Available online at www.sciencedirect.com

ScienceDirect

Comprehensive Psychiatry 55 (2014) 1855–1861www.elsevier.com/locate/comppsych

Correlates of psychiatric hospitalization in a clinical sample of Canadianadolescents with bipolar disorder

Joshua Shapiroa,b, Vanessa Timminsa, Brenda Swampillai a, Antonette Scavonea,Katelyn Collingera, Carolyn Boulosa, Jessica Hatcha, Benjamin I. Goldsteina,⁎

aDepartment of Psychiatry, Sunnybrook Health Sciences Centre, Toronto, ON, CanadabWilfrid Laurier University, Waterloo, ON, Canada

Abstract

Objective: To identify factors associated with psychiatric hospitalization among adolescents with bipolar disorder (BD).Methods: Participants were 100 adolescents, ages 13–19, who fulfilled DSM-IV criteria for bipolar I disorder [(BD-I), n = 26], bipolar IIdisorder [(BD-II), n = 40], or operationalized criteria for BD not otherwise specified [(BD-NOS), n = 34], via the Schedule for AffectiveDisorders and Schizophrenia, Present and Lifetime version (KSADS-PL). Demographic, clinical, and family history variables were measuredvia clinical interview with the participant and a parent or guardian.Results: The lifetime prevalence of psychiatric hospitalization was 50%. Significant predictors of psychiatric hospitalization in univariateanalyses included older age, BD-I, history of suicide attempt, psychosis, lifetime use of second generation antipsychotics (SGAs), lithium,SSRI antidepressants and any medication. BD-II was negatively associated with psychiatric hospitalization. In multivariable analyses, olderage, history of suicide attempt, psychosis and use of SGAs were positively associated with hospitalization, whereas BD-II was negativelyassociated with hospitalization.Conclusions: Psychiatric hospitalization in adolescents with BD is highly prevalent and associated with older age and proxies for greater illnessseverity. Further studies are needed to identify strategies for reducing the need for psychiatric hospitalizations among adolescents with BD.© 2014 Elsevier Inc. All rights reserved.

1. Introduction

Bipolar disorder (BD) is a recurrent and impairing illnessthat is associated with substantial burden of mood symptomsand comorbidity among adults and youth [1–3]. BD oftenrequires high treatment usage [4], andmore than 75% of adultswith BD report one or more lifetime psychiatric hospitaliza-tions [5]. With widespread decreases in the number ofpsychiatric beds, psychiatric hospitalization in recent years isoften precipitated by acute crises that necessitate hospitaliza-tion for the purpose of risk containment. As such, psychiatrichospitalization is often a decision driven by immediate needrather than individual preferences among patients, families,

⁎ Corresponding author at: Centre for Youth Bipolar Disorder,Sunnybrook Health Sciences Centre; 2075 Bayview Ave., Room FG-53.Tel.: +1 416 480 5328.

E-mail address: benjamin.goldstein@sunnybrook.ca (B.I. Goldstein).

http://dx.doi.org/10.1016/j.comppsych.2014.08.0480010-440X/© 2014 Elsevier Inc. All rights reserved.

and/or practitioners. Concerns about hospitalization relate tothe disruptive and distressing experience psychiatric hospital-ization can be for patients with BD and their families, and thesubstantial costs incurred [6,7]. The overall treatment costs ofpatients with BD have been found to exceed those for otherbehavioral and mood disorders [8–10], a disparity that isattributable in large part to hospitalizations which account for20% of these costs [11].

Investigation of clinical characteristics associated withhospitalization can potentially help align outpatient treatmentfocus toward preventing these outcomes and identify patientsat risk for hospitalization. Hospitalizations among adults withBD have been associated with substance use disorders (SUD)[12–15], rapid cycling [16,17], psychosis [18], previoushospitalization [19,20], BD-I subtype [21], female gender[22,23], various medications [11,16–18,24–28], minoritystatus [18], polypharmacy [27], and marital separation [13].

While there is a larger literature on this topic among adultswith BD, there is a growing body of literature regarding therates and correlates of psychiatric hospitalization among

1856 J. Shapiro et al. / Comprehensive Psychiatry 55 (2014) 1855–1861

adolescents with BD. Hospitalization rates for psychiatricdisorders in youth, especially BD, have increased markedly inthe past 20 years [29–33]. The annual rate of hospitalizationamong adolescents with BD in the US was found to be 39.6%in 2003 [8]. Annual rates in children and adolescents increased434% between 1997 and 2010 [29], which coincided with alarge increase in number of BD diagnoses for this age group[30,31]. Adolescents with BD are hospitalizedmore often thanthose with most other psychiatric and behavioral disorders[34–37], and rates of inpatient service use now exceed eventhose of adults with BD in some reports [22]. Female gender[38,39], comorbid health conditions [39] and having a parentwith SUD [38] have all been shown to be directly associatedwith hospitalization in adolescents with BD, while suicidal orself-injurious behavior [35], comorbid physical conditions ingeneral [39] and rapid cycling [38] are associated with use ofhighly restrictive treatment settings.

