project connect: linking youth with mental health concerns to community providers gail a. wasserman,...
TRANSCRIPT
Project Connect: Linking youth with mental health concerns to community providers
Gail A. Wasserman, PhD
Center for the Promotion of Mental Health in Juvenile JusticeColumbia University, Division of Child Psychiatry
www.promotementalhealth.org
Connecting Youth to Success: Doing Juveniles Justice in Minnesota
June 19, 2008
2
Strong Interest in ManagingSuicide Risk
in Juvenile Justice
Suicide risk in juvenile justice youth greater than for others
• History of aggressive or antisocial behavior• Access to weapons• Co-occurring mood and substance use
disorders• Increased school difficulties• Family Issues:
– Youth’s not living with parents and/or family discord
– Family history of mental health/substance use problems
• Poor problem solving skills
• Of all Utah youth suicides (<18 yrs), 80% had been in contact with the juvenile justice system in the 12m before death (Gray et al, 2002)
4
Pre-existing mental illness a strong predictor of suicide
• Over 90% of adolescents who commit suicide suffered from an associated psychiatric disorder at the time of death.
• In 63% of completed suicides, psychiatric symptoms developed more than a year prior to death.
• In only 4% of cases, psychiatric symptoms developed 3 months immediately prior to suicide.
• This means that mental health status is an important marker of suicide risk.
5
Current identification and management procedures for suicide risk in JJ insufficient• Recommendations (like those of
NCCHC, OJJDP) and procedures do not apply to most juvenile justice youths– Most juveniles with justice contact are not
confined, but managed in their communities
– Nationwide, only 16% of cases petitioned (9% of those arrested) result in secure placement, with the remainder returned to their communities
• Need for screening at juvenile justice entry points
6
Implementing standards lowers suicide risk for incarcerated youth
• OJJDP’s 2000 Juvenile Residential Facility Census (n=3690 facilities)
• Facilities with universal screening within first 24 hrs of intake reported significantly fewer serious suicide attempts (OR=.45, p<.01), – regardless of facility size or whether youth come
from another facility within the system
• Detention centers, privately owned facilities and those without on-site MH care reported significantly more serious attempts
CA Gallagher & A Dobrin (2005), JAACAP, 44(5):485-493
7
• Juvenile probations is an excellent public health setting in which to screen for suicide risk
• But… justice settings need to develop assessment strategies that differ from those used in clinical settings– High rates of disorder– Many youths in crisis– For many youths, the first opportunity for mental
health scrutiny• Minority youths, those from families with fewer
resources, less likely to access services in their communities
Implementing NYS’ Adolescent Suicide Prevention Plan
• SAMHSA-funded Project Connect– 4 NYS counties (Albany, Broome, Onondaga,
Orange)– Planning meetings between county probation
and mental health to develop linkage protocols
• Project Connect’s 3 phases– Baseline record review– 2 day training– Implementation and Follow-up
9
Project Connect relies on a public health
approach to mental health assessment • Proactive case identification
1. Systematic screening2. Consistent, sound and accurate approach to instruments (DISC, DPS)
• Clear protocols for how to move from assessments to treatment
3. Decision trees for referrals4. Local Resource Guides5. Two-day didactic training
• Evaluation of impact of new procedures on practices
10
Baseline data
• Charts for 584 new delinquent intakes, with 41 POs, reviewed – 70% Intake/diversion– 28% Probation supervision
• Average youth was white (46%) or African-American (42%) male (73%) 14 year old
• Two-thirds were first offenders• One third committed persons or weapons
offenses
584 (594) JD intakesseen by 41 (49) POs in4 counties
91 (15%) 96 (16%) receiving MH services at case opening
138 (24%) 134 (23%)
newly justice
identified
58 (10 %) 78 (13 %) additional justice referral or supportive action taken
355 (61%)364 (61%)not identified
33 (6 %) 18 (3%) no justice referral or additionalsupportive action taken
BL/Intervention
12
In Intervention, 96 of 594 intakes agreed to systematic MH screening via V-DISCDisorders that were to prompt referral
