crime, violence, and managing client and public safety michael l. dennis, ph.d., chestnut health...

Crime, Violence, and Managing Client and Public Safety

Michael L. Dennis, Ph.D., Chestnut Health Systems, Bloomington, IL

Presentation at “NEW DIRECTIONS TO HEALTHIER COMMUNITIES & METH SUMMIT”, September 28-30, 2005, Savannah Marriott Riverfront, Savannah, GA. Sponsored by the Georgia Council on Substance Abuse and the Georgia Department of Juvenile Justice, Office of Behavioral

Health Services. The content of this presentations are based on treatment & research funded by the Center for Substance Abuse Treatment (CSAT), Substance Abuse and Mental Health Services Administration (SAMHSA) under contract 270-2003-00006 using data provided by the CYT and

AMT grantees: (TI11320, TI11324, TI11317, TI11321, TI11323, TI11874, TI11424, TI11894, TI11871, TI11433, TI11423, TI11432, TI11422, TI11892, TI11888). The meta analysis of juvenile

offender intervention data was adapted from an earlier presentation by Mark Lipsey with his permission. The opinions are those of the author and do not reflect official positions of the

consortium or government. Available on line at www.chestnut.org/LI/Posters or by contacting Joan Unsicker at 720 West Chestnut, Bloomington, IL 61701, phone: (309) 827-6026, fax: (309)

829-4661, e-Mail: [email protected]

• To summarize the need for measuring substance use, crime and violence and its correlates

• To examine the utility of the GAIN’s Substance Problem for assessing the risk of relapse and recidivism

• To summarize the results of meta analyses of effective programs for juvenile offenders by Lipsey and colleagues

Goals of this Presentation

Adolescent Present with a Broad Range of Past Year Illegal Activity and Violence

Source: Adolescent Treatment Model (ATM) data

7478

82

69 7168

86

65

8580 81 81

939395

0

10

20

30

40

50

60

70

80

90

100

OP/IOP (n=560) LTR (n=390) STR (n=594)

Any illegal activity Property crimes Interpersonal crimes

Drug related crimes Acts of physical violence

Substance Abuse Treatment (particularly residential) Reduces Illegal Activity

40

50

60

Intake 3 6 9 12

Months from Intake

STR\t,s,ts

LTR\t,ts

OP\s

\a Source: Adolescent Treatment Model (ATM) data; Levels of care coded as Long Term Residential (LTR, n=390), Short Term Residential (STR, n=594), Outpatient/Intensive and Outpatient (OP/IOP, n=560);. T scores are normalized on the ATM outpatient intake mean and standard deviation. Significance (p<.05) marked as \t for time effect, \s for site effect, and \ts for time x site effect.

Background

• Substance use and crime are inter-related.• Self-report method is valid and useful for predicting

treatment placement, relapse and recidivism. • Typically, substance use measures have been used to

predict placement and relapse, while criminological measures have been used to predict recidivism.

• This is one of the first adolescent studies to look at the ability of substance use and criminological measures combined to predict placement, relapse, and recidivism in the same population or study.

12

3

4

56

78

9

10

a

b

c

d

Location of CYT/ATM Treatment Sites

Sponsored by: Center for Substance Abuse Treatment (CSAT), Substance Abuse and Mental Health Services Administration (SAMHSA), U.S. Department of Health and Human Services

Adolescent Treatment Model (ATM) Sites:1. Chestnut Health Systems, Bloomington, IL2. Dynamite Youth, New York, NY3. Four Corners Regional Adolescent Center/

University of Oklahoma Shiprock, NM4. Friends Institute/Epoch Counseling, Catonsville, MD5. Mountain Manor, Baltimore, MD6. Public Health Institute/Thunder Road, Oakland, CA7. Rand Corp./Phoenix Academy/Group Homes, Santa

Monica, CA8. University. of Arizona/IMPACT, Phoenix, AZ9. University of Arizona/La Cañada/7-Challenges/Drug

Court, Tucson, Az10. University of Miami/MDFT/The Village, Miami, FL

Cannabis Youth Treatment (CYT) Sites:a. Chestnut Health Systems, Madison County, ILb. Children’s Hospital of Phil., Philadelphia, PAc. Operation PAR, St. Petersburg, FLd. Univ. of Conn. Health Center, Farmington, CT

Evaluation • Target Population: Adolescents entering substance abuse

treatment. • Inclusion Criteria: 12 to 22 year old adolescents who present

for substance abuse treatment and received at least 2 outpatient sessions or 1 week of residential treatment.

