Bipolar Disorders 2001: 3: 165–173
`Printed in Ireland. All rights reser6ed
`
`Original Article
`
`Medication prescribing patterns for patients
`with bipolar I disorder in hospital settings:
`adherence to published practice guidelines
`
`Lim PZ, Tunis SL, Edell WS, Jensik SE, Tohen M. Medication pre-
`scribing patterns for patients with bipolar I disorder in hospital set-
`tings: adherence to published practice guidelines.
`Bipolar Disord 2001: 3: 165–173. © Munksgaard, 2001
`
`Objecti7e: The purposes of this paper were to examine the medication
`prescribing patterns for bipolar I disorder in hospital settings and to
`compare them to recently published expert consensus guidelines for
`medication treatment of bipolar disorder.
`
`Methods: Data were obtained from the 1996–2000 CQI+ SM Outcomes
`Measurement System, on patients age 18 or older admitted to psychi-
`atric inpatient units from over 100 medical–surgical hospitals. A total
`of 1864 patients with a primary discharge diagnosis of bipolar I or II
`disorder were identified from a large cohort of hospitalized patients.
`Patient characteristics were assessed at hospital admission and medica-
`tion usage, at discharge. The medication analysis focused on the 1471
`individuals with bipolar I mania or bipolar I depression (with or with-
`out psychotic features), representing 54% and 25% of admitted bipolar
`patients, respectively.
`
`Results: At admission, the typical bipolar patient (mean age 57) had
`experienced a relatively severe and chronic course of illness. The array
`of psychotropic agents used was broad, with no single prescribing pat-
`tern predominant. Only one in three bipolar I (manic or depressed)
`patients with psychotic features was discharged on medications recom-
`mended by expert guidelines as preferred or alternate recommended
`treatment. Absent psychotic features, this dropped to one in six pa-
`tients. Surprising was the relatively high use of antidepressants for pa-
`tients with mania, particularly those without psychotic symptoms.
`
`Conclusions: Results suggest that a substantial proportion of patients
`with bipolar I disorder are discharged from hospitals on medications
`not generally recommended by current practice guidelines.
`
`Prudence Z Lima, Sandra
`L Tunisa, William S Edellb,
`Sarah E Jensikb and
`Mauricio Tohena,c
`a Lilly Research Laboratories, Indianapolis,
`IN, USA, b Mental Health Outcomes,
`Lewisville, TX, USA, c McLean Hospital,
`Harvard Medical School, Belmont, MA,
`USA
`
`Key words: bipolar I disorder – hospital
`settings – medication prescribing patterns
`
`Received 7 May 2000, revised and
`accepted for publication 7 March 2001
`
`Corresponding author: Sandra L Tunis,
`PhD, Lilly Corporate Center, Indianapolis,
`IN 46285, USA, e-mail:
`tunis–sandra@lilly.com
`
`Disease impact
`Bipolar I disorder is a chronic and often devastat-
`ing disease that affects about 1.0% of the general
`population (1–6). Severe physical and psychosocial
`impairments are common (2, 7), which have been
`associated with high rates of medical and psychi-
`atric care, family dysfunction, divorce, substance
`abuse and suicide (8–10). To illustrate this, bipolar
`disorder was ranked sixth among the top ten lead-
`ing worldwide causes of disability-adjusted life
`years in 1990 for persons aged 15–44 years (11).
`
`The direct and indirect economic costs of bipo-
`lar disorder are staggering (12). One decade ago, it
`was estimated that the total costs of the disorder
`exceeded $45 billion in the USA alone (13). More
`recent work suggests that the health care costs for
`bipolar disorder exceed those for depression or
`diabetes (14). The mortality rates for untreated
`bipolar disorder are comparable to many types of
`heart disease and cancer, with rates of death by
`suicide (around 19%) perhaps the highest for any
`psychiatric or medical disorder (9).
`
`165
`
`1 of 9
`
`Alkermes, Ex. 1053
`
`

`
`Lim et al.
`
`Treatment guidelines
`
`The successful management of bipolar I disorder
`depends on early detection, appropriate pharmaco-
`logic and psychosocial treatments, and an under-
`standing of the long-term cyclic, recurrent, and
`relapsing nature of
`the disorder. Recurrent
`episodes of mania and depression, often secondary
`to medication noncompliance (15–18), have a de-
`teriorative effect on patient functioning, response
`to treatment and prognosis (7, 9, 19–22). To
`provide guidance to psychiatrists and other mental
`health professionals about the most appropriate
`treatments for patients with bipolar disorder, vari-
`ous practice guidelines and treatment algorithms
`have been developed. The hope is that such guide-
`lines will reduce unnecessary variation in clinical
`practice, facilitate rational clinical decision-mak-
`ing, improve quality of care, increase efficiency of
`treatment and allow for systematic examination of
`specific treatments on outcomes (23).