The purpose of this report is to expand the literature onmental health service use among adolescents with BD byinvestigating characteristics associated with psychiatrichospitalization in a treatment-seeking sample from a sub-specialty clinic at a tertiary academic health sciences centerin Canada. Most previous studies on the topic of psychiatrichospitalizations among youth with BD are derived fromAmerican populations, and differences in the two nations'health care systems [40,41] suggest the need to examine thistopic among Canadian adolescents specifically. In particular,the absence of universal healthcare to date may renderbarriers that are not present in socialized healthcare [42],which may cause for differences in factors associated withpsychiatric hospitalization.

2. Methods

2.1. Sample and setting

The present study includes 100 participants, ages 13–19,with BD-I (n = 26), BD-II (n = 40), or BD-not otherwisespecified (NOS; n = 34). Participants were seeking outpa-tient assessment and/or treatment at a sub-specialty clinic in atertiary academic health sciences center. Written informedconsent was provided by parent(s) and adolescent beforestudy commencement. Research ethics board approval wasobtained. At least one parent/guardian for each adolescentalso participated.

2.2. Subject assessment

Diagnoses (current and lifetime) were determined via theSchedule for Affective Disorders and Schizophrenia, Presentand Lifetime version (KSADS-PL) [43], which incorporatesinformation from adolescents and their parent(s). Similarly,lifetime history of psychiatric hospitalization was ascer-tained during the KSADS-PL interview. Primary cause ofhospitalization was determined through review of availablepatient charts. All interviewers had a bachelor's or master's

degree in a mental health field and completed comprehensiveKSADS training under the supervision of the senior author(B.G.). The KSADS Depression Rating Scale (DEP-P) [44]and KSADS Mania Rating Scale (MRS) [45] were used inplace of the mood sections of the KSADS-PL. Diagnoseswere determined with the consideration of all availableinformation and clinical judgment was used when conflictinginformation was provided. Diagnoses were confirmed by aconsensus meeting with a child psychiatrist followingcompletion of the KSADS-PL interview (B.G.).

This study employed operationalized criteria for BD-NOSas described in the Course and Outcome of Bipolar Youth(COBY) study [46]: elevated and/or irritable mood, plus(i) two DSM-IV manic symptoms (three if only irritablemood is reported), (ii) change in functioning, (iii) mood andsymptom duration of at least 4 h during a 24-h period, and(iv) at least four cumulative 24-h periods of episodes over theparticipant's lifetime that meet the mood, symptom severity,and functional change criteria.

The age of onset of an individual's BD was consideredto be when the participant first met DSM-IV criteria for amanic, hypomanic, ormajor depressive episode, orwhen he/shefirst met study criteria for BD-NOS.

A self-reported medical history questionnaire and thepost-traumatic stress disorder (PTSD) screen within theKSADS-PL were used to obtain history of physical andsexual abuse. A Safety Assessment Form, an interviewer-administered questionnaire, was used to determine lifetimeaggression and suicidality that may have occurred outside ofthe context of a depressive episode. The Conflict BehaviorQuestionnaire (CBQ) self-report assesses family conflict andwas completed by parents regarding their adolescent [47].Psychiatric status and history of first- and second-degreerelatives was obtained through an interview with theadolescent and parent(s), using the Family History Screen[48]. Socioeconomic status was determined via the 4-factorHollingshead Scale [49].

2.3. Data analysis

Variables were screened for their association with lifetimepsychiatric hospitalization using chi-square tests for cate-gorical variables and t-tests for continuous variables. For thepurpose of this analysis, the dimensional CBQ, LPI andCALS questionnaires were dichotomized as high scoreversus low score on the basis of a median split. Demographicand clinical characteristics and comorbidities that wereassociated with psychiatric hospitalization (p b 0.1) in theunivariate analyses were entered into a backward eliminationWald logistic regression model in order to examine the uniquecontribution to variance in psychiatrics hospitalizations associ-ated with each predictor, controlling for the effects of otherpredictors. All p values are based on two-tailed tests with asignificance level ofα = 0.05. Statistical correction for multiplecomparisons was not applied as information on the correlatesof psychiatric hospitalization in this population is scarce,

1857J. Shapiro et al. / Comprehensive Psychiatry 55 (2014) 1855–1861

especially since youth with BD-II and BD-NOS wereincluded. The Statistical Package for the Social SciencesVersion 20 (SPSS) was used to perform statistical analyses.

3. Results

3.1. Socio-demographic characteristics of adolescents withbipolar disorder who had a psychiatric hospitalization

Socio-demographic characteristics of all participants areportrayed in Table 1. Of the 100 adolescents with BD, 50%had at least one lifetime psychiatric hospitalization. Partici-pants with, versus without, a history of psychiatric hospital-ization were significantly older. Gender, race, socioeconomicstatus and living with both natural parents were notsignificantly associated with psychiatric hospitalization.