• Data Sources: Self-report measures of diagnosis and outcome collected with the Global Appraisal of Individual Needs (GAIN).

• Participants: 2007 adolescents recruited from 14 sites around the U.S. and interviewed at 3, 6, 9 and 12 months later (98% completed 1 plus interview; 92% completed 12 month interview).

Intensity of Juvenile Justice System Involvement

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Severity

Detention 14+ days (n=433)

Probation/parole and urine monitoring 14+ days (n=472)

Other current arrest or JJ status (n=303)

Other detention, parole, or probation (n=374)

Past arrest or JJ status (n=170)

Past year illegal activity (n=298)

Source: CYT & ATM Data

LowHiRow %

Intensity by Level of Care

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Short Term Residential

Long Term Residential

Outpatient/IOP

Step Down OP

Total

Detention 14+ days (n=433) Probation/ Parole and urine monitoring 14+ days (n=472)Other detention, parole, or probation (n=374) Other current arrest or JJ status (n=303)Past arrest or JJ status (n=170) Past year illegal activity (n=298)


Row %

76%

24%

14%

52%

31%

3%

52%

17%

14%

8%

0% 20% 40% 60% 80% 100%

Male

Female

Under 15

15 to 16

17 to 18

Over 18

White

African American

Hispanic

Native American

Demographic Characteristics


Row %

Demographics by Intensity

Detention 14+ days (n=433) Probation/parole and urine monitoring 14+ days (n=472)Other detention, parole, or probation (n=374) Other current arrest or JJ status (n=303)Past arrest or JJ status (n=170) Past year illegal activity (n=298)


0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Female Caucasian AfricanAmerican

Hispanic NativeAmerican

Other

Females and Caucasians more likely in lower

intensityMinorities More

Likely to be in higher intensity

Col %

Demographics by Intensity (continued)



0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Age 11-15 Years Age 15-17 Years Age 18+ Years Single Parent

High Severity More likely to be 15-17 years olds and

from Single Parent Families

Youngest least likely to be in

the system

Col %

Substance Use Characteristics

87%

12%

32%

44%

20%

43%

63%

70%

0% 20% 40% 60% 80% 100%

First Use Under Age 15

Under age 10

5+ years of use

Prior Treatment

Multiple Prior Tx

Entering Residential

Dependence in Past Year

Weekly Use in Past 90 Days


Row %

Substance Use Disorder Diagnosis by Intensity


Source: CYT & ATM Data; a\ Self report for past year

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Any Substance Disorder\a Dependence\a Abuse\a

Current Intensity Inversely related to Substance Use Severity

Past Involvement a Mix of Severity

Col %

External Diagnoses by Intensity



0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Any External Conduct Disorder ADHD

Col %

Multiple Co-Occurring Disorders are Common in all levels of JJ involvement

Internal Diagnoses/Problems by Intensity


Source: CYT & ATM Data \b n=1838 because some sites did not ask trauma questions

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Any Internal\b Major Depression Suicide Ideation GeneralizedAnxiety

Trauma\b

Curvilinear Relationship between Intensity and Internal Distress

Col %

Pattern of Co-occurring Disorders by Intensity



0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

None Internal Only External Only Both

Most Internal Distress is Multi-morbid with External (and Substance Use) Disorders

Col %

Legal Characteristics

88%

59%

79%

43%

18%

10%

55%

22%

0% 20% 40% 60% 80% 100%

Other illegal activity in Pst Yr

Violent crimes in Pst Yr

History of Arrest

Past 90 day Arrest

Awaiting a trail

Coming from Detention

On Probation/Parole

Oustanding fines/restitution


Row %

Crime/Other Problems by Intensity



0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

AnyVictimization

High levels ofVictimization

Any crime High Crime/Violence

Homeless orRunaway

High HealthProblems

Focus of JJ Detention

Often Both Perpetrator and Victim

Stress Can lead to higher rates of health

problems

Also higher incidents of

Running away

Col %

Substance Problem Scale (SPS)

