Family-based Crisis Intervention with SuicidalAdolescents in the Emergency Room: A Pilot
StudyElizabeth A. Wharff, Katherine M. Ginnis, and Abigail M. Ross
The prevailing model of care for psychiatric patients in the emergency room (ER) is eval-uation and disposition, with little or no treatment provided. This article describes theresults of a puot study of a famuy-based crisis intervention (FBCI) for suicidal adolescentsand their families in a large, urban pédiatrie ER. FBCI is an intervention designed tosufficiently stabilize patients within a single ER visit so that they can return home safelywith their families. Of the 100 suicidal adolescents and their families in the sample, 67met eligibility criteria for FBCI. Demographic and clinical characteristics and dispositionoutcomes from the sample were compared with those obtained retrospectively from amatched comparison group (N= 150). Statistical analyses compared group inpatient ad-mission rates and disposition outcomes. Patients in the puot cohort were significandy lesslikely to be hospitalized than were those in the comparison group (36 percent versus 55percent). Only two ofthe patients in the FBCI cohort were hospitalized immediately afterreceiving the intervention during their ER visit. FBCI with suicidal adolescents and theirfamilies during a single ER visit is feasible and safely limits the need for inpatient psychiat-ric hospitalization, thereby avoiding disruption offamily, academic, and social activitiesand increasing use of less intrusive and more cost-effective psychiatric treatment.
KEYWORDS; crisis intervention;family intervention; suicidal adolescents; suicide
As the adolescent suicide rate has been in-creasing over the last several decades(Centén for Disease Control and Preven-
tion, 1998, 2007a, 2007b, 2008; Office of DiseasePrevention and Health Promotion, 2000), therehas been a parallel increase (as high as 59 percent)in pédiatrie emergency room (ER) usage rates byadolescents in need of mental health evaluationsin the United States (Breslow, Erickson, & Cava-naugh, 2000; EUison, Hughes, & White, 1989;Hughes, 1993; Page, 2000; Sills & Bland, 2002;Stewart, Spicer, & Babl, 2006). Suicidality in ado-lescents has been the most significant factor in themajority of ER visits for behavioral health con-cerns (Stewart et al., 2006) and the most commonpresenting problem for adolescents subsequentlyadmitted to an inpatient psychiatric unit (Brooker,Ricketts, Bennett, & Lemme, 2007).
Although the number of psychiatric ER visitshas increased substantiaUy (Bruffaerts, Sabbe, &Demyttenaere, 2004; Hughes, 1993; Larkin,Claassen, Emond, PeUetier, & Camargo, 2005),child mental health service avaUabUity has not
kept pace, resulting in longer ER wait times andstays for patients (American CoUege of EmergencyPhysicians, 2008), Ekely contributing to a phe-nomenon termed psychiatric "boarding" (Man-sbach, Wharff, Austin, Ginnis, & Woods, 2003)that has gained notoriety in the popular press(Holmberg, 2007; Katz, 2006; Kowalczyk, 2007;Trafford, 2000). Boarding describes a patient whois in psychiatric crisis and requires inpatient hospi-talization but for whom there is no available inpa-tient psychiatric bed (Mansbach et al., 2003). In arecent survey of ER medical directon, over 70percent reported boarding psychiatric patients as aroutine practice, with nearly 40 percent doing soa minimum of once a week (American CoUege ofEmergency Physicians, 2008).
In current practice, the standard of care inemergency psychiatry is evaluation and dispositionwith little or no treatment provided at the time ofpresentation (Bruffaerts, Sabbe, & Demyttenaere,2008). Psychiatric ER protocol is a noteworthydeviation from triage practice in standard emer-gency care, in which the most acute patients are
doi: 10.1093/SW/SWS017 O 2012 National Association of Social Wo-kers 133
prioritized and receive the most rapid and inten-sive care. Historically, there has been little focuson psychiatric treatment within the emergencysetting, often due to time pressures to move pa-tients through the ER and the prevailing treat-ment philosophy that psychiatric treatment ofsuicidal patients requires admission to a locked in-patient facility.
A number of studies evaluating specialized in-terventions occurring within the context of theER have yielded significant increases in after-caretreatment compliance among psychiatric patients(Rotheram-Borus et al., 1996; Spooren, VanHeeringen, & Jannes, 1998) and reductions in de-pressive symptomology (R.otheram-Borus, Piacen-tini, Cantwell, BeUn, & Song, 2000) and suicideattempts (Huey et al., 2004). None, however, havepiloted or evaluated a single-session intervendonthat occurs exclusively within the ER.
Though limited data on the cost-effectiveness ofalternatives to inpatient hospitalization are avail-able (Lamb, 2009; Shepperd et al., 2009),community-based interventions like multisystemictherapy (MST) show promising results; specifically,in a randomized controlled trial of 116 adolescentsmeeting criteria for inpatient hospitalization re-ceiving either home-based MST or inpatienthospitalization, higher levels of patient satisfaction,improvement in family functioning, and reduc-tions in externalizing symptoms were reported inthe MST group than in the group receivinginpatient hospitalization (Henggeler et al., 1999).Because the ER is frequently a critical point ofcontact for suicidal adolescents to receive access toservices, we developed a farrdly-based crisis inter-vention (FBCI) for use exclusively in the ER,with the explicit goal of decreasing acute symp-toms and sending more suicidal adolescents homesafely with their families.
