Early findings from NIDA’s Clinical Trials Network “Women and Trauma” Study Denise Hien, Ph.D. Senior Research Scientist, Social Intervention Group, Columbia.

Early findings from NIDA’s Clinical Trials Network “Women and Trauma” Study

Denise Hien, Ph.D.Denise Hien, Ph.D.

Senior Research Scientist, Social Intervention Group,Columbia University School of Social Work

Executive Director, Women’s Health Project Treatment and Research Center, Addiction Institute of New York, St. Luke’s\Roosevelt Hospital Center

115th Annual American Psychological Association MeetingSan Francisco, CaliforniaAugust 18, 2007

PLEASE DO NOT CITE CONTENTS OF PRESENTATION WITHOUT PERMISSION OF THE AUTHOR

“The past isn’t dead, it isn’t even past.”

-William Faulkner

Scope of the Problem1 in 2 women in the U.S. experience some type of traumatic event (Kessler, 1995)

Approximately 33% of females under age 18 experience sexual abuse (Finkelhor, 1994; Wyatt, 1999)

Prevalence rates of PTSD in community samples have ranged from 6% to 36% (Breslau, 1991; Kilpatrick, 1987; Norris, 1992; Resnick, 1993)

Studies have documented PTSD rates among substance using populations to be between 14%-60% (Brady, 2001; Donovan,

2001; Najavits, 1997; Triffleman, 2003)

Neurobiological Changes in Response to Traumatic Stress

Limbic System -- Hippocampus and Amygdala (Affect and Memory, e.g, Ledoux, 2000; van der

Kolk, 1996)

Neurotransmitters and Peptides (Numbing and Depression, e.g., Pitman, 1991, Southwick, 1999)

Changes in Hormonal System (HPA axis) (Arousal, e.g., Yehuda, 2000)

Historical Context for the Study of Trauma & Addiction

1970 1980 1990

DSM-IIIR broadens criteria for

PTSD

Miller’s work with criminal justice

population

Women’s movement and

grassroots advocacy for

battered women

Crack/ Cocaine epidemic

PTSD studies with male veterans

with/out substance use

Fullilove’s snowball sample

Historical Context for the Study of Trauma & Addiction

1990 1995 2000 2007

Herman’s Trauma

and Recovery published

Violence declared

public health epidemic

Chilcoat and Breslau self-medication

model

Kendler et al. co-twin study suggesting causal link

between abuse and SUD

National awareness of

PTSD and addiction following 9/11/01

SAMSHAfindings

published

Manualized integrated

trauma and SUD tx

RCTs of integrated PTSD and

SUD tx

Differences between Co-morbid PTSD vs. PTSD-only behavioral treatmentsAddition of components focused on coping and cognitive restructuring related to substance use (cravings and relapse triggers)

Concurrent Model : Additional components may be

integrated and delivered concurrently

Sequential Model: Initial phase may focus on

substance abuse related symptoms in preparation for

working on trauma related symptoms later

PTSD/SUD Behavioral Treatments

ATRIUM: Addictions and Trauma Recovery Integrated Model (Miller & Guidry, 2001)

Seeking Safety (Najavits, 1998; www.seekingsafety.org)

TARGET - Trauma Affect Regulation: Guidelines for Education and Therapy (Ford; www.ptsdfreedom.org)

Transcend (Donovan et al., 2001)

CTPCD - Concurrent Treatment of PTSD and Cocaine Dependence (Back et al., 2001) SDPT/ARTS: Substance Dependence PTSD Therapy/Assisted Recovery from Trauma and Substances (Triffleman et. al, 1999)

Washington Node Residence XII

New York Node ARTC Long Island Node

Lead Node

New England Node LMG Programs

South Carolina Node Charleston Center

Florida Node The Village

Florida Node Gateway Community

Ohio Valley Node Maryhaven

NIDA Clinical Trials Network Women & Trauma Sites

CTN Long Island Node Team

Denise Hien, Lead InvestigatorEdward Nunes, Node PIGloria Miele, Training DirectorLisa Cohen, Protocol ManagerAimee Campbell, Project DirectorJennifer Lima, Node CoordinatorHuiping Jiang, Statistician David Liu, NIDA Liaison