`Some guidelines have been developed by panels
`of recognized experts [International Psychophar-
`macology Algorithm Project (24); European Al-
`gorithm Project (25, 26)]. Others have used panels
`of experts combined with additional detailed com-
`mentary from clinicians in the field [American Psy-
`chiatric Association (APA) (8); Canadian Network
`for Mood and Anxiety Treatments (CANMAT)
`(27, 28); Department of Veterans Affairs (29)]. The
`Texas Medication Algorithm Project (TMAP) (23,
`30) was generally based on the 1996 Expert Con-
`sensus Guidelines (31) and the 1994 APA Guideli-
`nes. The 2000 Expert Consensus Guidelines (32)
`were developed most
`rigorously, using survey
`methodology (33) with 65 leading American ex-
`perts on bipolar disorder, and the RAND modified
`Delphi approach for ascertaining expert consensus
`(34).
`For the purpose of this descriptive analysis, ac-
`tual practice patterns for bipolar I mania and bipo-
`lar I depression were compared with the 2000
`Expert Consensus Guidelines (32, 35). Table 1
`summarizes the preferred strategies and alternate
`strategies for pharmacotherapies recommended by
`the 2000 Expert Consensus Guidelines for bipolar I
`mania (with, or without psychotic features) and for
`bipolar I depression (with, or without psychotic
`features).
`The specific mood stabilizers recommended for
`all clinical presentations of bipolar mania or bipo-
`lar depression include lithium, valproate and carba-
`mazepine.
`In addition, a new anticonvulsant
`lamotrigine is recommended for bipolar depression.
`Atypical antipsychotics are preferred by the Expert
`Consensus Guidelines over conventional antipsy-
`
`166
`
`chotics as first-line adjunctive treatment for bipolar
`mania and for psychotic bipolar depression, partic-
`ularly when long-term use (2–5 months following
`resolution of acute episode) is required. Specific
`antidepressant agents recommended for bipolar
`depression include buproprion, the selective sero-
`tonin reuptake inhibitors (SSRIs) and venlaxafine.
`No guidance regarding specific benzodiazepines to
`use with bipolar patients is provided.
`
`Study purpose
`
`While the treatment recommendations specified in
`published guidelines may not address the specific
`and unique needs of every bipolar patient, and are
`not intended to represent standards of medical care
`per se (36), they do provide a helpful clinical
`framework. Describing the psychotropic agents
`prescribed for bipolar patients admitted to a psy-
`chiatric unit in acute-care hospitals throughout the
`USA can be informative regarding current treat-
`ment practices in this important setting. The spe-
`cific objective of
`this
`study was
`to profile
`clinician-prescribing practices for bipolar patients
`within hospital settings, and to delineate the extent
`to which they are consistent with recently pub-
`lished treatment guidelines.
`
`Methods
`Data source
`
`Retrospective analyses were conducted on data
`collected between January 1, 1996 and March 31,
`2000. Data came from the CQI+ SM Outcomes
`
`Table 1. Pharmacologic therapies recommended by Expert Consensus
`Guidelines (32)
`
`Bipolar I mania with psychotic features
`Mood stabilizer and antipsychotic
`Mood stabilizer, antipsychotic and benzodiazepine
`
`Bipolar I mania without psychotic features
`Mood stabilizer alone
`Mood stabilizer and benzodiazepine
`Mood stabilizer and antipsychotic
`
`Bipolar I depression with psychotic features
`Mood stabilizer, antipsychotic and antidepressant
`Electroconvulsive therapy
`Mood stabilizer and antipsychotic
`Mood stabilizer and antidepressant
`
`Bipolar I depression without psychotic features
`Mood stabilizer and antidepressant
`Mood stabilizer alone
`Electroconvulsive therapy
`
`Preferred treatments specified within the Expert Consensus
`Guidelines (32) are italicized. Alternate treatments specified within
`the Expert Consensus Guidelines (32) are not italicized.