3.2. Clinical and familial characteristics of adolescents withbipolar disorder who had a psychiatric hospitalization

A comparison of clinical characteristics distinguishingthose who have a lifetime psychiatric hospitalization to thosewho have not is portrayed in Table 2. BD subtype wassignificantly associated with psychiatric hospitalization. Apost-hoc univariate analysis of the three BD subtypesrevealed that patients with BD-I were significantly morelikely to have a psychiatric hospitalization, while patientswith BD-II were significantly less likely. Participants with,versus without, lifetime psychiatric hospitalization weresignificantly more likely to have lifetime psychosis, lifetimesuicide attempt, and trended toward lower prevalence ofcomorbid attention deficit hyperactivity disorder (ADHD)(p = 0.07). Participants with lifetime psychiatric hospitali-zation were also significantly more likely to have lifetime useof lithium, SSRI antidepressants, second generation antipsy-chotics (SGAs) and any medication. First- and second- degreefamily history of mania or hypomania, major depressiveepisode, ADHD, conduct disorder, anxiety and substancedependence is displayed in Table 3. No significant between-

Table 1Demographic characteristics of 100 adolescents with bipolar disorders (BD) with

Overall sample(n = 100)

Psychiatric ho(n = 50)

Age, mean ± SD 16.2 ± 1.5 16.6 ± 1.3Socioeconomic status, mean ± SD (n = 96) 48.5 ± 12.9 47.9 ± 13.4

n (%) n (%)

Sex (males) 33 (33) 13 (26)Race (white) 86 (86) 40 (80)Intact familyb 56 (56) 26 (52)

If a variable has n b 100 due to missing information, the actual n value is noted b⁎⁎⁎ p b 0.001.

a Two-tailed Pearson χ2 analyses were performed for categorical variables ab Subject lives with both biological parents.

group differences in prevalence of these familial conditionswere observed.

3.3. Logistic regression analysis

Demographic and clinical characteristics with p b 0.1 inthe univariate analysis were included in a logistic regressionto investigate their unique contributions to variance inpsychiatric hospitalization. Variables examined included:age, BD-I subtype, BD-II subtype, suicide attempt,psychosis, ADHD comorbidity, lifetime use of anymedication, lifetime use of SGAs, lithium, and SSRIantidepressants. Psychiatric hospitalization served as thedependent variable for this analysis.

History of suicide attempt [odds ratio (OR) = 10.4, 95%confidence interval (CI): 2.7–40.7, p = 0.001] and use ofSGAs (OR = 7.0, 95% CI: 2.3–21.8, p = 0.001) weresignificantly associated with psychiatric hospitalization.The associations of psychiatric hospitalization with olderage (OR = 1.5, 95% CI: 1.0–2.4, p = 0.05) and psychosis(OR = 3.4, 95% CI: 0.8–14.3, p = 0.095) were reduced tostatistical trends. The negative association with BD-II(OR = 0.3, 95% CI: 0.10–1.0, p = 0.05) was also reducedto a statistical trend.

Information on primary cause was available for 79 totalhospitalizations, representing 45 adolescents with BD.Information was completely or partially missing for 6participants, accounting for approximately 18 hospitaliza-tions. As information regarding reason for hospitalizationrelied on chart review and was not ascertained systematical-ly, these data are presented for descriptive purposes and werenot subjected to further statistical analysis. According tochart review, mania or psychotic mania was the mostcommon primary cause of hospitalization (35% of cases),followed by suicidal ideation (19%), suicide attempt (18%)and depression (17%). A total of 11% of the hospitalizationswere attributable to various other reasons including self-injurious behavior, aggression, medication misuse or optimi-zation, anxiety, food restriction, and catatonia.

versus without psychiatric hospitalization.

spitalization No psychiatric hospitalization(n = 50)

Statistica p

15.8 ± 1.5 −3.1 0.003⁎⁎⁎

49.1 ± 12.5 0.48 0.63

n (%)

20 (40) 2.2 0.1446 (92) 3.0 0.0830 (60) 0.65 0.42

eside the variable. Otherwise, it is assumed that all variables have n = 100.

nd two-tailed t-tests conducted for continuous variables.

Table 2Clinical characteristics, comorbidities, medication history, and life events of 100 adolescents with bipolar disorders (BD) with versus without psychiatric hospitalization.