FBCI is based on the assumptions that an inpa-tient hospitalization is not necessarily the mosthelpful level of psychiatric care for adolescentswith suicidal ideation/behavior, that families andcaregivers are able to provide support to an adoles-cent fanuly member if given both an opportunityand effecdve tools to use, and that a family thatlearns to support an adolescent while he or she is incrisis wül be empowered to provide ongoingsupport once the acute psychiatric crisis subsides.Based on an integradon of cognidve-behavioral skillbuilding, psychoeducadon, therapeudc readiness.
and safety planning, FBCI uses nonjudgmentalcollaboradon (Madsen, 1999) to stabilize padentsand provide psychiatric intervention in the ER forboth the adolescent and the family, thereby de-creasing a patient's level of risk and increasing thecapacity of the family to maintain the patient athome with appropriate therapeutic supports.
This two-part puot study explored (1) thesafety and feasibility of FBCI in a population ofadolescents presenting with suicidal complaints ina large urban pédiatrie ER and (2) dispositionoutcomes between the pilot sample and a com-parison sample obtained retrospectively during theidentical calendar period immediately precedingthe FBCI study period. It was hypothesized thatFBCI during an ER visit would prove both feasi-ble and safe and that rates of inpatient psychiatrichospitalization in the sample of patients presendngduring the FBCI study period would be lowerthan those in the retrospective cohort sample.
METHOD
Study DesignThis puot study was conducted in the BostonChildren's Hospital ER, in which approximately1,000 patients in psychiatric crisis are seen annually.Nearly 40 percent of these patients present•with chief complaints of depression or suicidalideation/behavior. During an 18-month period(January 1, 2001, through June 30, 2002), 100suicidal adolescents and their families were re-cruited to participate in the pilot study of FBCIwhen presenting to the ER. The sample was ob-tained consecutively. Padents were excludedwhen they met at least one of the followingfive criteria: (1) not currently living with a fanuly,(2) presenting to the ER unaccompanied by afamily member, (3) intoxicated/sedated at thetime of ER presentation, (4) presenting with cog-nitive limitations that prohibited FBCI participa-tion (that is, severe psychosis or significantdevelopmental delay), and (5) presenting duringan overnight shift (11:00 P.M. through 8:00 A.M.,Monday through Friday) or during weekendhours (5:00 P.M. Friday through 8:00 A.M.Monday), because FBCI-trained staff were notavailable to administer the intervention duringthese ER shifts. Informed consent and assent wereobtained from all patients and families prior topatients receiving both a standard psychiatric
134 Social Work VOLUME 57, NUMBER 2 APRIL 2012
evaluation and FBCI. After finishing the standardpsychiatric evaluation with the chUd and famUy,the social worker used her best chnical judgmentto make the decision about whether the patientcould benefit from FBCI. If the evaluating socialworker had any uncertainty about whether thepatient was appropriate for FBCI, a supervisor wasavaUable for consultation.
A pUot design was selected because the entireER psychiatry staff was trained in the interven-tion protocol, rendering random assignment offamihes to standard or specialized ER care im-possible. Prior to commencing the study, aU ERpsychiatry social work staff members were trainedin FBCI protocol by the creators of the inter-vention. FBCI staff were required to attendweekly meetings to review cases with the crea-tors of the intervention. Fidelity to the interven-tion was measured using a checklist requiringcompletion of each of the four core essentialcomponents of FBCI. Interrater rehabUity wasestablished prior to study start. FBCI staff metweekly during the 18-month pUot study periodto maintain interrater reHabUity. The safety ofFBCI was measured by the number of FBCIpatients who reported incidence of a suicideattempt or completion during the three-monthfoUow-up period. FeasibUity was measured byour abUity to adequately train ER staff in FBCIprotocol and implement the single-session inter-vention within the context of a busy ER. Fideh-ty to the intervention was measured using achecklist requiring completion of each of thefour core essential components of FBCI. Demo-graphic and clinical characteristics and dispositiondata from the pUot sample were then comparedwith data obtained retrospectively from a cohortof suicidal adolescents presenting consecutivelyto the ER during the previous 18-month calen-dar period (fanuary 1, 1999, through June 30,2000). The Boston ChUdren's Hospital institu-tional review board approved the pUot study.
Recruitment and Consent for the PilotStudyAU patients and famUies presenting to the ERduring the study period received standard emer-gency care, or treatment as usual (TAU). Thisprocess began with a medical examination by anER physician. Once medical clearance was ob-tained and a psychiatric consultation was
requested, participants were approached by a psy-chiatry research assistant to obtain informedconsent/assent. Once consent was obtained, thepatient and famUy were asked to complete somebrief psychometric measures—the ChUdren'sDepression Inventory (CDI) (Kovacs, 1982), theHopelessness Scale for ChUdren (HSC) (Spirito,WUhams, Stark, & Hart, 1988) and the FamUyAdaptability and Cohesion Evaluation Scale II(FACES II) (Olson, Portner, & Lavee, 1982)—toassess depression, hopelessness, and flexibUity offamUy system, respectively. AU psychometric mea-sures used have demonstrated strong reliabUityand vahdity (Kazdin, Rogers, Sc Colbus, 1986;Kovacs, 1992; Olson et al., 1983). Famihes werealso asked to complete a comprehensive famUyself-report form, which coUected demographicand historical information. Patients and familieswere not excluded if the famUy or adolescent didnot complete the assessment forms. AU formswere avaUable in Spanish and English. Interpreterswere avaUable for famUies whose first languagewas other than English or Spanish. The StudyFlow is depicted in Figure 1.
Determination of Inclusion for FBCIOn completion of the standard emergency psy-chiatry evaluation with the adolescent and family,the evaluating chnician, either a licensed master's-or doctoral-level social worker, reviewed the casewith the attending psychiatrist to determine theappropriate level of psychiatric care. FamUies wereoffered FBCI only if the evaluating social worker,attending psychiatrist, and famUy were in agree-ment that FBCI might enable the adolescent toreturn home safely. The decision to offer FBCIwas based on both the acuity of the adolescent'ssuicidality and the capacity for galvanizing envi-ronmental supports avaUable to and within thefamUy system. If the evaluating social worker, at-tending psychiatrist, and famUy concurred thatFBCI might help the adolescent return homesafely with his or her famUy, the social workerthen proceeded with FBCI. Adolescents whowere not offered FBCI were hospitalized at an in-patient psychiatric facUity.