`
`2 of 9
`
`Alkermes, Ex. 1053
`
`

`
`Prescribing patterns and adherence to bipolar practice guidelines
`
`Measurement System database developed by Men-
`tal Health Outcomes in Lewisville, TX. The system
`tracks information on patients, age 18 and older,
`admitted to psychiatric inpatient units from over
`100 acute care medical–surgical hospitals in 31
`states.
`Patients with a diagnosis of bipolar I or II
`disorder represented a subset (n=1864 or 7.0%) of
`a large national cohort of patients (n=26559),
`hospitalized in psychiatric treatment programs
`across the USA. Included in the subset were bipo-
`lar patients aged 18–64 years admitted to adult
`psychiatric units (n=869), and patients aged 55
`years or older admitted to geropsychiatric units
`(n=995).
`
`Assessment procedures
`Programs participating in the CQISM+ System
`invite the first 17 patients admitted per month
`(regardless of diagnosis) to complete a series of
`assessments within 72 h of admission, within 48 h
`of discharge and at 3 months (geriatric programs)
`or 6 months (adult programs) following hospital
`discharge. Overall, the assessments include several
`self report, informant report, and clinician/staff-
`rating scales. For this study, demographic charac-
`teristics and clinical status variables of all patients
`with a primary discharge diagnosis of bipolar I or
`bipolar II disorder were selected from an admission
`questionnaire, completed by patients and infor-
`mants. Discharge medications were identified by a
`medication usage questionnaire, which was com-
`pleted by the hospital staff based on the patients’
`medical
`records. Clinician-prescribing patterns
`were analyzed only for the two largest subsets of
`bipolar patients [i.e. patients with bipolar I mania
`(ICD-9 Codes: 296.00–296.06) or with bipolar I
`depression (ICD-9 Codes: 296.50–296.56)].
`
`Results
`Patient demographics and diagnosis
`
`Patient demographics at admission are shown in
`Table 2. Overall, the bipolar group was 63% fe-
`male, predominantly Caucasian (90%), educated at
`a High School level or beyond (73%) and averaged
`57 years old (SD=20; range 18–102). About half
`(49%) were separated, divorced or widowed, with
`only one-third (33%) currently married. Most were
`unemployed or ‘retired’ (85%), and were living in
`their private residence (75%) prior to hospital
`admission.
`Bipolar I mania (n=1007 or 54%) and bipolar I
`depression (n=464 or 25%) (with or without psy-
`chotic features) represented the majority of bipolar
`
`Table 2. Admission assessment of demographic variables for bipolar I and
`II inpatients
`
`Variable
`
`1. Mean age9SD
`
`2. Race
`% Caucasian
`% African American
`
`3. Gender
`% Female
`
`4. Work status
`% Unemployed or ‘retired’
`
`5. Marital status
`% Never married
`% Married
`% Separated, divorced or widowed
`
`6. Living arrangement
`% Living at home (alone or with others)
`% Nursing home
`% Board and care
`% Other (e.g. retirement facility,
`hospital)
`
`7. Type of payora
`% Medicaid
`% Medicare
`% Private insurance
`
`8. Educational status
`% High School graduates or above
`
`Bipolar cohort
`(n=1864)
`
`57920
`
`90
`6
`
`63
`
`85
`
`18
`33
`49
`
`75
`9
`6
`10
`
`19
`57
`28
`
`73
`
`a Response choices within this category were not mutually exclu-
`sive.
`
`patients discharged from the hospital. Smaller
`groups of patients had a primary discharge diagno-
`sis of bipolar I mixed (n=306 or 16%) or bipolar
`II disorder (n=87 or 5%).
`
`Clinical status at hospital admission
`
`The typical patient in our overall bipolar sample
`had experienced a severe and chronic clinical his-
`tory. The most common secondary DSM-IV Axis
`I/II discharge diagnoses documented in the medi-
`cal record were personality disorder (14%) and
`substance-related disorder (11%). The majority
`(81%) had previously been hospitalized at least
`once for psychiatric reasons, with about 40% hav-
`ing been hospitalized four or more times for psy-
`chiatric reasons during their lifetime. In addition,
`approximately one-third reported attempting sui-
`cide at least once within their lifetime.
`
`Discharge medication prescribing patterns
`
`The frequencies (by percentage) of different phar-
`macologic therapies prescribed for bipolar I mania
`
`167
`
`3 of 9
`
`Alkermes, Ex. 1053
`
`

`
`Lim et al.