Overall sample(n = 100), n (%)

Psychiatric hospitalization(n = 100), n (%)

No psychiatric hospitalization(n = 100), n (%)

Statistica p

Clinical characteristicsBD subtype 21.3 b0.001⁎⁎⁎

BD-I 26 (26) 23 (46) 3 (6) 21.9 b0.001⁎⁎⁎

BD-II 40 (40) 13 (26) 27 (54) 9.7 0.008⁎⁎

BD-NOS 34 (34) 14 (28) 20 (40) 2.8 0.25Psychosis 24 (24) 18 (36) 6 (12) 7.9 0.005⁎⁎

Self-injurious behavior 47 (47) 23 (46) 24 (48) 0.04 0.84Suicidal ideation 61 (61) 33 (66) 28 (56) 1.1 0.31Suicide attempt 25 (25) 18 (36) 7 (14) 6.5 0.01⁎

Physical or sexual abuse 17 (17) 10 (20) 7 (14) 0.64 0.42Comorbidity

ADHD 41 (41) 16 (32) 25 (50) 3.3 0.07SUD 33 (33) 20 (40) 13 (26) 2.2 0.14ODD 36 (36) 15 (30) 21 (42) 1.6 0.21Conduct disorder 10 (10) 5 (10) 5 (10) b0.01 N0.99Anxiety disorder 72 (72) 37 (74) 35 (70) 0.2 0.66Panic disorder 17 (17) 7 (14) 10 (20) 0.6 0.42

Medication historyLifetime medication 71 (71) 43 (86) 28 (56) 10.9 0.001⁎⁎

Lifetime antimanic/anticonvulsantsb 13 (13) 8 (16) 5 (10) 0.8 0.37Lifetime 2nd generation antipsychoticsc 53 (53) 38 (76) 15 (30) 21.2 b0.001⁎⁎⁎

Lifetime lithium 20 (20) 17 (34) 3 (6) 12.3 b0.001⁎⁎⁎

Lifetimes SSRI antidepressantsd 32 (32) 21 (42) 11 (22) 4.6 0.03⁎

Lifetime non-SSRI antidepressantse 14 (14) 6 (12) 8 (16) 0.3 0.56Lifetime stimulantsf 16 (16) 8 (16) 8 (16) b0.01 N0.99

Life events and functioningPolice contact or arrest 43 (43) 24 (48) 19 (38) 1.0 0.31CBQ (parent; high) (n = 94) 50 (53) 22 (48) 28 (58) 1.0 0.31CALS (parent; high) (n = 90) 47 (52) 50 55 0.2 0.60CALS (adolescent; high) (n = 90) 46 (51) 21 (48) 25 (54) 0.4 0.53LPI identity confusion (% high) (n = 92) 50 (54) 19 (43) 31 (65) 4.2 0.04⁎

LPI interpersonal chaos (% high) (n = 92) 48 (52) 20 (46) 28 (58) 1.5 0.22LPI impulsivity (% high) (n = 92) 50 (54) 22 (50) 28 (58) 0.6 0.42LPI emotional dysregulation (% high) (n = 92) 50 (54) 20 (46) 30 (63) 2.7 0.1

BD = bipolar disorder; BD-I = bipolar I disorder; BD-II = bipolar II disorder; BD-NOS = bipolar disorder not otherwise specified; ADHD = attention-deficithyperactivity disorder; SUD = substance use disorder; ODD = oppositional defiant disorder; SSRI = selective serotonin reuptake inhibitor; CBQ = ConflictBehavior Questionnaire; CALS = Children Affective Liability Scale; LPI = Life Problems Inventory.If a variable has n b 100 due to missing information, the actual n value is noted beside the variable. Otherwise, it is assumed that all variables have n = 100.

⁎ p b 0.05.⁎⁎ p b 0.01.⁎⁎⁎ p b 0.001.

a Two-tailed Pearson χ2 analyses were performed for categorical variables.b Antimanic anticonvulsants = valproic acid, divalproex, carbamazepine.c Second generation antipsychotics = risperidone, olanzipine, aripiprazole, ziprasidone, seroquel.d SSRI antidepressants = zoloft, paroxetine, prozac, fluvoxamine, citalopram, lexapro.e Non-SSRI antidepressants = wellbutrin, remeron, effexor, cymbalta.f Stimulants = ritalin, concerta, adderall, dexedrine.

1858 J. Shapiro et al. / Comprehensive Psychiatry 55 (2014) 1855–1861

4. Discussion

This study found that 50%of the Canadian adolescents withBD in our sample had a history of psychiatric hospitalization.The current study is the only one to our knowledge to examinea wide range of correlates of psychiatric hospitalization inCanadian adolescents with BD. We found older age, BD-I,history of attempted suicide, psychosis, lifetime use of SGAs,lithium, SSRI antidepressants and any medication to bepositively associated with psychiatric hospitalization, whileBD-II was negatively associated with hospitalization. History

of suicide attempt and of exposure to SGAs were the variablesmost robustly associated with psychiatric hospitalization inmultivariable analyses.

In descriptive information derived from chart review,nearly 90% of hospitalizations were precipitated by mania,suicidality, or depression without suicidality. This mirrorstrends in the literature on adults, such as one study whichfound mania accounting for 50% of hospitalizations,depression for 25% and mixed episodes for 10% [20], andanother that found suicidality accounting for 48%, mania for27% and depression for 16% [50]. In adolescents, mania

Table 3First- and second-degree family history among 100 adolescents with bipolar disorders (BD) with versus without psychiatric hospitalization.