– Class I (Emergency)• Recent suicide attempt• Ideation + plan or subthreshold Mood or SU disorder
– Class II (Crisis)• Recent ideation w/o plan or subthreshold disorder
and• Can agree on safety plan
– Class III (Non-Crisis)• Any Substance Use Disorder• Any Mood Disorder (MDD, Mania, Dysth)• PTSD• Panic Disorder• Any of the above, at “Serious” subthreshold levels
13
County MHPS Rating (US HRSA, SAMHSA)1. The area is a rational area for the delivery of MH services
2. One of the following conditions prevails:– The area has either
• Population-to-core-MH-professional 6,000:1 and a population-to-psychiatrist 20,000:1
• Population-to-core-MH-professional 9,000:1• Population-to-psychiatrist 30,000:1
– The area has unusually high needs for mental health services, and has
• Population-to-core-MH-professional 4,500:1 and a population-to-psychiatrist 15,000:1
• Population-to-core-MH-professional 6,000:1• Population-to-psychiatrist 20,000:1
3. MH professionals in contiguous areas are over-utilized, excessively distant or inaccessible to residents of the area under consideration
14
Measuring PO Mental Health Competence• Most measures administered directly before
and after training – Mental Health Knowledge (33 items)– Self-efficacy
• How well POs believed they could identify youths’ mental health concerns and link them to service providers
• 25 5-point Likert items, among POs = .85• Adaptation of the Vanderbilt Mental Health Services
Efficacy Questionnaire (Bickman, Heflinger, Northrup, Sonnischen, & Schilling, 2004)
– Perceived competency (12 Likert-scale items)• “How well do you think you can identify a youth’s
anxiety disorder?”
15
Who gets identified in Baseline?
• Logistic regression considering youth and PO characteristics, PO MH competency, and county MHPS Rating
16
Independent contributions to BL MH identification
Measure OR Significance
Control variables < .001
Receiving Rx at opening
3.15 < .01
Youth characteristics < .08
Repeat offender 2.36 < .01
PO characteristics < .05
PO MH Competency < .001
Pre PC Knowledge 1.06 < .01
County MHPS Rating < .001
Partial vs. No Shortage
14.1 < .001
17
In Baseline, characteristics of youths, POs, and the mental health system predict identification
• Repeat offenders were almost 2.5 times as likely to be newly identified
• For every item increase in a PO’s knowledge score, the youth on that PO’s caseload were 6% more likely to be newly identified
• JDs in counties designated as not having a shortage of mental health professionals, compared to those in a shortage county, were more than 14 times as likely to be newly identified
18
BL MH identification relates to a range of factors (42.3% of variance explained)
Youth char's (3.4% )
PO dem/ occ chars(3.8% )PO MH competency(10.4% )County MHPS Rating(5.6% )Control Vars (18.9% )
Unexpl Var (57.7% )
19
Detention referral study: Lopez-Williams, 2006
• Gatekeeper staff more likely to refer repeat offenders, and females
• Referred and non-referred not different in actual symptom levels
• Concluded that without systematic screening, gatekeepers rely on “proxy” measures of mental health need
• Leads to inefficient decision-making
20
Factors often linked to mental health concerns
• Gender– Girls, even in justice samples, have higher
rates of mood (e.g., depression) and anxiety disorders (e.g., trauma exposure), suicide attempt risk
• Violent behavior– Those with violence histories, including current
offense, at increased risk of suicide attempt
• Substance use– Substance abusers at increased risk of suicide
attempt
21
Intervention data
• Charts for 594 new delinquent intakes, with 49 POs, were reviewed & compared to BL
• No differences across conditions in most demographic or offense characteristics– Baseline youths more likely to be repeat
offenders (34% v 28%, p < .05)– Baseline youths were less likely to come
from a county with a Mental Health Professional Shortage (31% v 40%, p < .01)
Some linkage practices improve after training
Baseline (n=583)
Intervention(n=594)
MH/SU in supervision plan
185 (31.7%) 199 (33.5%)
MH/SU services court-ordered
100 (17.1%) 79 (13.3%)
PO referred for non MH/SU services***
173 (29.7%) 109 (18.3%)
PO implemented MH referral*
68 (11.7%) 100 (16.8%)
PO confirmed initiation *
96 (16.5%) 120 (20.2%)
MH/SU services accessed***
90 (55.6%) 125 (79.1%)
Linkage practices more strongly affect service access in Intervention; among those with new referrals…
MH/SU Services Accessed?