`
`and bipolar I depression patients, presented sepa-
`rately for those with or without psychotic features,
`are listed in Tables 3 and 4. The vast majority of
`patients (89–98%) were discharged on at least one
`psychotropic agent. However, the array of agents
`used alone or in combination was broad. No single
`pharmacologic therapy was overwhelmingly pre-
`ferred by clinicians in the treatment of either bipo-
`lar I mania or bipolar I depression.
`
`Bipolar I mania
`
`The most common treatments for bipolar I mania
`with psychotic features (n=181) were a mood
`stabilizer/antipsychotic combination (29%), mood
`stabilizer alone (9%), mood stabilizer/antipsy-
`chotic/benzodiazepine combination (9%), antipsy-
`chotic alone (8%) or mood stabilizer/antipsy-
`chotic/antidepressant combination (8%).
`In contrast, bipolar I mania without psychotic
`features (n=796) was most commonly treated
`with an antidepressant alone (20%), a benzodi-
`azepine/antidepressant combination (10%) or a
`mood stabilizer/antipsychotic combination (9%).
`
`While only 2% of patients with bipolar I mania
`with psychotic features were prescribed no psy-
`chotropic agents at discharge, a higher percentage
`of patients with bipolar I mania without psychotic
`features received no psychotropic agents at dis-
`charge (11%).
`
`Bipolar I depression
`
`The most common treatments for bipolar I depres-
`sion with psychotic features (n=46) were an an-
`tipsychotic/antidepressant combination (20%), a
`mood stabilizer/antipsychotic/antidepressant com-
`bination (20%), an antidepressant alone (13%), or
`mood stabilizer/antipsychotic/antidepressant/ben-
`zodiazepine combination (11%). Bipolar I depres-
`sion without psychotic features (n=398) was most
`commonly treated with a mood stabilizer/antipsy-
`chotic/antidepressant combination (14%), a mood
`stabilizer/antidepressant/benzodiazepine combina-
`tion (11%), a mood stabilizer/antidepressant com-
`bination (10%) or an antidepressant alone (10%).
`The percentage of patients with bipolar I depres-
`sion with, or without psychotic features who were
`
`Table 3. Medication treatments for patients diagnosed with bipolar I mania
`
`Percentage
`
`Bipolar I mania with
`psychotic features
`(n=181)
`
`Bipolar I mania without
`psychotic features
`(n=796)
`
`55
`
`2
`20
`
`9
`2
`3
`
`17
`
`10
`
`5
`4
`3
`
`02
`
`10
`11
`2
`
`Mono-therapy
`Mood stabilizera alone
`Antipsychotic alone
`Benzodiazepine alone
`Antidepressant alone
`
`Dual-therapy
`Mood stabilizer and antipsychotic
`Mood stabilizer and benzodiazepine
`Mood stabilizer and antidepressant
`Antipsychotic and benzodiazepine
`Antipsychotic and antidepressant
`Benzodiazepine and antidepressant
`
`Poly-therapy
`Mood stabilizer, antipsychotic and benzodiazepine
`Mood stabilizer, antipsychotic and antidepressant
`Mood stabilizer, benzodiazepine and antidepressant
`Antipsychotic, benzodiazepine and antidepressant
`Mood stabilizer, antipsychotic, benzodiazepine and antidepressant
`
`Other psychotropic combinations (not including those listed above)
`No psychotropic therapy
`ECT with or without pharmacologic therapies
`
`9
`8
`1
`1
`
`29
`4
`1
`6
`4
`1
`
`9
`8
`2
`3
`7
`
`5
`2
`1
`
`Treatments recommended by Expert Consensus Guidelines (32) are highlighted in bold for each diagnostic subset.
`a Mood stabilizer includes lithium, valproate, carbamazepine, gabapentin.