Overall sample (n = 100),n (%)

Psychiatric hospitalization(n = 50), n (%)

No psychiatric hospitalization(n = 50), n (%)

Statistica p

Mania/hypomania 53 (53) 30 (60) 23 (46) 2.0 0.16MDE 78 (78) 39 (78) 39 (78) b0.01 N0.99ADHD 28 (28) 13 (26) 15 (30) 0.2 0.66Anxiety 54 (54) 29 (58) 25 (50) 0.6 0.42SUD 45 (45) 23 (46) 22 (44) 0.04 0.84

MDE = major depressive episode; ADHD = attention-deficit hyperactivity disorder; CD = conduct disorder; SUD = substance use disorder.a Two-tailed Pearson χ2 analyses were performed for categorical variables.

1859J. Shapiro et al. / Comprehensive Psychiatry 55 (2014) 1855–1861

(irritability in particular) has been cited as the most commonreason for psychiatric hospitalization [51].

In the current study, 50% of adolescents with BD had alifetime psychiatric hospitalization, which is somewhatlower than a recent study that found a 63% prevalence [52].The higher rate of psychiatric hospitalization found in theCOBY study may be due to a greater percentage of subjectswith BD-I, and/or may relate to aforementioned differencein the American and Canadian healthcare systems. Thehospitalization rates in this sample of adolescents werehigher than in studies of mixed samples of children andadolescents that have investigated patients with BD-I(41.2%) [53], and patients of all BD subtypes (41.6%) [39],and considerably higher than rates found for children,ages 12 and under, for whom reported rates are closer to20% [35,51]. Indeed, in the current study of exclusivelyadolescents we found that older age was associatedwith increased prevalence of hospitalization. More than75% of adults with BD report at least one psychiatrichospitalization [5], though prevalence in older patients is atleast partially attributable to accumulated opportunity forpsychiatric hospitalization.

Participants with BD-I were significantly more likely tohave a psychiatric hospitalization than those with other BDsubtypes. This result is expected, as a criterion for BD-Idiagnosis is a manic episode, which is often differentiatedfrom hypomania by the necessity for hospitalization. Indeed,mania was the most common primary cause of psychiatrichospitalization. As follows, participants with BD-II weresignificantly less likely to have a psychiatric hospitalizationthan those with other BD subtypes. BD-II has beenpreviously associated with lower hospitalization ratescompared to BD-I in adults [21,54], and in a study of acombined sample of children and adolescents [55]. TheCOBY study in particular found prevalence of psychiatrichospitalization among the combined sample of youth withBD-I, BD-II, and BD-NOS to be 66.1%, 53.3%, and 28.8%,respectively. However, given that more than 50% ofhospitalizations were precipitated by depression or suicid-ality, it appears that there may be other factors at play that areyet unknown.

This study found that adolescents with, versus without, ahistory of suicide attempt(s) were significantly more likely tohave a lifetime psychiatric hospitalization. Suicide attempt

was the primary cause of 18% of hospitalizations. This is inline with other research on adolescent BD that finds suicideattempts related to more use of restrictive treatment settings[35]. Psychosis was significantly associated with psychiatrichospitalization. Psychosis has been previously shown topredict psychiatric hospitalization in adults [18], and patientswith psychotic BD have been found to experience a firsthospitalization at a younger age than individuals with BD andno psychosis [56]. Psychosis can accompany mania, whichwas responsible for half of the psychiatric hospitalizations inthis sample.

Adolescents with a history of psychotropic medication usewere significantly more likely to have a history of psychiatrichospitalization. Due to the retrospective methodologyemployed, the present study cannot determine whethermedications were initiated prior to, during, or followinghospitalization. We found risk for hospitalization to bestrongly associated with use of SGAs, which was significantin both the univariate and multivariate analyses. Adultswith BD treated with SGAs have been previously shown tobe at higher risk for hospitalization [28] compared tothose not on SGAs. Lithium use was also significantlyassociated with lifetime psychiatric hospitalization. We founda significant association between SSRI antidepressant use andpsychiatric hospitalization, which has been previously shownin adults [27]. As the current study did not characterize theseverity, duration, or number of previous depressive episodes,this finding may be confounded by indication. Indeed,depressive episodes increase risk for hospitalization amongadults [12,25].

Previous studies on psychiatric hospitalization in adoles-cent BD have found significant association with femalegender [38,39], although absence of gender difference inhospitalization rates has also been reported [57]. The currentstudy did not find a significant association for gender inregards to psychiatric hospitalization. Males were numeri-cally less likely to have been hospitalized, although thisassociation was not statistically significant. Given that oursample of 100 individuals was comprised of only 33%males, it is possible that our results would have beensignificant given a larger sample.

Finally, the current study found a statistical trend (p =0.07) toward less psychiatric hospitalizations for BDadolescents with, versus without, comorbid ADHD. A recent

1860 J. Shapiro et al. / Comprehensive Psychiatry 55 (2014) 1855–1861

study found that comorbid ADHD is not associated withincreased hospitalizations in children with BD [58]. Theneed exists for further research on ADHD-BD comorbidityin regards to psychiatric hospitalization.