Baseline Intervention
MH/SU in supervision plan
83 (56%) 115 (79%)
MH/SU services court-ordered
44 (63%) 48 (77%)
PO implemented MH referral
40 (68%) 86 (86%) **
** 2 (1) = 7.82, p < .005
24
MH access increased in Intervention, even adjusting for the role of PO implementing
Measure OR Significance
Condition (BL/Intervention)
2.53 < .005
Days chart open 1.004 < .02
Implement referral
1.82 < .05
Analyses restricted to youths not in Rx at case opening (n=974)
25
After training, characteristics of youths, POs, and the mental health system predict service access
Among those not in treatment at case opening
• Youth in the Intervention condition were
2.5 times as likely to access services• For every 10 days a youth’s chart was
open for review, s/he was 4% more likely to access services
• Youths whose PO implemented the referral were almost twice as likely to access services
26
Of those screened, 36% endorsed 1+ disorders, or substantial suicide risk
96 Youth screened(DISC records)
PC “Disorders”
N=20 (20.8%) 15 (16%) w Dx
5 (5%) w Suicidal B Other DisordersN=15 (16%)
No DisorderN=61 (63.5%)
5 (33.3%) already in Rx
2 (13.3%) newly referred
7/15 (47%) to MH
12 (20%) already in Rx15 (16%) newly referred27/61 (44%) to MH
7 (35%) already in Rx11 (55%) newly referred18/20 (90%) to MH
27
Targeting referrals to high risk group
Baselinen=583
Post training, w systematic screening (V-DISC) n=93
PC Dx or suicide risk
(n=19)
Other Dx(n=15)
No Dx(n=59)
Already in Rx
84 (14%) 7 (37%) 5 (33%) 12 (20%)
Newly referred
123 (21%)
11 (58%) 2 (13%) 15 (25%)
Σ “in MH system”
207 (36%)
18 (95%) 7 (47%) 27 (46%)
BL refs < Post training z=4.18, p < .00001 (36% vs 56%)BL refs < PC Dx refs z=11.24, p < .00001 (36% vs 95%)
28
BL MH/SU reasons were listed for only 65.5% (150/229) of identified youth
Dx reasons, 123
Non-Dx reason only,
27
No MH/ SU reason:
identified, 79
No reason: not
identified, 355
With V-DISC, identification of any mental health or suicide concerns 2x high
0
5
10
15
20
25
30
35
40
45
DBD SUD Int Any Fam Eval Suic
PC No V-DISC:N=584
PC DISC: N=89
NYS V-DISC:N=230
30
Without systematic screening
• About half of expected are identified (49%)– 54% of those with substance problems– 50% of those with disruptive problems– 20% of those with internalizing problems– 8% of those with suicide risk
• Other research shows that families and other gatekeepers identify externalizing problems more readily than internalizing problems– For “hidden” concerns such as suicide risk and
depression, need direct youth input to identify
31
Conclusion:
• To increase accurate identification and MH access, agencies need – Universal Screening– Systematic decision rules – Standard practices
Materials used in or developed for Project Connect
33
1., 2. Systematic screening via sound instruments
• Project Connect relies on the Voice-DISC to aid in proactive identification of youth suicide and mental health risk
34
The DISC-IV• Comprehensive: up to 30
DSM-IV diagnoses and multiple timeframes (i.e., past year, past month, whole life)
• Printout of provisional diagnoses available immediately
• Most widely tested child psychiatric assessment instrument
• Administration time is approximately 1 hour
35
Advantages of the Voice DISC for Juvenile Justice Settings
• Increased disclosure of suicide risk and substance use
• Requires little or no reading skill -- Self administered format: youth hears questions over
headphones and keys in responses on computer
• Minimal staff support requirements• Rates identified comparable to
systematic interviewer-based procedures
• Allows for ready aggregation of prevalence data across individuals
36
3. Decision trees to systematize referral