FBCIThe thoretical underpinnings of this single-sessionER intervention come from cognitive-behavioral,narrative, and family systems therapies, with an
WHARFF, GINNIS, AND ROSS / Family-based Crisis Intervention with Suicidal Adolescents in the ER 135
Figure 1: Study Flow Chart
Patient presents to ER withsuicidality and is medically cleared
ExclusionCriteria:Acuity ofsuicidality
• Study team obtains informedconsent and administers CDI, HSC,
and FACES II• Clinician performs standard
psychiatric evaluation anddetermines FBCI eligibility
Patient is ::iot FBCIeligible: Disposition ofinpatient tospitalization(« = 33)
Exclusion Critetia:• Not currently living
with a family• Presenting to ER
unaccompanied by afamily member• Intoxicated/
sedated at the time ofER presentation• Severe cognitive
limitations thatprohibited FBCIparticipation• Presenting during an
ovemight/ weekendshift
Patient is FBCI eligible(n = 67)
Clinician does FBCI inER: Meets with patientand family bothindividually and together
Clinician consults withattending psychiatrist,adolescent, and family todetermine whether thepatient can retum homesafely
Patient is hospitalized (n =
2)Patient returns home safely(« = 65)
Note: ER = emergency room; CDI = ChildreiTs Depressio • Inventory; HSC = Hopelessness Scale for Children; FACES II = family Adaptability and Cohesion Evaluation Scale II;FBCI = family.based crisis intervention.
overaU approach of nonjudgmental collaboration,as described by Madsen (1999;. First, the socialworker holds separate meetings with the adoles-cent and family to assess the sequence of eventsand differing perceptions leading to the suicidalproblem. During these meetings, the socialworker uses a narrative approach to help eachparty tell his or her story. The social worker alsoexplores what each party feels would be necessaryfor the adolescent to retum home safely with hisor her family. Next, the social worker meets with
the whole family together, attempting to con-struct a single, unified perception of the problemusing the same narrative approach. We refer tounified perception of the problem as the "jointcrisis narrative." During the meeting, the socialworker assesses family roles and the potential flex-ibUity and adaptabiUty of the family system, usingcUnical interventions to both facilitate andimprove communication among family members.The social worker uses cognitive—behavioral ther-apeutic approaches, including relaxation.
136 Social Work VOLUME 57, NUMBER 2 APRIL 2012
problem-solving, and cognitive reframing tech-niques to shift negative atttibutions. The socialworker also works with the chud and family toproblem solve around any specific dilemmas asneeded. In the family meeting, the social workertties to help the family and adolescent work to-gether to improve intrafamilial communication, tosafety plan, and to effect additional changes thatwill enable the adolescent to feel safe at home.The goal of FBCI is to effect changes that wOlreduce the acute symptoms that brought the ado-lescent to the ER and increase the family's aware-ness of the problem and sense of efficacy to helptheir chud, thereby avoiding inpatient admissionsand further disruptions of the adolescent's hfe.After completing FBCI, the social worker againconsults with the supervising psychiattist toreview the case and obtain consensus that thepatient is able to return home safely. Patients aredischarged home only when the patient, family,attending psychiattist, and assessing social workeragree that this is the best disposition for theadolescent.
Follow-up AssessmentsFive follow-up assessments were completed by astudy clinician via telephone at one-day, one-week, two-week, one-month, and three-monthintervals from the date of the ER visit. Follow-upassessments served the dual purpose of obtaininginformation about the patient's level of function-ing and facilitating acquisition of additional sup-portive services as needed. Follow-up assessmentswere completed only for those adolescents andfamilies who were discharged home after theirER visit. Data on incidence of subsequent psychi-attic evaluations and inpatient hospitalizationswere also collected at these five follow-upintervals.
Retrospective Comparison GroupDemographic and clinical charactetistics and dis-position outcomes of patients in the pilot samplewere compared retrospectively with adolescents{N= 150) who presented consecutively to thesame ER with complaints of suicidal behavior/ideation duting the corresponding 18-month cal-endar petiod immediately preceding the puotstudy petiod Qanuary 1, 1999, through June 30,2000). Retrospective cohort patients met thesame inclusion and exclusion ctitetia as those in
the pilot sample. Standard psychiattic assessmentinformation—including demographic informa-tion. Diagnostic and Statistical Manual of Mental Dis-orders (4th ed.) {DSM-IV) (Ametican PsychiatticAssociation, 1994) diagnoses, and disposition de-tennination—that pertained to each patient in theretrospective compatison group was obtainedthrough a medical record review.
Statistical AnalysesFrequency disttibutions and means were calculat-ed for demographic data in both the puot andretrospective cohort compatison samples (seeTable 1). Mean CDI, HSC, and FACES II adapt-ability and cohesion scores were computed for thesample using a dichotomous disposition outcome(inpatient hospitahzation/aU others) as the depen-dent vatiable (see Table 2). Between-groups dif-ferences in mean CDI, HSC, and FACES IIscores were examined using independent sampleÍ tests (see Table 2).
Chi-square analyses were used to examine dif-ferences in demographic vatiables and dispositionoutcomes between the pilot and retrospectivecohort samples. Disposition outcomes were cate-gotized by level of support (inpatient, intensiveoutpatient, and outpatient services). Analyses ofdisposition outcome rates between the pilot andretrospective cohort samples are reported inTable 3.