The current study has several limitations. The sample wasderived from a sub-specialty clinic in a tertiary hospital,based in a diverse urban environment, and may not berepresentative of populations using different services orliving in different regions. In addition, psychiatric hospital-ization was determined based on interviewing adolescentsand their parents, and confirmatory hospital records were notsystematically reviewed. Information regarding cause ofhospitalization was not obtained systematically, whichlimited the usefulness of the data in statistical analysis.Finally, as this studied employed a cross-sectional, retro-spective approach, it could not be determined for mostcorrelates of psychiatric hospitalization whether they servedas antecedents or consequences of hospitalization. Thecurrent study examined a broad spectrum of variables.Although the multivariable analyses offer some protectionagainst spurious findings, the possibility of type I errorremains. Given the importance of the topic and the paucity ofavailable data on this topic, we opted for this approach over amore highly conservative approach.

5. Conclusions

Many findings in the current study are congruent with thecurrent literature in the US regarding psychiatric hospital-ization in both adolescent and adult BD. The narrowed focuson psychiatric hospitalization is infrequent in the literature,especially in regards to adolescents, and identified severalnovel correlates that warrant further investigation. Psychiat-ric hospitalization is often a crisis-oriented intervention that,although often necessary, is also often disruptive to patientsand families. Moreover, young patients may be particularlyaffected by psychiatric hospitalizations due to the impact onthe continuity of schooling and of social and recreationalactivities. Psychiatric hospitalizations are also exceedinglycostly to the healthcare system. Taken together, this suggeststhat identifying strategies to prevent hospitalization maybenefit all stakeholders, including patients, families, and thehealthcare system. This study identifies a number ofcharacteristics associated with psychiatric hospitalizationamong Canadian adolescents with BD that, if replicated byothers and in prospective samples, may help to inform thedevelopment of strategies and practices to reduce the needfor psychiatric hospitalization.

References

[1] Goldstein BI. Recent progress in understanding pediatric bipolardisorder. Arch Pediatr Adolesc Med 2012;166(4):362-71.

[2] Birmaher B, Axelson D, Goldstein B, StroberM, Gill MK, Hunt J, et al.Four-year longitudinal course of children and adolescents with bipolar

spectrum disorder: the course and outcome of bipolar youth (COBY)study. Am J Psychiatry 2009;166(7):795-804.

[3] Hilty DM, Brady KT, Hales RE. A review of bipolar disorder amongadults. Psychiatr Serv 1999;50(2):201-13.

[4] Depp CA, Jeste DV. Bipolar disorder in older adults: a critical review.Bipolar Disord 2004;6(5):343-67.

[5] Lish JD, Dime-Meenan S, Whybrow PC, Price RA, Hirschfeld R. TheNational Depressive and Manic-depressive Association (DMDA)survey of bipolar members. J Affect Disord 1994;31(4):281-94.

[6] Heru AM, Ryan CE. Burden, reward and family functioning ofcaregivers for relatives with mood disorders: 1-year follow-up. J AffectDisord 2004;83(2):221-5.

[7] Geller JL. The last half-century of psychiatric services as reflected inpsychiatric services. Psychiatr Serv 2000;51(1):41-67.

[8] Peele PB, Xu Y, Kupfer DJ. Insurance expenditures on bipolardisorder: clinical and parity implications. Am J Psychiatr 2003;160(7):1286-90.

[9] BuschAB,Yoon F, Barry CL, Azzone V,Normand S-LT, GoldmanHH,et al. The effects of mental health parity on spending and utilization forbipolar, major depression, and adjustment disorders. Am J Psychiatr2013;170(2):180-7.

[10] Simon GE, Unützer J. Health care utilization and costs among patientstreated for bipolar disorder in an insured population. Psychiatr Serv1999;50(10):1303-8.

[11] Kim E, Maclean R, Ammerman D, Jing Y, Pikalov A, You M, et al.Time to psychiatric hospitalization in patients with bipolar disordertreated with a mood stabilizer and adjunctive atypical antipsychotics: aretrospective claims database analysis. Clin Ther 2009;31(4):836-48.

[12] Kozma C, RupnowM. P. 2.079 Evaluation of the relationship betweenantipsychotics and hospitalization in bipolar disorder. Eur Neuropsy-chopharmacol 2004;14:S266-7.

[13] Hoblyn J, Balt S, Woodard S, Brooks J. Substance use disorders as riskfactors for psychiatric hospitalization in bipolar disorder. PsychiatrServ 2009;60(1):50-5.

[14] Cassidy F, Ahearn EP, Carroll BJ. Substance abuse in bipolar disorder.Bipolar Disord 2001;3(4):181-8.

[15] Kessing LV. The effect of comorbid alcoholism on recurrence in affectivedisorder: a case register study. J Affect Disord 1999;53(1):49-55.

[16] Gianfrancesco F, Rajagopalan K. PMH11 Antipsychotic use and hos-pitalization in bipolar or manic patients. Value Health 2005;8(3):387-8.