`
`168
`
`4 of 9
`
`Alkermes, Ex. 1053
`
`

`
`Table 4. Medication treatments for patients diagnosed with bipolar I depression
`
`Prescribing patterns and adherence to bipolar practice guidelines
`
`Percentage
`
`Bipolar I depression with
`psychotic features
`(n=46)
`
`Bipolar I depression without
`psychotic features
`(n=398)
`
`Mono-therapy
`Mood stabilizera alone
`Antipsychotic alone
`Benzodiazepine alone
`Antidepressant alone
`
`Dual-therapy
`Mood stabilizer and antipsychotic
`Mood stabilizer and benzodiazepine
`Mood stabilizer and antidepressant
`Antipsychotic and benzodiazepine
`Antipsychotic and antidepressant
`Benzodiazepine and antidepressant
`
`Poly-therapy
`Mood stabilizer, antipsychotic and benzodiazepine
`Mood stabilizer, antipsychotic and antidepressant
`Mood stabilizer, benzodiazepine and antidepressant
`Antipsychotic, benzodiazepine and antidepressant
`Mood stabilizer, antipsychotic, benzodiazepine and antidepressant
`
`Other psychotropic combinations (not including those listed above)
`No psychotropic therapy
`ECT with or without pharmacologic therapies
`
`0
`7
`0
`13
`
`7
`0
`2
`4
`20
`4
`
`7
`20
`2
`2
`11
`
`0
`2
`2
`
`4
`2
`2
`10
`
`5
`2
`10
`1
`5
`6
`
`3
`14
`11
`3
`8
`
`10
`6
`3
`
`Treatments recommended by Expert Consensus Guidelines (32) are highlighted in bold for each diagnostic subset.
`a Mood stabilizer includes lithium, valproate, carbamazepine, gabapentin.
`
`prescribed no psychotropic agents at discharge was
`2% and 6%, respectively.
`
`Most frequently prescribed agents within medication
`category
`
`Table 5 lists the percentages of bipolar I mania or
`bipolar I depression patients prescribed specific
`medications within each of four medication cate-
`gories (mood stabilizers, antipsychotics, antide-
`pressants, benzodiazepines). For example, of those
`with mania who were prescribed a mood stabilizer
`(n=405), 59% were prescribed valproate and 38%
`were prescribed lithium. For those with either ma-
`nia or depression, valproate and lithium were the
`most frequently prescribed agents. When benzodi-
`azepines were prescribed,
`lorazepam and clon-
`azepam were used most commonly. For patients
`with bipolar I mania, trazodone and sertraline were
`used in approximately equal proportions. For
`those with bipolar I depression, sertraline was used
`most often of the antidepressants. Of the antipsy-
`chotic agents prescribed for bipolar I mania, rispe-
`ridone or conventional agents were used most
`frequently, while for bipolar I depression olanzap-
`ine or conventional agents were prescribed most
`often.
`
`Electroconvulsive therapy (ECT)
`
`Of note, a very small percentage of patients with
`bipolar I mania (3%) or bipolar I depression (5%)
`(with or without psychotic features) received elec-
`troconvulsive therapy (ECT), despite its well-docu-
`mented efficacy in the treatment of both phases of
`the disorder (37–39).
`
`Adherence to Expert Consensus Guidelines
`
`Of the patients with bipolar I mania and having
`psychotic features, 38% were prescribed a recom-
`mended pharmacological therapy. More specifi-
`cally, 29% were prescribed a mood stabilizer and
`an antipsychotic, while 9% were prescribed a com-
`bination of a mood stabilizer, an antipsychotic and
`a benzodiazepine (refer again to Table 3). Seven-
`teen percent were discharged on a single mood
`stabilizer or an antipsychotic agent. Of note, about
`a quarter of the patients (26%) in this diagnostic
`subgroup were discharged without
`a mood
`stabilizer.
`Even more strikingly, only 16% of the bipolar I
`mania diagnostic subgroup (without psychotic fea-
`tures) were prescribed a recommended pharmaco-
`logic therapy at discharge (see Table 3) consisting
`
`169
`
`5 of 9
`
`Alkermes, Ex. 1053
`
`

`
`Lim et al.
`
`of a mood stabilizer and an antipsychotic (9%),
`mood stabilizer alone (5%), or mood stabilizer and
`benzodiazepine (2%). The high use of antidepres-
`sant agents, alone or in combination, in about half
`of the patients with bipolar I mania (without psy-
`chotic features) was surprising and somewhat per-
`plexing. In addition, two-thirds of the patients
`(66%) were discharged without a mood stabilizer,
`with 11% being discharged on no psychotropic
`agent.