RESULTS
DemographicsThe pilot sample included 100 adolescents ages13 to 18 years presenting to a large urban ERwith symptoms of suicidality. A total of 144 suici-dal adolescents presented to the ER duting theFBCI study petiod, 44 of whom were excludedfrom participating in the study due to aforemen-tioned exclusion ctitetia, lack of available researchassistants, or lack of available FBCI-trained staff.
Among the 100 adolescents participating in thepilot study, 76.0 percent were female. Meanpatient age was 15.6 {SD= 1.5) years. Sixty-fivepercent of patients self-identified as white, 16percent self-identified as black, 11 percent self-identified as Hispanic/Latino, 3 percent self-identified as biracial, 2 percent self-identified asAsian, and 3 percent self-identified as being ofanother race. Demographic data obtained
WHARFF, GINNIS, AND ROSS / Family-based Crisis Intervention with Suicidal Adolescents in the ER 137
Table 1: Demographics of Intervention and Comparison Groups
iiteryention'(/V= 100)
VariableComparison (/V={150) ¡
GenderFemaleMale
Race/ethnicity
AsianBlackHispanic/LatinoWhiteBiracialOther
Living arrangement
ParentsOther relativeFoster care
Primary languageEnglish
SpanishOther
Legal custody
ParentsDSS/DCFOther relativeOther
Primary DSM-IVxás I diagnosis
Total depressive disordersBipolar disorder
Other mood disordersAnxiety disorders/PTSDOther'
76
24
2
16
11
65
3
3
96
4
0
89
7
4
96
1
2
1
76
5
1
8
II
76.0
24.0
2.0
16.0
11.0
65.0
3.0
3.0
96.0
4.0
0.0
89.0
7.0
4.0
96.0
1.0
2.0
1.0
76.0
5.0
1.0
8.0
11.0
111
39
4
26
15
97
2
6
139
11
1
138
11
1
140
5
1
0
105
10
2
10
23
74.0
26.0
2.7
17.3
11.0
64.7
1.3
4.0
92.7
7.3
0.7
92.0
7.3
0.7
93.3
3.3
0.7
0.0
70.0
6.7
1.3
6.6
15.3Notes: The average age in years was 15.60 (SD=1.4S) for the intervention group and 15.50 {SD=1.47) for the comparison group. DSS = Department of Social Sen/ices; DCF =Department of Children and Families; DSM-IV= Diagnostic and Statistical Manual of Mental Disorder (4th ed.); PTSD = posttraumatic stress disorder.'Includes eating disorder, psychosis, substance abuse behavioral disorders, attention-deficit/hyperactivity disorder, and somatoform disorders.
Table 2: Pilot Sample CDI, FACES II, and
HSC Scores
Total CDI
Patient cohesion
Family cohesion
Patientadaptability
Familyadaptability
Hopelessness
InpatientAll othersInpatientAll othersInpatient
All others
InpatiencAll others
InpatientAll othersInpatient
All others
346131
593360
3254
32582760
31.5623.26
47.1947.6354.7356.02
38.6640.56
44.7244.81
11.157.52
9.8639.83
12.38
12.9258.5968.54
9.46
9.07
7.106.324.364.44
<.OO1
.876
.490
.364
.952
.001
Notes: W=100. CDI = Children's Depression Inventory; HSC = Hopelessness Scale forChildren; FACES II = Family Adaptability and Cohesion Evaluation Scale II.
Table 3: FBCI and Comparison Group
Disposition Outcomes, in Percentages
DispositionOutcomeInpatientIntensive outpatientOutpatientOther
FBCI(/V=100)
35"21"
430
Comparison(W=150)
555
373
Note: FBCI = family.based crisis intervention."Reduction in hospitaiization rate: p<.0001.''increase in intensive outpatient referrai: p < .001.
retrospectively from the comparison sample (150
suicidal adolescents presenting consecutively to
the ER during the corresponding previous
18-month calendar period) are presented in
Table 1. Patients in the comparison sample did
not differ significantly in age, race and ethnicity,
138 Social Work VOLUME 57, NUMBER 2 APRJL 2012
living arrangements, primary language, legalcustody status, caregiver relationship status,primary DSM-IV diagnosis, or reported familyhistory of depression from their counterparts inthe pilot study. Exclusionary criteria werematched between samples.
On compledon of the initial psychiatric evalua-tion, 67 percent of adolescents (n = 67) were eligi-ble to receive FBCI. The remaining 33 percent(n = 33) who did not receive FBCI were hospital-ized due to the acuity of their suicidality. Of the67 patients who received FBCI, 97.0 percent{n = 65) were not hospitalized. Only two padentswho received FBCI were hospitalized after theirER visit. These patients were unable to engage insafety planning during FBCI and thus requiredhospitalization. Statistically significant differencesin depressive symptom severity occurred betweenpatients who were admitted to an inpatient unitand those who were not. Mean CDI scores for34 patients with an inpatient disposition (31.50[SD = 69.86]) were significantly greater than themean CDI scores for 60 patients who did not(23.26 [5D = 9.83]). CDI scores were not avail-able for two patients who were ineligible forFBCI and for three who received the interventionand were discharged home. Similarly, mean HSCscores were higher among patients who were hos-pitalized (n = 27 [p = .001]). HSC scores v/ere notavailable for eight patients deemed ineligible forFBCI and for five patients who received FBCIand were discharged home.
Neither CDI nor HSC scores differed signifi-cantly between patients receiving an inpatienthospitalization and those receiving intensive out-patient treatment. Differences in patient andfamily FACES II cohesion or adaptability subscalesdid not approach significance for any dispositioncategory (see Table 2).