• Project Connect relies on systematic decision trees to guide POs in connecting youths to mental health services, safely and efficiently
37
Referral Urgency Classification
• Class I: Emergency Room – Life-Threatening Emergency:
• Youth requires immediate intervention to prevent death or serious harm to self or others
• Youth requires immediate evaluation within a safe environment
– Non-Life Threatening Emergency• Youth requires rapid intervention to prevent acute
deterioration which might compromise the youth’s safety• Face to face evaluation should take place within 6 hours of
identified need
• Class II– Crisis
• Class III– Non-critical
38
Class I: Immediate Triage/Emergency Clinical Care
Y e s N o t physically/em o tio nally
capable
N o t w illing toaccept E R
services
N o t availab le
Leave a m essage
R ep o rt to C P S
C all 9 1 1 to arrange transpo rta tio n to a w alk-in c lin ic o r a9 3 9 em ergency ro o m fo r psych em ergencies. If neither o fthese are availab le , arrange transpo rta tio n to the nearest
availab le E R .
W alk-in c lin ic , 9 3 9 psych em ergency ro o m o r E R availab le?
N oY e s
E R Y o uth to b e held in p o lic e c us to d yuntil an ER b ec o m es availab le
Is the pa re nt…• present o r can be reached to co m e in A SA P ?• physically/em o tio nally capable o f tak ing child to E R ?• w illing to accept em ergency/m o bile m enta l health serv ices?
If th e y o u th h a s e ith e r ...• Su ic idal ideatio n (in the past 4 w eeks) and a p lan (in the past 4 w eeks)• Su ic idal ideatio n (in the past 4 w eeks) and a su ic idal a ttem pt ( in the past 4 w eeks)• Su ic idal ideatio n (in the past 4 w eeks) and a su ic idal a ttem pt (prio r to past 4 w eeks) and a po sitive o r severe su b-thresho ld d iagno sis o f M o o d o r SU D , regardless o fim pairm ent
• Su ic idal a ttem pt in the past 4 w eeks
Is the yo u th in trea tm ent?
Y es N oC a ll M H P ro v id e ran d in fo rm th em o fth e cu rren t s itu a tio n
C all w ith in 5 hrs . o f w hen the yo u thleaves yo u r o ffice to co nfirm thatyo u th received em ergency services.If parent transpo rted the yo u th , andthey d id no t sho w u p, repo rt to C P S
39
Class II: Crisis – Clinical Evaluation within 24
Hours
Y e s N o t physically/em o tio nally
capable
N o t w illing toco m m it
N o t available
Leave a m essage
R e p o rt to C P S
N oY e s
E R
Is the pa re nt…• present o r can be reached to co m e in ASA P ?• physically/em o tio nally capable and w illing to co m m it to a safety p lan?
•If the yo u th has su icidal ideatio n (in the past fo u r w eeks)w ith no plan and he/she can agree o n a safety plan
Is the yo u th in treatm ent?
Y es N oC a ll M H P ro v id e ra n d in fo rm th e m o fth e c u rre n t s itu a tio n
• Safety planestablished &im plem ented
•Set u p a p sychevalu atio n/crisisreferral w ith in2 4 hrs
If yo u th no t w illing toagree o n a safety p lan
C all 9 1 1 to arrange transpo rta tio n to a w alk-in c lin ico r a 9 3 9 em ergency ro o m fo r psych em ergencies.If neither o f these are available , arrange transpo rta tio n tothe nearest available E R .
W alk-in c lin ic , 9 3 9 psych em ergency ro o m o r E R available?
C all pro vider 1 ho u r after the v isit w asto o ccu r to co nfirm that yo u th receivedem ergency services. If parent w as totake the yo u th and the yo u th d id no t receive the service repo rt to C P S
Y o u th to be held in po lice cu sto dyu ntil an E R beco m es available
40
Class III: Non-crisis – Clinical Confirmation & Referral for MH
Services
N o t w illing toaccept
N o t availab le
Leave a m essage
Is the pa re nt…• present o r can be reached to co m e in A SA P ?• w ill ing to accept the referra l?