Participating Nodes and CTPs

Node Node PI(s) Protocol PI CTP Site PI Location

Florida

Jose Szapocznik & Daniel Santisteban

Lourdes Suarez-Morales

The Village Michael Miller Miami, FL

Gateway Community

Candace Hodgkins

Jacksonville, FL

New England

Kathleen Carroll

Melissa Gordon

LMG Programs Samuel Ball Stamford, CT

New York John RotrosenMarion Schwartz

Addiction Research & Treatment Corporation

Robert Sage Brooklyn, NY

Ohio Valley Gene SomozaGreg Brigham

Maryhaven Greg BrighamColumbus, OH

South Carolina

Kathleen BradyTherese Killeen

Charleston Center Mark CowellCharleston, SC

WashingtonDennis Donovan & Betsy Wells

Betsy Wells Residence XII Karen Canida Kirkland, WA

CTN Women & Trauma: A Unique Opportunity

Chance to conduct a “practical” clinical trial on SS while maintaining a rigorous controlShorter treatment window and doseOpen-group, rolling admissions format paralleling real worldCommunity treatment providers as research cliniciansTreatment as usual while receiving research intervention

Study Aims

To assess the effectiveness of adding Seeking Safety (SS) and Women’s Health Education (WHE) groups to ongoing substance abuse treatment.

To evaluate the transportability of a 12- session group version of SS in community drug/alcohol treatment settings.

Pre-Post Control Group Design

Pre-Treatment

1 - 4 Weeks

Treatment

6 Weeks

Post Treatment Follow-up

46 Weeks

1 Week 3 Month 6 Month 12 Month

Pre-screening, Screening, Baseline, Randomization, Individual Session w/ Counselor

12 Twice Weekly Group Sessions (rolling admission)

Treatment Groups

Seeking Safety (SS)Short term, manualized treatmentCognitive BehavioralFocused on addiction and trauma

Women’s Health Education (WHE)Short term, manualized treatmentPsychoeducationalFocused on women’s health information and issues

Seeking Safety

Developed as a group treatment for PTSD/SUD womenBased on CBT models of SUDs, PTSD treatment, women’s treatment and educational researchEducates patients about PTSD and SUD’s and their interactionGoals include abstinence and decreased PTSD symptomsFocuses on enhancing coping skills, safety and self-careActive, structured treatment - therapist teaches, supports and encouragesCase management

Najavits, 2002; www.seekingsafety.org

Women’s Health Education

EmpowermentInformation is empowering

Self-careSubstance abuse and trauma interfere with ability to care for oneself

Exposure to traumatic stress can affect people on many different levels of functioning including:

emotionalbehavioral physical

There is significant overlap of PTSD and physical symptoms

In the national comorbidity survey, use of medical care services was highest in PTSD and panic disorder patients (Kessler, 1995)

Assessment Domains (1 of 2)Demographics

Substance Use Disorder DiagnosisCIDI – Composite International Diagnostic Interview

Substance UseSUI – Substance Use Inventory (Primary)

Biological – Urine/Saliva Screen (Primary)

ASI – Addiction Severity Index

HIV Risk Behaviors: RBS – Risk Behavior Survey

Health and Family Network (ASI; add-on questions)

Assessment Domains (2 of 2)

PTSD Diagnosis and SymptomsCAPS – Clinician Administered PTSD Scale (Primary)

PSS-SR – Post Traumatic Stress Symptoms – Self Report

Trauma Exposure: Lifetime Events Checklist

Psychiatric Symptoms: BSI – Brief Symptom Inventory

Service Utilization: Non-study mental health, medical, substance abuse treatment services

In-Treatment Measures(baseline, weekly thru treatment,1 week post)

PTSD Symptoms (PSS-SR)

Biologically Confirmed Substance Abstinence and Proportion of Days Used

Substance Use Inventory (SUI)

Urine Drug Screen (UDS)

Saliva Alcohol Screen (ST)

Service Utilization (NSMS)

Medication (ASL)

Adverse Events (AE)

Participant Eligibility Criteria

Inclusionfemale, 18 - 65 years oldused an illicit substance within the past six months and have a current diagnosis of illicit drug/alcohol abuse or dependencePTSD or Sub-threshold PTSDenrolled at participating community treatment program

Exclusion advanced stage medical disease (AIDS, TB)impaired mental status (MMSE: less than or equal to 21)significant risk of suicidal/homicidal intent or behaviorhistory of schizophrenia-spectrum diagnosisactive psychosis (prior 2 months)involved in PTSD-related litigationrefuses to be audio or videotaped

EnrollmentInitial Eligibility Screen

N=1,963

Eligible for Full ScreenN=1,212 (62%)

IneligibleN=751

Completed ScreenN=541 (45%)

No/Incomplete ScreenN=671

Completed BaselineEligible for Randomization

N=379 (70%)

IneligibleN=162

RandomizedN=353 (93%)

Not Randomized N=26

Sample Characteristics (N=353)

Variable percent or M (S.D.)