Suicidal adolescents and families who presentedto the ER during the FBCI pilot period v/ere sig-nificantly less likely to be admitted to an inpatientpsychiatry unit than were members of thematched sample who presented during the com-parison period. Sixty-five percent of suicidal pa-tients presenting during the study period weredischarged home, whereas only 44.7 percent ofthe comparison cohort (n = 67) were dischargedhome. Adolescents and their families presentingto the ER during the pilot study period v/ere sig-nificandy more likely to receive a referral to
intensive outpatient services (acute day treatmentprograms and intensive home-based therapies) atdischarge fi'om the ER (21.0 percent [n = 21]venus 5.3 percent [n = 8], p<.001) than weretheir TAU counterparts in the retrospectivecohort (see Table 3).
Of the 65 patients and families who receivedfoUow-up assessments at five separate intervals asa component of the intervention protocol, 43(66.1 percent) were reached at one day, 44 (67.7percent) were reached at one week, 42 (64.6percent) were reached at two weeks, 44(67.7 percent) were reached at one month, and 36(55.4 percent) were reached at three months. Atotal of 55 patients (84.6 percent) were reached atleast once during the foUow-up period. Pearsonchi-square tests revealed no significant differencesbetween patients reached at foUow-up and thoseunable to be reached in age, gender, primary axis Idiagnosis, CDI or HSC scores, or insurance cate-gories. No patients reported incidence of attempt-ed or completed suicide during the three-monthfoUow-up period.
None of the patients for whom data wascoUected at the one-day foUow-up required aninpatient hospitalization. At the three-monthfoUow-up, seven patients reported requiring aninpatient hospitalization since the initial ER visit(12.7 percent), only two (3.6 percent) of whomwere hospitalized because of suicidal complaints.Other reasons for hospitalization included transi-tion between partial hospitalization placements,decompensadon due to schizophrenia, self-injuriousbehavior (nonsuicidal), and psychiatric evaluadonrequired prior to entering child protecdve custody.
DISCUSSIONConsidering the increasing rates and high costs ofadolescent psychiatric hospitalization and an in-creasingly overburdened health-care system, inves-tigations evaluating the efficacy of therapeuticinterventions occurring within the ER are essen-tial. To date, FBCI is the only standardized single-session crisis intervention for suicidal adolescentsevident in the literature that has been designed forand pUoted within the ER to demonstrate feasi-bility, acceptabUity, and significant reductions ininpatient hospitalization rates relative to a demo-graphicaUy matched, retrospectively obtainedcomparison sample.
WHARPF, GINNIS, AND ROSS / Family-based Crisis Intervention with Suicidal Adolescents in the ER 139
Avoidance of inpatient psychiatric admission forsuicidal adolescents has several benefits for the in-dividual adolescent, the familŷ and the mentalhealth system. Although the sodetal stigma associ-ated with mental health problems has beenreduced somewhat, the stereotypical view of ado-lescent inpatient psychiatric care depicted inpopular literature and films continues to prevail.An inpatient admission may negatively affect anindividual or family's beliefs about recovery(Hellzen & Lilja, 2008), the capacity to be safe inthe world, or the family's ability to provide a safeand supportive environment for their chud. Con-venely, community-based supports may allow achud to refrain from developir.g a "dependencyupon the hospital environment or from beingstigmatized" (Shepperd et al, 2009, p. 3). AnER-based crisis intervention provides the adoles-cent and family with the message that, despite thesuicidal ideation/behavior with, which they pre-sented, there are skills that famlies can learn thatwill enable the adolescent to alleviate his or herdistress and thus remain at home. The family feelsempowered to be the coordinator of and partici-pant in their child's care, comparable to the foun-dational empowerment model used by thecommunity intensive therapy team in the UnitedKingdom (Darwish, Salmon, Ahi^a, & Steed, 2006).roCI provides psychoeducation to promote en-gagement in therapy and (¡mraiy understanding oftreatment. In addition to the tangible parts of the in-tervention, FBCI provides hope for a family thatarrived at the ER overwheln^ed, anxious, andworried for a child's survival. During the study, sev-eral families expressed relief and gmdtude for the carethat they received in the ER and noted duringfollow-ups that family communication and function-ing in home and school domains had improved.
The absence of significant differences betweenfamily adaptability or cohesion TACES II scores)and hospitalization in the intervention group wasan unexpected finding. Even the most seeminglyinflexible and uncommunicati'^e families couldengage with a skilled clinician to participate inpsychiatric treatment of their child. Using Parad's(1965) crisis theory approach, ve posit that eventhe most rigid family system is open to changeduring a crisis. FBCI allows cHricians to take ad-vantage of this opportunity, thereby avoiding un-necessary psychiatric hospitalization.
The significant increase in referrals for intensiveoutpatient treatment in the puot sample indicatesthat these adolescents were clearly in need of in-tensive mental health support; however, FBCIenabled chnicians to join with families to providethem with the tools needed to care for their chil-dren safely at home, allowing this intensive treat-ment to occur outside the hospital. As previouslynoted, FBCI incorporates cognitive therapeutictechniques to reframe negative attributions. Arecent study of a 12-session cognitive-behavioraltreatment for suicide prevention that also includescognitive refcaming as a key component has dem-onstrated feasibility in a population of suicidal ad-olescents (Stanley et al., 2009), providing furtherempirical support for cognitive-behavioral treat-ment techniques for suicidahty specifically. FBCIcould be a part of a growing number of morecost-effective alternatives to inpatient hospitaliza-tion—such as multisystemic therapy (Henggeler &Borduin, 1990; Henggeler et al., 1997; Hueyet al., 2004; Schoenwald, Ward, Henggeler, &Rowland, 2000) and rapid-response outpatientmodels (Greenfield, Larson, Hechtman, Rous-seau, & Platt, 2002)—that have been shown to beas feasible as inpatient hospitalization for treatingsuicidality in adolescents presenting to the ER.The absence of significant differences in HSC andCDI scores between those receiving dispositionsof an inpatient hospitalization and intensive out-patient services within the puot study group indi-cates that safety can be estabhshed at home foreven severely depressed adolescents.