•If the yo u th 's V o ice D ISC sho w s a po sitive o r severe su b-thresho ld d iagno sis fo r m ajo r depressio n/dysthym ia , a lco ho l,dru g, and o ther su bstance abu se, m ania / hypo m ania ,po st-trau m atic s tress d iso rder, and/o r panic d iso rder
Is the yo u th in trea tm ent?
Y es N o
F o llo w u p w ith yo u th 's pro viderto see if yo u th is a ttendingm ental health trea tm ent and sharew ith them the resu lts o f yo u th 'sV o ice D ISC
M ake parents aw areo f the benefits o fm ental healthtreatm ent
If the yo u th no t a ttending,m ake yo u th and parentsaw are o f the benefits o fm ental health trea tm ent
K eep ra is ing theissu e
M ake referra lw ith yo u th fo rnext availab le M Happo intm entC o nsider w ith yo u th
and parents m akinga referra l to ano therpro vider
Y es
M ake referra lw ith yo u th andparents fo r nextavailable M Happo intm ent
F o llo w u p w ith parent to seeif yo u th info rm ed them o fthe referra l
C all the p ro vid erw ithin 72 hrs . o fthe ap p o intm ent toc o nfirm that yo uthfo llo w ed thro ughw ith the referral
C all pro vider w ith in 7 2 hrs . o f theappo intm ent to co nfirm that yo u thfo llo w ed thro u gh w ith referra l
41
4. Developing local Resource Guides
• Project Connect relies on Resource Guides to describe county services– Contact information– Staffing– Insurance accepted– Hours of operation– Disorders treated– Transportation access
42
5. Project Connect Training
• Enhance probation officers’ knowledge of: – Suicidal behavior and correlated risks– Specific mental health disorders– Evidence-based treatments for these disorders– Community mental health resources for youth
• Coach probation officers on how use – Effective screening techniques for identifying
youth– Effective communication techniques for
referring youth with mental health conditions
• Assist probation officers to implement new skills and knowledge into practice
43
88 PO’s attended trainings
•Fall-winter 2006/2007•The average PO was 41 years old•Approximately 60% were female•An average of 10.2 years in
probation•39% had prior experience working in
a MH setting•74% had no prior MH inservice
training
All PO’s learned, but those without prior MH experience learned significantly more
2022242628303234
Befo
re Tra
ining
Afte
r Tra
ining
Test
Sco
re Prior MHEmploymentNo Prior MHEmployment
Overall, a 17.2% increase (p < .001)
Training increased overall perceived MH Competency significantly (4.3%)
• How well do you think you can……– Identify an anxiety disorder– Explain to family the need for MH services– Act on a mental health problem
• 11 of 12 items increased, 6 significantly so– For example, there was a 10.8% increase in
POs’ perceived ability to identify an anxiety disorder
• 90% were either “Satisfied” or “Very Satisfied” with the training overall
47
Similar results in AL probations pilot
• 40 probation staff in Jefferson County, AL (12 hour, 2 day training)
• Training increased POs’ mental health knowledge
• Training improved PO Attitudes about their MH competency
• Favorably evaluated MH curriculum• Felt that training was likely to
positively impact AL POs’ relationship with youths
48
AL training alters recommendations, saliency and satisfaction (N=866 youths)
0 20 40 60 80
PO recs MH Services
Court recs MH Services
MH Info V Imp inDisposition
PO V Satis w MH info
PO V Satis w SU info
Post-Training (1 yr)
Baseline (3 mo)
49
Conclusions
• There is a high rate of mental health concerns at each level of juvenile justice processing
• Problems are measurable and addressable at intake
• Targeted gatekeeper training increases probation practices that promote access to mental health services
• A public health model can be applied to mental health issues across diverse juvenile justice settings
• Failure to do so consistently results in under-identification of the burden of mental health need and suicide risk