`Turning to bipolar I depression patients with
`psychotic features,
`fewer than one-third (31%)
`were discharged with a recommended pharmaco-
`logic therapy (see Table 4). Of the recommended
`strategies, clinicians preferred the combination of a
`mood stabilizer, an antipsychotic and an antide-
`pressant (20%), versus a) the combination of a
`mood stabilizer and an antipsychotic (7%), b) the
`combination of a mood stabilizer and an antide-
`pressant (2%) or c) ECT (with or without pharma-
`cological agents) (2%). Of note, about half of the
`patients (52%) were discharged on no mood stabi-
`lizer. Virtually all (98%) patients within this diag-
`
`Table 5. Most common psychotropic agents prescribed at discharge for
`bipolar I manic and bipolar I depressed patients by medication category
`
`Percentages
`
`Medicationsa
`
`Bipolar I mania
`(with or without
`psychosis)
`
`Bipolar I depression
`(with or without
`psychosis)
`
`(n=272)
`50
`40
`12
`10
`
`(n=205)
`37
`27
`8
`32
`
`(n=329)
`18
`9
`15
`14
`10
`9
`
`(n=179)
`37
`29
`12
`10
`10
`
`Mood stabilizers
`Valproate
`Lithium
`Gabapentin
`Carbamazepine
`
`(n=405)
`59
`38
`12
`6
`
`(n=442)
`Antipsychotics
`26
`Olanzapine
`37
`Risperidone
`6
`Quetiapine
`Conventional agents 36
`
`Antidepressants
`Sertraline
`Fluoxetine
`Bupropion HCl
`Venlafaxine HCl
`Mirtazapine
`Trazodone
`
`Benzodiazepines
`Lorazepam
`Clonazepam
`Alprazolam
`Bisoprolol
`Hydroxyzine HCl/
`Hydroxyzine
`Pamoate
`
`(n=517)
`21
`13
`9
`9
`7
`22
`
`(n=321)
`37
`31
`12
`10
`9
`
`a Medications are not mutually exclusive.
`
`170
`
`nostic subgroup were treated with a psychotropic
`agent.
`Of the patients with bipolar I depression, without
`psychotic features, only 17% were discharged with
`a recommended pharmacologic therapy (including
`ECT) (see Table 5). The combination of a mood
`stabilizer and an antidepressant was prescribed
`more frequently (10%) than was a mood stabilizer
`alone (4%) or ECT (3%). Although the majority
`(94%) was prescribed a psychotropic agent, 37%
`were discharged without a mood stabilizer.
`
`Discussion
`Bipolar I disorder is among the most challenging
`conditions
`that
`clinicians
`face, with many
`monotherapeutic medication approaches rarely ef-
`fective (40, 41). This research was undertaken to
`describe the primary medication treatment patterns
`for inpatients discharged with a diagnosis of bipo-
`lar I disorder (mania or depression), and to deter-
`mine the extent to which clinicians treating these
`patients are following published treatment guide-
`lines derived from empirical studies and expert
`consensus.
`Because the guidelines are intended to advise
`rather than mandate specific forms of treatment,
`we did not expect clinicians to demonstrate full
`compliance with all recommendations. Nonethe-
`less, we were surprised by the enormously wide
`array of pharmacologic treatments employed that
`often differed substantially from the most current
`set of published medication treatment guidelines
`for bipolar disorder. Only about one out of three
`bipolar I (manic or depressed) patients with psy-
`chotic features was discharged on medications rec-
`ommended as preferred or alternate treatments for
`that disorder. Even more troubling is the finding
`that only one in six patients with bipolar I (mania
`or depression) without psychotic features was dis-
`charged on recommended pharmacologic agents.
`These findings suggest that a substantial number
`of patients with bipolar I disorder were discharged
`from the hospital on treatment regimens that 1)
`did not include medications with demonstrated
`effectiveness for their illness and 2) are not gener-
`ally recommended as preferred or alternate treat-
`ments by current clinical practice and expert
`consensus guidelines. As the findings are taken
`from a national cohort of acute care medical–sur-
`gical facilities with inpatient psychiatric programs
`that treat bipolar patients, one cannot attribute
`these findings to an idiosyncratic pattern of pre-
`scribing at a single location or geographic region.
`Perhaps the most surprising finding was the rela-
`tive high use of antidepressant agents alone or with
`other psychotropic agents. Thirty percent of bipo-
`
`6 of 9
`
`Alkermes, Ex. 1053
`
`

`
`Prescribing patterns and adherence to bipolar practice guidelines
`
`lar I mania patients with psychotic features and
`58% of bipolar I mania patients without psy-
`chotic features were prescribed an antidepressant
`at discharge. It has been reported that antidepres-
`sants can induce hypomania, full-blown mania or
`rapid cycling in bipolar patients (42). For that
`reason, none of the published guidelines recom-
`mend use of antidepressants for the manic phase
`of the illness. One set of guidelines (29) even
`states explicitly that if an individual is receiving
`an antidepressant while in a manic, hypomanic or
`mixed state, the antidepressant should be discon-
`tinued. The only exception may be the use of
`trazadone to treat mania-related insomnia. Traza-
`done was,
`indeed, one of the most commonly
`prescribed antidepressants. Other frequently pre-
`scribed antidepressants were the SSRIs
`(most
`commonly sertraline or fluoxetine), as well as
`buproprion and venlafaxine. We had expected
`somewhat higher use of buproprion, since it is
`often prescribed as a first-line treatment for bipo-
`lar depression and according to some authors,
`has a relatively low likelihood of causing a switch
`to mania (32).