The SPS (alpha=.88) is a count of 16 past year symptoms basedon• three common screening questions (S9c-e), • two questions related to substance “induced” psychological

or health disorders (S9f-g),• lay versions of the DSM-IV/ICD-9 criteria for substance

abuse (S9h-m), • Lay versions of the DSM-IV/ICD-9 criteria for substance

dependence (S9n-u).The latter also forms the Substance Dependence Subscale (SDS;alpha=.82). The SPS symptom count severity is triaged as Low(0 past year symptoms), Moderate (1 to 9 symptoms) or High (10to 16 symptoms) severity.

Crime and Violence Scale (CVS)

• The CVS (alpha=.90) is a count of 29 past year symptoms from two subscales:

– The General Conflict Tactic Subscale (GCTS; alpha = .88) - based on the National Family Violence Survey and work by Murray Strauss.

– The General Crime Subscale (GCS; alpha = .86) - based on the National Household Survey on Drug Abuse lay terms for the Uniform Crime Report categories.

• CVS symptom count severity is triaged as:– Low (0 to 2 past year symptoms), – Moderate (3 to 6 symptoms), or– High (7 to 29 symptoms) severity.

Distribution of SPS by CVS Risk Groups

Low

Mod.

High

LowMod

.High0%

20%

40%

Per

cent

of

Tot

al

(n=

2007

)

Substance Problem

Scale

Crime and Violence

Scale


Moderate to high severity substance use

and crime/ violence

problems are common

Validation of the SPS and CVS subgroups

• Endorsement of each items and subscales increased with the shift from low to moderate to high.

• For the Substance Problem Scale (SPS) severity subgroups:– Shifting from low to moderate was associated with sharp increases in the

screener questions (c-e), continued use in spite of getting into fights or legal problems (m), and time spent on getting/using/recovering from substance use (s).

– Shifting from moderate to high was associated with more of the above and greater increases in the substance dependence and substance induced disorder symptoms.

• For Crime/Violence Scale (CVS) severity subgroups:– Shifting from low to moderate was associated with increased oral violence,

property crime, and drug related crime.– Shifting from moderate to high was associated with even more of these

things, as well as more physical violence and interpersonal (aka violent) crimes.

• Next we looked at their predictive validity separately and together

Probability of Being Placed in Residential Treatment at Intake

Low

Mod.

High

LowMod

.High0%

20%

40%

60%

80%

100%


Substance Problem

Scale

Crime and Violence

Scale

Pro

babi

lity

of

Res

iden

tial

Pla

cem

ent

Substance Problem Severity predicted

residential placement

Crime/ Violence did not predict residential placement

Probability of Using at Month 12

Low

Mod.

High

LowMod

.High0%

20%

40%

60%

80%

100%

Pro

babi

lity

of

Usi

ng a

t Mon

th 1

2


Substance Problem

Scale

Crime and Violence

Scale


RelapseHowever knowing both was the best predictor

(Intake) Crime/

Violence did not predict

relapse

Subsequent Violence, Victimization, and Illegal Activity (by self and others) is one of the Major

Environmental Predictors of Relapse

RecoveryEnvironment

Risk

SocialRisk

FamilyConflict

FamilyCohesion

SocialSupport

SubstanceUse

Substance-RelatedProblems

Baseline

Baseline

Baseline Baseline

.32.18

-.13

.21

-.08

.32

.19

.22

.32

.22

.17

.11

.43

.77

.82

.74 .58

-.54

-.09

.19

Source: Godley et al (2005)

Model Fit CFI=.97 to .99 RMSEA=.04 to .06

Recall that the effects of treatment are mediated by the extent to which they lead to actual changes in the recovery environment or peer group

Includes days of aggression towards others and victimization by

others

Includes substance use, fighting, and illegal

activity by peers

Crime/Violence and Substance Problems Interact to Predict Recidivism

Low

Mod.