[17] Gianfrancesco F, Rajagopalan K, Goldberg JF, Wang RH. Hospital-ization risks in the treatment of bipolar disorder: comparison ofantipsychotic medications. Bipolar Disord 2007;9(3):252-61.

[18] Al Jurdi RK, Schulberg HC, Greenberg RL, Kunik ME, Gildengers A,Sajatovic M, et al. Characteristics associated with inpatient versusoutpatient status in older adults with bipolar disorder. J GeriatrPsychiatry Neurol 2012;25(1):62-8.

[19] Kessing LV, Munk-Jørgensen P. Does type of first contact indepressive and bipolar disorders predict subsequent hospitalisationand risk of suicide? J Affect Disord 2004;83(1):65-71.

[20] Edman G, Backlund L, Adler M, Hallgren J, Sparen P, Osby U.Psychiatric hospitalization in bipolar disorder in Sweden. EurPsychiatry 2008;23:S227-8.

[21] Parker G, Fletcher K, McCraw S, Futeran S, Hong M. Identifyingantecedent and illness course variables differentiating bipolar I, bipolarII and unipolar disorders. J Affect Disord 2013;148(2-3):202-9.

[22] DeppCA, Lindamer LA, FolsomDP, Gilmer T, HoughRL,Garcia P, et al.Differences in clinical features and mental health service use in bipolardisorder across the lifespan. Am J Geriatr Psychiatry 2005;13(4):290-8.

[23] Goldstein BI, Levitt AJ. A gender-focused perspective on healthservice utilization in comorbid bipolar I disorder and alcohol usedisorders: results from the national epidemiologic survey on alcoholand related conditions. J Clin Psychiatry 2006;67(6):925-32.

[24] Jing Y, Johnston SS, Fowler R, Bates JA, Forbes RA, Hebden T.Comparison of second-generation antipsychotic treatment on psychi-atric hospitalization in Medicaid beneficiaries with bipolar disorder.J Med Econ 2011;14(6):777-86.

1861J. Shapiro et al. / Comprehensive Psychiatry 55 (2014) 1855–1861

[25] Lazarus A, Lage M, Pesa J. PMH53 Hospitalization rates duringcombination therapy with atypical antipsychotics in bipolar disorder.Value Health 2004;7(3):279-80.

[26] Kim E, You M, Pikalov A, Van-Tran Q, Jing Y. One-year risk ofpsychiatric hospitalization and associated treatment costs in bipolardisorder treated with atypical antipsychotics: a retrospective claimsdatabase analysis. BMC Psychiatry 2011;11(1):6.

[27] Bresnahan B, Ling D, Neslusan C, White A, Wang S, Crown W. Therisk of hospitalization for patients with bipolar disorder. EurNeuropsychopharmacol 2001;11:S210-1.

[28] Guo JJ, Keck Jr PE, Li H, Jang R, Kelton CM. Treatment costs andhealth care utilization for patients with bipolar disorder in a largemanaged care population. Value Health 2008;11(3):416-23.

[29] Pfuntner A, Wier LM, Stocks C. Most frequent conditions in UShospitals, 2010; 2013.

[30] Blader JC. Acute inpatient care for psychiatric disorders in the United States,1996 through 2007. Arch Gen Psychiatry 2011;68(12):1276-83.

[31] Blader JC, Carlson GA. Increased rates of bipolar disorder diagnosesamong US child, adolescent, and adult inpatients, 1996–2004. BiolPsychiatry 2007;62(2):107-14.

[32] Harpaz-Rotem I, Leslie DL, Martin A, Rosenheck RA. Changes inchild and adolescent inpatient psychiatric admission diagnosesbetween 1995 and 2000. Soc Psychiatry Psychiatr Epidemiol2005;40(8):642-7.

[33] OlfsonM, Gameroff MJ, Marcus SC, Greenberg T, Shaffer D. Nationaltrends in hospitalization of youth with intentional self-inflicted injuries.Am J Psychiatr 2005;162(7):1328-35.

[34] Peele PB, Axelson DA, Xu Y, Malley EE. Use of medical andbehavioral health services by adolescents with bipolar disorder.Psychiatr Serv 2004;55(12):1392-6.

[35] Mendenhall AN, Demeter C, Findling RL, Frazier TW, Fristad MA,Youngstrom EA, et al. Factors influencing mental health serviceutilization by childrenwith serious emotional and behavioral disturbance:results from the LAMS study. Psychiatr Serv 2011;62(6):650-8.

[36] Mendenhall AN. Predictors of service utilization among youthdiagnosed with mood disorders. J Child Fam Stud 2012;21(4):603-11.

[37] Wozniak J, Spencer T, Biederman J, Kwon A, Monuteaux M, RettewJ, et al. The clinical characteristics of unipolar vs. bipolar majordepression in ADHD youth. J Affect Disord 2004;82:S59-69.

[38] Rizzo CJ, Esposito-Smythers C, Swenson L, Birmaher B, Ryan N,Strober M, et al. Factors associated with mental health serviceutilization among bipolar youth. Bipolar Disord 2007;9(8):839-50.

