
`boxes” (rectangles). The “go to” circles at the various ter-
`minal steps of the algorithms indicate that DSM-IV12-
`based diagnostic criteria for a particular condition have
`been met, and the user is then routed to the appropriate
`diagnosis-specific module. An alphabetical letter appear-
`ing within a box indicates that there is an accompanying
`text annotation, as described in the Method section above.
`The Core Module algorithm is sufficiently comprehen-
`sive and flexible to meet clinicians’ needs in assessing all
`individuals with suspected bipolar spectrum conditions.
`Specifically, it is designed to triage individuals who
`present for treatment with or without current medications,
`to evaluate individuals with cyclothymia, and to accom-
`modate individuals with bipolar disorder who present for
`treatment while not in a major mood episode. With regard
`to this last group, the relevant annotations (annotations E
`and F) indicate the necessity of long-term treatment with
`mood stabilizers for individuals with bipolar disorder. Ac-
`cordingly, the data for prophylactic efficacy of the avail-
`able agents are reviewed in detail along with a discussion
`of the costs and benefits.
`The Core Module also presents an overview of psycho-
`social interventions for bipolar disorder. These include
`psychoeducation, formal psychotherapy, and psychoso-
`cial rehabilitation. The guidelines specify psychoedu-
`cation for all individuals and formal psychotherapy or
`psychosocial rehabilitation for selected individuals, de-
`pending on the clinical situation.
`
`Manic, Hypomanic, or Mixed Episode Module
`As seen in Figure 2, the clinical state oval indicates
`that individuals in the Manic, Hypomanic, or Mixed Epi-
`sode Module (Module E) meet DSM-IV12 criteria for one
`of these episodes and are free of causative general medi-
`cal condition, substance intoxication, or substance with-
`drawal. The clinician must then determine the appropriate
`setting of care, initiate psychoeducational tasks, evaluate
`for other psychosocial interventions, and ensure normal
`thyroid functioning. Subsequent actions involve evaluat-
`ing the status of current medications, making medication
`adjustments, and monitoring additional symptoms such as
`insomnia and anxiety.
`The guidelines recommend that if an individual is in
`a manic, hypomanic, or mixed state and is receiving
`
`One personal copy may be printed
`
`J Clin Psychiatry 60:1, January 1999
`
`13
`
`5 of 13
`
`Alkermes, Ex. 1069
`
`
`
`Bauer et al.
`
`Figure 2. Current Manic, Hypomanic, or Mixed Episode (Module E)
`1
`
`Person meets DSM-IV criteria for manic, hypomanic,
`or mixed episode and is free of any general medical
`condition, intoxication, or withdrawal
`
`2
`
`Determine appropriate setting for care
`[A]
`Begin psychoeducational tasks and evaluation for ongoing
`psychotherapy and/or psychosocial rehabilitation
`[B]
`Ensure normal thyroid function
`[C]
`
`3
`
`4
`
`5
`
`© C
`
`Is person taking
`antidepressants or
`mania-inducing
`medication?
`
`N
`
`Is person
`experiencing
`insomnia/agitation, or
`is anxiety present?
`[G]
`
`Y
`
`Y
`
`Reduce/stop
`antidepressants
`[D]
`
`N
`
`7
`
`Consider
`benzodiazepine
`treatment
`[H]
`
`opyright 1999 Physicians Postgraduate Press, Inc.
`
`Response?
`[E]
`
`8
`
`Institute prophylaxis
`and consider
`psychosocial
`rehabilitation
`[F]
`
`Go to Bipolar
`with Psychotic
`Features
`Module H
`
`Go to
`Box 19
`
`14
`
`Response?
`[E]
`
`Y
`
`Y
`
`6
`
`9
`
`10
`
`12
`
`15
`
`16
`
`Are psychotic
`features present?
`[I]
`
`N
`
`N
`
`Was person
`previously
`treated with mood
`stabilizers?
`
`Y
`
`Response to
`previous mood
`stabilizer
`treatment
`regimen?
`
`N
`
`Initiate and optimize
`mood stabilizer
`Reassess in
`2–3 weeks
`[J]
`
`One personal copy may be printed
`
`Y
`
`11
`
`13
`
`N
`
`Y
`
`Initiate and optimize
`mood stabilizer
`Assess in 2–3 weeks
`[J]
`
`Restart previous
`regimen
`
`N
`
`Y
`
`Response?
`[E]
`
`N
`
`17
`
`Add/change mood
`stabilizer until stable or
`consider alternative
`therapy
`[K]
`
`18
`
`Taper or discontinue
`neuroleptic and
`benzodiazepines
`(if applicable)
`[N]
`
`19
`
`Continue prophylaxis
`and consider
`psychosocial
`rehabilitation
`[F]
`
`14
`
`J Clin Psychiatry 60:1, January 1999
`
`6 of 13
`
`Alkermes, Ex. 1069
`
`
`
`Clinical Practice Guidelines for Bipolar Disorder
`
`Figure 3. Current Bipolar Depressive Episode (Module F)
`
`1
`
`Person meets DSM-IV criteria for bipolar
`depressive episode free of any general medical
`condition, intoxication, or withdrawal.
`
`2
`
`Determine appropriate setting for care
`[A]
`Begin psychoeducational tasks and evaluation for ongoing
`psychotherapy and/or psychosocial rehabilitation
`[B]
`Ensure normal thyroid function
`[C]
`
`© C
`
`Go to
`Bipolar with
`Psychotic
`Features
`Module H
`
`opyright 1999 Physicians Postgraduate Press, Inc.
`
`Is person
`experiencing
`anxiety, agitation, or
`insomnia?
`[D]
`N
`
`Are psychotic
`features
`present?