The follow-up component of the FBCI proto-col also yielded promising results. Of the 55 pa-tients (85.9 percent) reached during thethree-month follow-up period, none required animmediate inpatient hospitalization (within oneweek). Remarkably, only two FBCI patients (3.6percent) reached during the three-monthfoUow-up period required an inpatient hospitali-zation due to suicidal ideation or behavior.
The medical system uses a model of stabihza-tion in the ER whenever possible and admissiononly when necessary. FBCI and other crisis inter-vention protocols could help the mental healthsystem move to a similar model in which inpatientadmission is no longer the default position.Current trends indicate that the majority of ado-lescents presenting with suicidal ideation/behavior
140 Social Work VOLUME 57, NUMBER 2 APRIL 2012
are admitted to psychiatric inpatient facilities(Brooker et al., 2007). Inpatient hospitaUzation ismuch more costly than other alternatives thathave been shô wn to be as effective in reducingsuicidal ideation/behavior in adolescents (Gould,Greenberg, Velting, & Shaffer, 2003; Henggeleret al., 2003).
This change in mindset is critical for the system toappropriately respond to and care for patients alongthe continuum of care. Patients who can go homewith intensive outpatient follô w-up do not board inthe ER, where they •wül receive minimal psychiatrictreatment. Instead, they go home with outpatientservices in place, allowing them to engage in treat-ment more quickly than they would if they weresitting in the ER awaiting an inpatient bed.
LIMITATIONSThere were several limitations to this study.Though random assignment to FBCI or TAUconditions would have been preferable, FBCIinvolved changing clinicians' practice within theER; it was not possible to randomly assign famiUesto treatment and control groups. The closest avail-able approximation to a control group was a retro-spectively obtained sample from the most recentcorresponding 18-month calendar period prior tocUnician training in FBCI. The retrospective com-parison component of the study also prohibitedacquisition of data on the frequency of ER visits inthe TAU group; thus, ER recidivism rates couldnot be analyzed comparatively between the twosamples. The foUô w-up duration of three monthsis also a Umitation. Though alternatives to inpa-tient hospitalization have been shown to be moreeffective in extemaUzing symptom reduction andimprovement in family functioning than inpatienthospitalization (Henggeler et al., 1999; Schoen-wald et al., 2000), research indicates that thesegains may be relatively short-lived (Henggeleret al., 2003). We hypothesize that FBCI mayreduce ER recidivism and hospitaUzation rates.Studies of pédiatrie emergency psychiatry servicesindicate that multiple presentations account for 19percent to 36 percent of ER visits and that ap-proximately 50 percent of repeat •visits occurwithin one month of the prê vious presentation.Though data were unavailable on foUow-up hos-pitalizations for the comparison group, rates ofrepeat ER presentations were much lower in theFBCI sample than in other samples. Currently, we
are conducting a randomized cUnical trial that ex-amines the efficacy of FBCI, long-term gains, andeffects of FBCI on ER recidî vism rates. Althoughwe did place foUow-up phone caUs to aU studyparticipants, no famiUes required assistance in ac-cessing additional services at any of the foUow-uptime points; however, other services (beyondthose recommended as part of the discharge plan)were not controUed for in this study. Futurestudies should control for additional service useand variability.
CDI, HSC, and FACES II scores were not col-lected posttreatment or during foUow-up and,therefore, could not be analyzed. No psychomet-ric measures were obtained for the TAU group atany time point, as there was no practical way toproceed because of the nature of the populationpresenting to the ER and human subjects issues.In addition, frequency and dose of each of thefour core components of FBCI were not coUectedas part of this study. Future studies exploring theefficacy and effectiveness of FBCI should incorpo-rate the frequency and dose of each of these corecomponents.
CONCLUSIONS
Patients who received FBCI were significantly lessUkely to be hospitaUzed than were their compari-son cohort counterparts. Suicidal adolescents pre-senting in crisis to the ER were able to be senthome safely with appropriate therapeutic supports.This puot study demonstrates that a single-visitmodel of crisis intervention for suicidal adoles-cents and their famiUes deUvered in the ER cansufficiently stabilize an adolescent and familysystem, regardless of cohesion and adaptabiUtylevels, to enable a safe discharge home.
It is essential to begin to use a model of crisisintervention for suicidal adolescents and theirfamilies to provide reUef from their acute symp-toms with the least amount of family disruption.Empirical ê vidence has yet to document the supe-riority of inpatient care in effectively reducingrates of suicidal ideation, nonlethal attempts, orcompleted suicides among adolescents (Gouldet al, 2003). FBCI benefits the adolescent andfamily and simultaneously alleviates an overbur-dened mental health system by limiting use ofscant inpatient resources, keeping them availablefor those who truly need them. FBCI is a stan-dardized protocol that could be used by crisis
WHARFF, GINNIS, AND ROSS / Family-based Crisis Intervention with Suicidal Adolescents in the ER 141
clinicians across contexts to provide this kind ofintervention, and it may be a cost-effective andadvantageous alternative to inpatient hospitaliza-tion for both patients and providers.
REFERENCESAmerican College of Emergency Phyäcians. (2008).
ACEP Psychiatric and Substance Aiiuse Survey 2008.Retrieved firom http://www.acep.org/uploadedFiles/ACEP/Advocacy/federal_issues/PsychiatricBoardingSummary.pdf
American Psychiatric Association. (1994). Diagnostic andstatistical manual of mental disorders (4th ed.). Washing-ton, DC: Author.