`A partial explanation for the high use of an-
`tidepressants among bipolar manic patients may
`be found from a recent study (43), which con-
`cluded that 40% of a sample with a hospital dis-
`charge diagnosis of bipolar disorder had been
`diagnosed with unipolar depression prior to ad-
`mission. What is unknown is why antidepressants
`were continued once the bipolar manic diagnosis
`was made. An alternative explanation is that the
`discharge diagnosis was miscoded in the medical
`record (e.g. bipolar mania was coded when bipo-
`lar depression was the more accurate diagnosis).
`It should also be noted that diagnoses were most
`commonly made by the
`treating psychiatrist
`based on clinical observation, a less
`reliable
`method for assigning a diagnosis than is use of a
`more structured diagnostic interview.
`Not surprisingly for a hospitalized cohort, an-
`tipsychotic agents were commonly used both for
`bipolar I patients with psychotic features (74%
`manic; 78% depressed) as well as for those with-
`out psychotic features
`(33% manic; 41% de-
`pressed). Of
`those receiving an antipsychotic
`agent, about
`two-thirds were receiving second-
`generation agents, most commonly olanzapine or
`risperidone, with the other third prescribed a con-
`ventional
`antipsychotic
`agent.
`It
`should be
`pointed out that one limitation of the 2000 Ex-
`pert Consenus Guidelines is that they confine the
`role of antipsychotic agents to adjunctive treat-
`ment of bipolar disorder. This fails to take into
`account the relatively recent approval of olanzap-
`
`ine as a first-line monotherapeutic treatment for
`acute mania. This
`development
`should
`be
`reflected in future revisions to published treat-
`ment guidelines.
`We also found that a third or more of patients
`with bipolar I depression were prescribed a benzo-
`diazepine. Although benzodiazepines such as lo-
`razepam and clonazepam can play an important
`role in helping to attenuate the nonspecific symp-
`toms associated with an acute manic episode (e.g.
`insomnia, agitation, hyperactivity and anxiety)
`(40), the risk of abuse of anxiolytic agents among
`individuals prone to high rates of concurrent sub-
`stance use should be monitored closely (44).
`The variety of pharmacologic treatment thera-
`pies prescribed to inpatients with bipolar I disor-
`der suggests that practitioners in the field do not
`yet share the degree of consensus implied by pub-
`lished treatment guidelines. We know from other
`areas of medicine that the mere existence of a
`guideline does not mean that it will be easily or
`readily transferred into actual clinical practice
`(45). There are many barriers that may contribute
`to the relative lack of adherence. Gilbert et al.
`(23) offer such factors as physician reluctance to
`be ‘told’ what to do, the lack of necessary train-
`ing and ongoing consultation required, the com-
`plexity and sophistication of some guidelines, the
`perception of additional work required to use
`guidelines, the perceived ‘static’ nature of guide-
`lines and the failure of many patients to adhere
`to guideline-directed treatment. Moreover, a re-
`cent study suggests that about one in six psychia-
`trists was unaware that APA practice guidelines
`for specific psychiatric conditions even existed,
`and almost half had not read any of the guideli-
`nes specific to bipolar disorder (46).
`The degree to which the findings regarding
`medication prescribing patterns
`for
`inpatients
`with bipolar I disorder generalize to less restric-
`tive environments, such as the outpatient setting,
`is unknown. Future research is also needed to
`understand the extent to which adherence to pub-
`lished practice guidelines in psychiatry actually
`improves patient outcomes (47) in the treatment
`of this complex, challenging and often devastat-
`ing disorder.
`
`Acknowledgements
`The assistance of Yoav Gershon, MD, and John S
`Kennedy, MD, in providing valuable comments is grate-
`fully acknowledged. Portions of this work were presented
`at the American Psychiatric Association in Washington,
`DC (May 1999), the Third International Conference on
`Bipolar Disorder in Pittsburgh, PA (June 1999) and the
`Association for Health Services Research in Chicago, IL
`(June 1999).