`[F]
`
`4
`
`6
`
`Y
`
`Y
`
`Consider
`benzodiazepine
`treatment
`[E]
`
`Treat the
`psychosis
`
`N
`
`Initiate/optimize
`mood stabilizer.
`Reassess in 2–4 weeks
`[G]
`
`9
`
`N
`
`Institute prophylaxis
`and consider psycho-
`social rehabilitation
`[H]
`
`Continued
`symptoms?
`
`Y
`Initiate lithium or
`antidepressant
`treatment. Reassess
`response in 2–4 weeks
`[I]
`
`One personal copy may be printed
`
`3
`
`5
`
`7
`
`8
`
`10
`
`11
`
`14
`
`16
`
`Assess need for
`continued anti-
`depressant treatment
`[K]
`
`13
`
`Initiate prophylaxis and
`consider psychosocial
`rehabilitation
`[H]
`
`Augment or combine
`antidepressants
`[M]
`
`Y
`
`Y
`
`12
`
`15
`
`17
`
`Response?
`[J]
`
`N
`
`Partial
`response?
`[L]
`
`N
`
`Switch
`antidepressants
`[N]
`
`Response?
`[J]
`
`N
`
`20
`
`Utilize ECT or
`alternative therapies
`until response
`[O]
`
`Y
`
`18
`
`Assess need for
`continued anti-
`depressant treatment
`[K]
`
`19 Continue prophylaxis
`and consider psycho-
`social rehabilitation
`[H]
`
`J Clin Psychiatry 60:1, January 1999
`
`15
`
`7 of 13
`
`Alkermes, Ex. 1069
`
`
`
`Bauer et al.
`
`© C
`
`Bipolar Depressive Episode Module
`The algorithm for the treatment of individuals with bi-
`polar depressive episode (Module F) is presented in Fig-
`ure 3. The guidelines recommend that the first step in the
`treatment of acute bipolar depression is to initiate, or, if
`the patient is already being treated, to optimize the current
`mood stabilizer (annotation G). Controlled studies have
`shown that lithium is the most effective mood stabilizer
`for the treatment of acute bipolar depression.17 While less
`extensive data exist for carbamazepine, this agent may be
`effective for patients who fail to respond to lithium.18,19
`There are no controlled data supporting the use of valpro-
`ate in the treatment of acute bipolar depression, but it has
`been suggested that valproate may be effective for depres-
`sive symptoms associated with mixed states.16
`If depressive symptoms do not improve or are only
`partially responsive after a period of 2 to 4 weeks, the ad-
`dition of lithium should be considered for patients treated
`with carbamazepine or valproate prior to the initiation of
`an antidepressant (annotation I). The combined use of
`lithium and carbamazepine has been shown to enhance
`efficacy in bipolar depression.20 Although possibly help-
`ful, there are fewer data to support the efficacy of lithium
`and valproate or carbamazepine and valproate.
`Antidepressants should be used conservatively when
`treating patients with bipolar disorder. This recommenda-
`tion is based on the increased risk of antidepressant-
`induced mania with which these agents have been associ-
`ated.21 It is likely that all antidepressants can induce
`mania in susceptible patients.22 Furthermore, antidepres-
`sant treatment may have a negative impact on the natural
`course of bipolar disorder by inducing rapid cycling and
`mixed states.21,23 Although it has been reported that the tri-
`cyclic antidepressants (TCAs) have a greater propensity
`to induce mania than the selective serotonin reuptake in-
`hibitors (SSRIs), this report was derived from post hoc
`reanalysis of data collected for different purposes.24 On
`the other hand, data from a small prospective controlled
`trial indicated that bupropion is less likely to induce ma-
`nia than desipramine25; some data from open studies also
`support this finding.18,26
`Despite the inherent risks, it is often necessary to ad-
`minister antidepressants to patients with severe or recur-
`rent bipolar depression. In such cases, the guidelines rec-
`ommend that antidepressants be administered at the
`lowest effective dose for the shortest time possible and
`that a mood stabilizer always be coadministered. In addi-
`tion, frequent evaluation for the emergence of manic
`symptoms is required. Should manic symptoms occur, the
`antidepressant should be reduced or discontinued.
`The relative efficacy of the various antidepressants in
`bipolar depression has not been extensively studied.
`Treatments for which controlled data exist include ECT,
`TCAs, monoamine oxidase inhibitors, and bupropion (an-
`notation I). Some investigators have reported that ECT
`
`One personal copy may be printed
`
`may be the most effective treatment.27 Tranylcypromine
`also may be particularly efficacious, especially for pa-
`tients with hypersomnia and hyperphagia.28 TCAs appear
`to be less effective; controlled studies have shown an
`overall response rate of approximately 55%.27 While clini-
`cal experience suggests that bupropion may have superior
`efficacy in bipolar depression, controlled data are lack-
`ing.29 Finally, there are no published controlled studies of
`the SSRIs or other newer antidepressants in bipolar de-
`pression. However, clinical experience indicates that both
`SSRIs and venlafaxine may be useful treatments.29
`
`Rapid Cycling Module
`The algorithm for the treatment of rapid cycling in in-
`dividuals with bipolar disorder (Module G) is presented in
`Figure 4. Rapid cycling is defined by DSM-IV12 as the oc-
`currence of 4 or more affective episodes of any polarity
`within a 12-month interval. Evidence indicates that indi-
`viduals with rapid cycling respond less well to mood sta-
`bilizers than do other types of bipolar patients. In addi-
`tion, certain treatments such as antidepressants may
`worsen the course of the disorder. Evidence also indicates
`that individuals with rapid cycling should be closely
`monitored for hypothyroidism, which may occur in higher
`frequency in this group compared with those individuals
`without rapid cycling.
`Because of these factors, identification of rapid cycling
`is critical for the effective management of individuals with
`bipolar disorde