High

LowMod

.High0%

20%

40%

60%

80%

100%


12 m

onth

rec

idiv

ism

Crime/ Violence predicted recidivism


recidivismKnowing both was the

best predictor

Substance Problem

Scale

Crime and Violence

Scale

Crime/Violence and Substance Problems Interact to Predict Violent Crime or Arrest

Low

Mod.

High

LowMod

.High


12 m

onth

rec

idiv

ism

T

o vi

olen

t cri

me

or a

rres

t

Substance Problem

Scale

Crime and Violence

Scale

0%

20%

40%

60%

80%

100%

Crime/ Violence predicted

violent recidivism

(Intake) Substance Problem Severity did

not predict violent recidivism

Knowing both was the best predictor

Discussion of SPS and CVS• The GAIN’s SPS and CVS scales appears to be face valid,

internally consistent and to have good construct validity.• While placement in residential treatment focuses on substance use

severity, CVS helps to predict relapse. This suggests the need to consider crime and violence more closely in placement decisions.

• Conversely, SPS helps to predict recidivism. This suggests the potential benefits of screening for substance use problems in juvenile justice settings.

• The next step is to combine these variables with other factors in a multivariate model.

• We also need to replicate these findings, preferably with a sample not presenting for treatment and with urine and record checks.

The Effectiveness of Programs for Juvenile Offenders

N ofOffender Sample Studies

Preadjudication (prevention) 178Probation 216 Institutionalized 90Aftercare 25

Total 509

Source: Adapted from Lipsey, 1997, 2005

Most Programs are actually a mix of components

Average of 5.6 components distinguishable in program descriptions from research reports

Intensive supervisionPrison visitRestitutionCommunity serviceWilderness/Boot campTutoringIndividual counselingGroup counselingFamily counselingParent counselingRecreation/sportsInterpersonal skills

Anger managementMentoringCognitive behavioralBehavior modificationEmployment trainingVocational counselingLife skillsProvider trainingCaseworkDrug/alcohol therapyMultimodal/individualMediation


Most programs have small effectsbut those effects are not negligible

• The median effect size (.09) represents a reduction of the recidivism rate from .50 to .46

• Above that median, most of the programs reduce recidivism by 10% or more

• One-fourth of the studies show recidivism reductions of 30% or more, that is, a recidivism rate of .35 or less for the treatment group compared to .50 for the control group

• The “nothing works” claim that rehabilitative programs for juvenile offenders are ineffective is false


Major Predictors of Bigger Effects

1. Chose a strong intervention protocol based on prior evidence

2. Used quality assurance to ensure protocol adherence and project implementation

3. Used proactive case supervision of individual

4. Used triage to focus on the highest severity subgroup

Impact of the numbers of Favorable features on Recidivism (509 JJ studies)


Usual Practice has little

or no effect

Some Programs Have Negative or No Effects on recidivism

• “Scared Straight” and similar shock incarceration program

• Boot camps mixed – had bad to no effect• Routine practice – had no or little (d=.07 or 6% reduction

in recidivism)• Similar effects for minority and white (not enough data

to comment on males vs. females)• The common belief that treating anti-social juveniles in

groups would lead to more “iatrogenic” effects appears to be false on average (i.e., relapse, violence, recidivism for groups is no worse then individual or family therapy)


Program types with average or better effects on recidivism

AVERAGE OR BETTER BETTER/BESTPreadjudication

Drug/alcohol therapy Interpersonal skills trainingParent training Employment/job trainingTutoring Group counseling

ProbationDrug/alcohol therapy Cognitive-behavioral

therapyFamily counseling Interpersonal skills trainingMentoring Parent training

TutoringInstitutionalized

Family counseling Behavior management Cognitive-behavioral therapy Group counselingEmployment/job training Individual counseling

Interpersonal skills training


Cognitive Behavioral Therapy (CBT) Interventions that Typically do Better than Practice in Reducing

Recidivism (29% vs 40%)

• Aggression Replacement Training• Reasoning & Rehabilitation• Moral Reconation Therapy• Thinking for a Change• Interpersonal Social Problem Solving• Multisystemic Therapy• Functional Family Therapy• Multidimensional Family Therapy• Adolescent Community Reinforcement Approach• MET/CBT combinations and Other manualized CBT