Breslow, R. E., Erickson, B. J., & Cavanaugh, K. C.(2000). The psychiatric emergency service: Wherewe've heen and where we're going. Psychiatric Quar-terly, 71, 101-121.
Brooker, C , Ricketts, T , Bennett, S., & Lemme, F.(2007). Admission decisions following contact withan emergency mental health assessment and interven-tion service. joHma/ of Clinical Nursing, 6, 1313-1322.
Bruffaerts, R., Sabbe, M., & Demytteaaere, K. (2004). At-tenders of a university hospital psychiatric emergencyservice in Belgium—General chaacteristics andgender differences. Social Psychiatry and Psychiatric Epi-demiology, 39, 146-153.
Bruffaerts, R., Sabbe, M., & DemytteBaere, K. (2008).Emergency psychiatry in the 21s! century: Criticalissues for the future. European Journal of EmergencyMedicine, 15, 276-278.
Centers for Disease Control and Prevention. (1998, August14). Youth risk behavior surveillixice—United States,1997. Morbidity and Mortality WeMy Report, 43(SS-3),1-18.
Centers for Disease Control and Prevention. (2007a, Sep-tember 7). Suicide trends among youths and youngadults aged 10-24 years—United States, 1990-2004.Morbidity and Mortality Weekly Report, 56, 905-908.
Centers for Disease Control and Prevention. (2007b).Web-based injury statistics query and reportingsystem. Atlanta, GA: U.S. Deparanent of Health andHuman Services.
Centers for Disease Control and Prevention. (2008, June6). Youth risk behavior surveillance—United States,2007. Morbidity and Mortality Weekly Report, 57(SS04),1-131.
Darwish, A., Salmon, G., Ahuja, A., &: Steed, L. (2006).The community intensive therapy team: Develop-ment and philosophy of a new service. Clinical ChildPsychology and Psychiatry, 11, 591-605.
ElHson, J. M., Hughes, D. H., & White, K. A. (1989). Anemergency psychiatry update. Hospital & CommunityPsychiatry, 40, 250-260.
Gould, M. S., Greenberg, T., Velting,D. M., & Shaffer,D. (2003). Youth suicide risk and preventive inter-ventions: A review of the past 10 years. Jowma/ of theAmerican Academy of Child and Adolescent Psychiatry,42, 386-405.
Greenfield, B., Larson, C , Hechtman,. L., Rousseau, C , &Platt, R. (2002). A rapid-response outpatient modelfor reducing hospitalization rates imong suicidal ado-lescents. Psychiatric Services, 53, 1574-1579.
Hellzen, O., & Lilja, L. (2008). Former patients' experi-ences of psychiatric care: A qualitative investigation.International Journal of Mental Health Nursing, 17,279-286.
Henggeler, S. W., & Borduin, C. M. (1990). Familytherapy and beyond: A multisystemic approach to treatingthe behavior problems of children and adolescents. PacificGrove, CA: Brooks/Cole.
Henggeler, S. W., Rowland, M. D., Halliday-Boykins, C ,Sheidow, A. J., Ward, D. M., Randall, J., et al.(2003). One-year follow-up of multisystemic therapyas an alternative to the hospitalization of youths inpsychiatric crisis. _/oHma/ of the American Academy ofChild and Adolescent Psychiatry, 42, 543-551.
Henggeler, S. W., Rowland, M. D., Pickrel, S. G., Miller,S. L., Cunningham, P. B., Santos, A. B., et al. (1997).Investigating family-based alternatives to institution-based mental health services for youth: Lessonslearned fi:om the pilot study of a randomized fieldtrial. Jowma/ of Clinical Child Psychology, 26, 226-233.
Henggeler, S. W., Rowland, M. D., Randall, J., Ward,D. M., Pickrel, S. G., Cunningham, P. B., et al.(1999). Home-based multisystemic therapy as an al-ternative to the hospitalization of youths in psychiatriccrisis: Chnical outcomes. JoMma/ of the AmericanAcademy of Child and Adolescent Psychiatry, 38,1331-1339.
Holmberg, D. (2007, June 10). The patients in the hall-ways. New York Times. Retrieved from http://w^vw.nytimes.com/2007/06/10/nyregion/nyregionspecial2/10Rhospitals.html
Huey, S. J. Jr., Henggeler, S. W., Rowland, M. D.,Halhday-Boykins, C. A., Cunningham, P. B.,Pickrel, S. G., et al. (2004). Multisystemic therapyeffects on attempted suicide by youths presenting psy-chiatric emergencies. Jouma/ of the American Academyof Child and Adolescent Psychiatry, 43, 183-190.
Hughes, D. H. (1993). Trends and treatment models inemergency psychiatry. Hospital & Community Psychia-try, 44, 927-928.
Katz, A. (2006, April 16). Patients wait for hours in hall-ways: Strain felt throughout state. New Haven Register,pp. 15-17.
Kazdin, A., Rogers, A., & Colbus, D. (1986). The Hope-lessness Scale for Children: Psychometric characteris-tics and concurrent validity. Journal of Consulting andClinical Psychology, 51, 241-245.
Kovacs, M. (1982). Children's Depression Inventory (CDI)short form. Toronto: Multi-Health Systems Inc.
Kovacs, M. (1992). Children's Depression Inventory (CDI)short form. Toronto, Canada: Multi-Health SystemsInc.
Kowalczyk, L. (2007, February 3, 2005). Romney's Med-icaid cuts hit hard, hospitals say. Boston Globe, pp. Al,C5.
Lamb, C. E. (2009). Alternatives to admission for childrenand adolescents: Providing intensive mental health-care services at home and in communities: Whatworks? Current Opinion in Psychiatry, 22, 345-350.