Age (years) 39.2 (9.3) Race/ethnicity

Hispanic or Latino 6.5 Black/African American 34.0

White 45.6 Mixed 13.3 Other 0.6

Marital status Married 33.3

Widowed/Divorced 29.3 Never Married 37.4

Education Years 12.5 (2.4)

PTSD Diagnosis and Severity at Baseline (N=353)

Variable percent or M (S.D.) PTSD Diagnosis

Full Subthreshold

80.4 19.6

CAPS Total Score 62.8 (19.4)

PSS-SR Severity 45.6 (10.8)

PSS-SR Frequency 38.7 (15.3)

Substance Use Disorders at Baseline (N=353)

Substance Use Diagnosis percent

Current Alcohol Use Disorder Diagnosis 62.0

Current Marijuana Use Disorder Diagnosis 35.4

Current Opioid Use Disorder Diagnosis 33.1

Current Cocaine Use Disorder Diagnosis 72.8

Current Stimulant Use Disorder Diagnosis 8.2

Note: not exclusive categories

Lifetime Trauma Exposure (N=353)

Event percent

Physical Assault Childhood Physical Abuse 58.7

Lifetime Physical Abuse 93.8 Sexual Assault

Childhood Sexual Abuse 70.1 Lifetime Sexual Violence 89.5

Captivity 40.3 Sudden, Violent Death 19.3 Life-threatening Illness 39.8 Transportation Accident 72.7 Natural Disaster 53.1

Data Analytic Approach for In-Treatment Outcomes

Mixed effect models were used to model the outcome measures of interest as a function of baseline levels, time, treatment, site and their interactions. A non-randomized sample of naturalistically-occurring “wait list” controls was examined with baseline as the pre-test and assessment during 1st week of treatment as the post-test.Analyses addressing impact of treatment attendance on outcomes were conducted.Between group comparisons of service utilization, medication, and adverse events were conducted.

Modeled PSS-SR Severity during Treatment

25

30

35

40

45

50

55

Basel

ine Wk1

Wk2

Wk3

Wk4

Wk5

Wk6

Post

SS WHE

Mea

n S

ever

ity

Modeled PSS-SR Frequency during Treatment

25

2729

31

3335

37

39

4143

45

SS WHE

Mea

n F

requ

ency

Modeled Abstinence Rate during Treatment

30

3540

45

5055

60

65

7075

80

Basel

ine Wk1

Wk2

Wk3

Wk4

Wk5

Wk6

Post

SS WHE

% A

bstin

ent

Modeled Days of Use during Treatment (past 7 days)

Day

s of

Use

(%

)

0

5

10

15

20

25

Basel

ine Wk1

Wk2

Wk3

Wk4

Wk5

Wk6

Post

SS WHE

Secondary Analyses: Six Week Pre- “Post” Changes for Non-Randomized Naturalistic Wait Group (N=20)

25

30

35

40

45

50

55

Basel

ine Wk1

Wk2

Wk3

Wk4

Wk5

Wk6

Post

SS WHE WAIT

Mea

n S

ever

ity

Secondary Analyses: Treatment Attendance Rates

Treatment Group*

N Mean (sd) Median

SS 170 6.3 (4.4) 7

WHE 172 5.9 (4.3) 6.5

*No significant differences between groups

Secondary Analyses: Effects of treatment attendance on post-treatment abstinence rates

Source DF 2 p-value

Baseline Abstinence 1 48.07 <.001

Race 3 2.39 n.s.

Site 5 48.95 <.001

Age 1 0.27 n.s.

Education 1 0.00 n.s.

Time 1 4.33 <.05

Treatment 1 0.04 n.s.