[39] Evans-LackoSE, SpencerCS,Logan JE,RileyAW.Patterns and predictorsof restrictive health care service use by youths with bipolar disorder. AdmPolicy Ment Health Ment Health Serv Res 2010;37(5):379-87.

[40] Katz SJ, Kessler RC, Frank RG, Leaf P, Lin E. Mental health care use,morbidity, and socioeconomic status in the United States and Ontario.Inquiry 1997:38-49.

[41] Vasiliadis H-M, Lesage A, Adair C, Wang P, Kessler R. Do Canadaand the United States differ in prevalence of depression and utilizationof services? Psychiatr Serv 2007;58(1):63-71.

[42] Katz SJ, Kessler RC, Lin E, Wells KB. Medication management ofdepression in the United States and Ontario. J Gen Intern Med 1998;13(2):77-85.

[43] Kaufman J, Birmaher B, Brent D, Rao U, Flynn C, Moreci P, et al.Schedule for affective disorders and schizophrenia for school-agechildren-present and lifetime version (K-SADS-PL): initial reliability

and validity data. J Am Acad Child Adolesc Psychiatry 1997;36(7):980-8.

[44] Chambers WJ, Puig-Antich J, Hirsch M, Paez P, Ambrosini PJ,Tabrizi MA, et al. The assessment of affective disorders in children andadolescents by semistructured interview: test-retest reliability of theSchedule for Affective Disorders and Schizophrenia for School-AgeChildren, Present Episode Version. Arch Gen Psychiatry 1985;42(7):696-702.

[45] Axelson D, Birmaher BJ, Brent D, Wassick S, Hoover C, Bridge J,et al. A preliminary study of the Kiddie Schedule for AffectiveDisorders and Schizophrenia for School-Age Children mania ratingscale for children and adolescents. J Child Adolesc Psychopharmacol2003;13(4):463-70.

[46] Birmaher B, Axelson D, Strober M, Gill MK, Valeri S, Chiappetta L,et al. Clinical course of children and adolescents with bipolarspectrum disorders. Arch Gen Psychiatry 2006;63(2):175-83.

[47] Prinz RJ, Foster S, Kent RN, O'Leary KD. Multivariate assessment ofconflict in distressed and nondistressed mother-adolescent dyads.J Appl Behav Anal 1979;12(4):691-700.

[48] Weissman MM, Wickramaratne P, Adams P, Wolk S, Verdeli H,Olfson M. Brief screening for family psychiatric history: the familyhistory screen. Arch Gen Psychiatry 2000;57(7):675-82.

[49] Hollingshead AB. Four factor index of social status; 1975.[50] Martinez JM, Marangell LB, Simon NM, Miyahara S, Wisniewski SR,

Harrington J, et al. Baseline predictors of serious adverse events at oneyear among patients with bipolar disorder in STEP-BD. Psychiatr Serv2005;56(12):1541-8.

[51] Biederman J, Faraone S, Wozniak J, Mick E, Kwon A, Aleardi M.Further evidence of unique developmental phenotypic correlates ofpediatric bipolar disorder: findings from a large sample of clinicallyreferred preadolescent children assessed over the last 7 years. J AffectDisord 2004;82:S45-58.

[52] Goldstein BI, Strober MA, Birmaher B, Axelson DA, Esposito-Smythers C, Goldstein TR, et al. Substance use disorders amongadolescents with bipolar spectrum disorders. Bipolar Disord 2008;10(4):469-78.

[53] Findling RL, Gracious BL, McNamara NK, Youngstrom EA, DemeterCA, Branicky LA, et al. Rapid, continuous cycling and psychiatric co-morbidity in pediatric bipolar I disorder. Bipolar Disord 2001;3(4):202-10.

[54] Vieta E, Gasto C, Otero A, Nieto E, Vallejo J. Differential featuresbetween bipolar I and bipolar II disorder. Compr Psychiatry 1997;38(2):98-101.

[55] Axelson D, Birmaher B, Strober M, Gill MK, Valeri S, Chiappetta L,et al. Phenomenology of children and adolescents with bipolarspectrum disorders. Arch Gen Psychiatry 2006;63(10):1139-48.

[56] Keck PE, McElroy SL, Havens JR, Altshuler LL, Nolen WA, FryeMA, et al. Psychosis in bipolar disorder: phenomenology and impacton morbidity and course of illness. Compr Psychiatry 2003;44(4):263-9.

[57] Biederman J, Kwon A, Wozniak J, Mick E, Markowitz S, Fazio V, et al.Absence of gender differences in pediatric bipolar disorder: findings froma large sample of referred youth. J Affect Disord 2004;83(2):207-14.

[58] Arnold LE, Demeter C, Mount K, Frazier TW, Youngstrom EA,Fristad M, et al. Pediatric bipolar spectrum disorder and ADHD:comparison and comorbidity in the LAMS clinical sample. BipolarDisord 2011;13(5–6):509-21.