Larkin, G. L., Claassen, C. A., Emond, j . A., Pelletier, A.J., & Camargo, C. A. (2005). Trends in U.S. emer-gency department visits for mental health conditions,1992 to 2001. Psychiatric Services, 56, 671-677.
Madsen, W. C. (1999). Collaborative therapy with multi-stressed families: From old problems to new futures.New York: Guilford Press.
Mansbach,J. M., Wharff, E., Austin, S. B., Ginnis, K., &Woods, E. R. (2003). Which psychiatric patientsboard on the medical service? Pediatrics, 11 í(6, Pt. 1),e693-e698.
Office of Disease Prevention and Health Promotion.(2000). Healthy People 2010. Rockville, MD: U.S.department of Health and Human Services.
Olson, D. H., McCubbin, H. I., Bames, H. L., Larsen, A.S., Muxen, J. J., & Wilson, M. A. (1983). Families:
142 Sociat Work VOLUME 57, NUMBER 2 APRIL 2012
What makes them work. Beverly Hills, CA: SagePublications.
Olson, D. H., Portner, J., & Lavee, Y. (1982). FamilyAdaptability and Cohesion Evaluation Scales II. St. Paul:University of Minnesota.
Page, D. (2000). Pédiatrie psychiatry: More blues in ED?Hospital & Health Networks, 74(8), 24.
Parad, H. J. (1965). Crisis intervention: Selected readings. NewYork: Fatnily Service Association of America.
Rotberam-Borus, M. J., Piacentini, J., Canrwell, C , Belin,T. R., & Song, J. (2000). Tbe 18-month impact ofan emergency room intervention for adolescentfemale suicide attempters. Jowma/ of Consulting andClinical Psychology, 68, 1081-1093.
Rotberam-Borus, M. J., Piacentini, J., Van Rossem, R.,Graae, E, Cantwell, C , Castro-Bianco, D., et al.(1996). Enbancing treatment adberence witb a spe-cialized emergency room program for adolescentsuicide attempters. Joumu/ ofthe American Academy ofChild and Adolescent Psychiatry, 35, 654-663.
Scboenwald, S. K., Ward, D. M., Henggeler, S. W., &Rowland, M. D. (2000). Multisystemic tberapy versusbospitalization for crisis stabilization of youtb: Place-ment outcomes 4 montbs postreferral. Mental HealthServices Research, 2, 3—12.
Sbepperd, S., Doll, H., Gowers, S., James, A., Fazel, M.,Fitzpatrick, R., et al. (2009, April 15). Alternatives toinpatient mental bealtb care for children and youngpeople. Cochrane Database of Systematic Reviews(2),CD006410. doi: 10.1002/14651858.CD006410.pub2
Sills, M. R., & Bland, S. D. (2002). Summary statistics forpédiatrie psychiatric visits to US emergency depart-ments, 1993-1999. Pediatrics, 110, pe40.
Spirito, A., Williams, C. A., Stark, L. J., & Hart, K.J.(1988). Tbe Hopelessness Scale for Cbildren: Psycho-metric properties witb normal and emotionally dis-turbed adolescents. Jouma/ of Abnormal ChildPsychology, 16, 445-458.
Spooren, D., Van Heeringen, C , &Jannes, C. (1998).Strategies to increase compliance witb out-patient af-tercare among patients referred to a psycbiatric emer-gency department: A multi-centre controlledintervention study. Psychological Medicine, 28,949-956.
Stanley, B., Brown, G., Brent, D., Wells, K., Poling, K.,Curry, J., et al. (2009). Cognitive-bebavioral tberapyfor suicide prevention (CBT-SP): Treatment model,feasibility, and acceptability. Jouma/ ofthe AmericanAcademy of Child and Adolescent Psychiatry, 48,1005-1013.
Stewart, C , Spicer, M., & Babl, F. E. (2006). Caring foradolescents witb mental health problems: Cballengesin tbe emergency department. Jouma/ of Paediatrics andChild Health, 42, 726-730.
Trafford, A. (2000, June 27). Boarder kids, on tbe edge.Washington Post, pp. 5—6.
Elizabeth A. Wharff, PhD, MSW, is director, KatherineM. Ginnis, MSW, MPH, is associate director, andAbigail Ross, MSW, MPH, is a research social worker,Emergency Psychiatry Service, Department of Psychiatry,Boston Children's Hospital. Funding for this pilot study wasprovided by the George Harrington Trust. The authors aregrateful to the adolescents and their families who participated inthe study and to the ER social workers who piloted family-based crisis intervention (FBCI) with them, including ArielBotta, Elizabeth Colton Notine, Christina Feith, Lara Kay,
Mary Kate Little, and Katie Naftzger. The authors alsothank David DeMaso, chief of psychiatry, and the ERnursing and physician staff at Boston Children's Hospital fortheir support during the FBCI study period. Address corre-spondence to Elizabeth A. Wharff, Department of Psychiatry,Boston Children's Hospital, Fegan 8, 300 Longwood Avenue,Boston, MA 02115; e-mail: elizabeth.wharf@childrens.harvard.edu.
Original manuscript received Aprii 12, 2010Final revision received August 27, 2010Accepted September 1, 2010
PRACTICE UPDATES
Are you involved in a program, researchproject, practice innovation, or other
effort that may interest readers? Send yourarticle (six double-spaced pages or fewer) asa Word document through the online portalat http://swj.msubmit.net (initial, one-timeregistration is required).
WHARFF, GINNIS, AND ROSS / Family-based Crisis Intervention with Suicidal Adolescents in the FR 143
Copyright of Social Work is the property of National Association of Social Workers and its content may not be
copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.