`
`171
`
`7 of 9
`
`Alkermes, Ex. 1053
`
`

`
`Lim et al.
`
`References
`
`1. National Institutes of Mental Health. Mental Health: A
`Report of the Surgeon General – Executive Summary.
`Rockville, MD: US Dept of Health and Human Services,
`1999; 47–48.
`2. American Psychiatric Association. DSM-IV: Diagnostic
`and Statistical Manual of Mental Disorders. Washington,
`DC, 1994.
`3. Kessler RC, McGonagle KA, Zhao S et al. Lifetime and
`12-month prevalence of DSM-III-R psychiatric disorders
`in the United States. Arch Gen Psychiatry 1994; 51: 8–19.
`4. Regier DA, Narrow WE, Rae DS et al. The de facto US
`mental and addictive disorders service system. Epidemio-
`logic catchment area prospective 1-year prevalence rates of
`disorders and services. Arch Gen Psychiatry 1993; 50 (2):
`85–94.
`5. Weissman MM, Bland RC, Canino GJ et al. Cross-na-
`tional epidemiology of major depression and bipolar disor-
`der. JAMA 1996; 276 (4): 293–299.
`6. Woods SW. The economic burden of bipolar disease. J
`Clin Psychiatry 2000; 61 (Suppl 13): 38–41.
`7. Gelenberg AJ, Hopkins HS. Report on efficacy of treat-
`ments for bipolar disorder. Psychopharmacol Bull 1993; 29
`(4): 447–456.
`8. American Psychiatric Association. Practice guideline for
`the treatment of patients with bipolar disorder. Am J
`Psychiatry 1994; 141 (12 Suppl): 1–36.
`9. Goodwin F, Jamison K. Manic-Depressive Illness. New
`York: Oxford University Press, 1990.
`10. Coryell W, Scheftner W, Keller M et al. The enduring
`psychosocial consequences of mania and depression. Am J
`Psychiatry 1993; 150 (5): 720–727.
`11. Murray CJL, Lopez AD. In: The Global Burden of Dis-
`ease. World Health Organization. Cambridge, MA: Har-
`vard University Press, 1996.
`12. Keck, Jr. PE, McElroy SL, Stanton SP, Bennett JA. Phar-
`macoeconomic aspects of the treatment of bipolar disor-
`der. Psychiatric Ann 1996; 27 (Suppl 7): S449–S453.
`13. Wyatt RJ, Henter I. An economic evaluation of manic-de-
`pressive illness – 1991. Soc Psychiatry Psychiatric Epi-
`demiol 1995; 30 (5): 213–219.
`14. Simon GE, Unutzer J. Health care utilization and costs
`among patients treated for bipolar disorder in an insured
`population. Psychiatric Services 1999; 50 (10): 1303–1308.
`15. Brady KT. Difficulties in diagnosis and management of
`bipolar disorder: three case presentations. J Clin Psychi-
`atry 2000; 61 (Suppl 13): 32–37.
`16. Greenhouse WJ, Meyer B, Johnson SL. Coping and medi-
`cation adherence in bipolar disorder. J Affect Disorders
`2000; 59 (3): 237–241.
`17. Jamison KR, Gerner RH, Goodwin FK. Patient and
`physician attitudes toward lithium: relationship to compli-
`ance. Arch Gen Psychiatry 1979; 36 (8 Spec No): 866–869.
`18. Weiss RD, Greenfield SF, Najavits LM et al. Medication
`compliance among patients with bipolar disorder and sub-
`stance use disorder. J Clin Psychiatry 1998; 59 (4): 172–
`174.
`19. Hopkins HS, Gelenberg AJ. Treatment of bipolar disor-
`der: how far have we come? Psychopharmacol Bull 1994;
`30 (1): 27–38.
`20. Prien RF, Gelenberg AJ. Alternatives to lithium for pre-
`ventive treatment of bipolar disorder. Am J Psychiatry
`1989; 146 (7): 840–848.
`21. Gelenberg AJ, Kane JM, Keller MB et al. Comparison of
`standard and low serum levels of lithium for maintenance
`
`172
`
`treatment of bipolar disorder. New Engl J Med 1989; 321
`(22): 1489–1493.
`22. Goldberg JF, Harrow M, Grossman LS. Course and out-
`come in bipolar affective disorder: a longitudinal