NOTE: Generally little or no differences in mean effect size between these brand names

Source: Adapted from Lipsey et al 2001, Waldron et al, 2001, Dennis et al, 2004

Implementation is Essential (Reduction in Recidivism from .50 Control Group Rate)

The effect of a well implemented weak program is

as big as a strong program implemented poorly

The best is to have a strong

program implemented

well

Thus one should optimally pick the strongest intervention that one can

implement wellSource: Adapted from Lipsey, 1997, 2005

Conclusions

• Research shows that intervention programs can be very effective for reducing the recidivism of juvenile offenders, even in routine practice

• Program selection and strong implementation are critical; otherwise effects quickly slide to zero (or worse)

• What evidence we have about the effects of programs in routine practice indicates that most are not very effective– there is plenty of room for improvement

Next Steps

• Currently working on evaluating RWJF reclaiming futures diversion projects, CSAT young offender re-entry projects, drug court projects and several individual juvenile justice projects

• Doing more work on predicting risk of recidivism and how they related to substance use disorders, co-morbidity, and environmental factors

Resources and References• Copy of these slides and handouts

– http://www.chestnut.org/LI/Posters/

• References citedDennis, M. L., Godley, S. H., Diamond, G., Tims, F. M., Babor, T., Donaldson, J., Liddle, H., Titus, J. C., Kaminer, Y., Webb, C.,

Hamilton, N., & Funk, R. (2004). The Cannabis Youth Treatment (CYT) Study: Main Findings from Two Randomized Trials. Journal of Substance Abuse Treatment, 27, 197-213.

Dennis, M. L., Titus, J. C., White, M., Unsicker, J., & Hodgkins, D. (2003). Global Appraisal of Individual Needs (GAIN) Administration guide for the GAIN and related measures. (Version 5 ed.). Bloomington, IL Chestnut Health Systems. Retrieve from http//www.chestnut.org/li/gain

Dennis, M.L., & White, M.K. (2003). The effectiveness of adolescent substance abuse treatment: a brief summary of studies through 2001, (prepared for Drug Strategies adolescent treatment handbook). Bloomington, IL: Chestnut Health Systems. [On line] Available at http://www.drugstrategies.org

Dennis, M. L. and White, M. K. (2004). Predicting residential placement, relapse, and recidivism among adolescents with the GAIN. Poster presentation for SAMHSA's Center for Substance Abuse Treatment (CSAT) Adolescent Treatment Grantee Meeting; Feb 24; Baltimore, MD. 2004 Feb.

Godley, M. D., Kahn, J. H., Dennis, M. L., Godley, S. H., & Funk, R. R. (2005). The stability and impact of environmental factors on substance use and problems after adolescent outpatient treatment. Psychology of Addictive Behaviors.

Lipsey, M. W. (1997). What can you build with thousands of bricks? Musings on the cumulation of knowledge in program evaluation. New Directions for Evaluation, 76, 7-24.

Lipsey, M.W. (2005). What Works with Juvenile Offenders: Translating Research into Practice. Adolescent Treatment Issues Conference, February 28, Tampa, FL

Lipsey, M.W., Chapman, G.L., & Landenberger, N.A. (2001). Cognitive-Behavioral Programs for Offenders. The ANNALS of the American Academy of Political and Social Science, 578(1), 144-157

Waldron, H. B., Slesnick, N., Brody, J. L., Turner, C. W., & Peterson, T. R. (2001). Treatment outcomes for adolescent substance abuse at four- and seven-month assessments. Journal of Consulting and Clinical Psychology, 69(5), 802-812.

White, M. K., Funk, R., White, W., & Dennis, M. (2003). Predicting violent behavior in adolescent cannabis users The GAIN-CVI. Offender Substance Abuse Report, 3(5), 67-69.

http://www.chestnut.org/LI/Posters/

crime, violence, and managing client and public safety michael l. dennis, ph.d., chestnut health...

Documents

substance abuse treatment

opiop n

substance use

presentation slide

violence source

treatment research

gains substance problem

atm outpatient intake