Treatment Attendance 1 7.47 <.001

Secondary Analyses: Effects of treatment attendance on post-treatment days of use

Source DF 2 p-value

Baseline Use 1 24.09 <.001

Race 3 1.48 n.s.

Site 5 14.20 0.01

Age 1 0.85 n.s.

Education 1 1.28 n.s.

Time 6 11.78 n.s.

Treatment 1 0.05 n.s.

Baseline Use*Time 6 24.89 <.001

Treatment Attendance 1 7.07 <.01

Service Utilization*Baseline Treatment

Service% Receiving % Receiving

Visits/Week M (SD)

Mental Health SS

WHE

73.7

80.7

83.8

78.9

1.30 (1.6)

1.50 (2.7)

Outpatient Medical SS

WHE

49.7

48.3

62.7

58.3

0.20 (0.5)

0.20 (0.4)

Inpatient Drug Treatment (# nights)

SS

WHE

20.0

14.8

3.5

4.5

0.04 (0.4)

0.05 (0.3)

ER/Hospital SS

WHE

18.9

17.7

24.7

32.7

0.04 (0.1)

0.05 (0.1)

12-Step Meeting SS

WHE

76.0

72.3

83.1

81.4

3.40 (4.1)

2.80 (3.7)

Total SS

WHE

91.4

93.2

97.9

95.5

5.00 (5.3)

4.60 (6.0)*Service use data collected in # visits unless otherwise noted; no differences between SS and WHE on any variable

Medication (past 30 days)

Baseline Post Treatment

% %

Prescribed for Psychiatric Problems SS

WHE

42.6%

40.1%

46.3%

31.3%

p-value 0.63 0.02

Adverse Events during Treatment

Adverse Event Descriptions

0

5

10

15

20

25

30

35

Wors

enin

g PTSD

Incr

ease

d Sub

stan

ce U

se

Anxie

ty/P

anic

Depre

ssiv

e Sym

ptom

sOth

er

Fre

quen

cy

SummaryDespite a number of differences in site characteristics, overall, both SS and WHE groups led by community substance abuse counselors can reduce PTSD symptoms at a statistically significant level.

Those in Seeking Safety groups improved significantly more quickly on PTSD symptoms over the course of six weeks of treatment.The number of sessions received significantly predict lowered levels of SUD symptoms, as did baseline severity of use.

Although participants in both groups experienced study related adverse events, there were no differences in rates between groups, and further, these did not impact treatment attendance or post-treatment outcomes.

Utilization of other services including outpatient drug and mental health visits, twelve step meetings, medical visits and hospitalizations did not differ across treatment groups.

Implications

With supervision, counselors in drug treatment programs can be trained to conduct and deliver Seeking Safety and Women’s Health groups safely.

Consideration of the impact of trauma-focused assessments should also be given in the treatment process; all participants received weekly self-reported assessments of their symptoms.

Being assigned to a group with other women who have been identified as having trauma or PTSD may also have a powerful effect on expectancies and outcomes, as those in Women’s Health also improved significantly on PTSD symptoms.

Longer-term groups and additional approaches may be needed to impact substance use and abuse.

THANK YOU!THANK YOU!

353 Women who participated in the study

28 Research Assistants, Assessors, and Coordinators

39 Therapists and Supervisors

11 Project Directors and Protocol Principal Investigators

27 Quality Assurance Monitors and Data Managers

NIDA U10DA13035

Appendices: Summary Data Tables

Comparison of Existing Trauma / SUD- Focused Treatment Research

Najavits, 1998

Triffleman, 2000

Brady et. al., 2001

Donovan, 2001

Hien et al.,

2004

N 27 Women

19 Mixed

39 Mixed

46 Men

107Women

Design No Control RCT No Control No Control RCT

Sessions, Timeframe & Modality

24, 3 mos., group

40, 5 mos.,

individual

164 mos.,

individual

603 mos., group

243 mos.,

individual

Tx Content SS SDPT, 12-Step

CTPSD, CBT, Exposure

CBT, RPT, Soc Support

SS, RPT, TAU

Follow Up 3 mos. 1 mo. 6 mos. 6/12 mos. 6/9 mos.

Results SU, PTSD & Depression SXs

SU, PTSD, Psych SXs

SU, PTSD & Depression SXs

SU, PTSD SXs

SU, PTSD, Psych SXs at 6 mos

Summary Statistics for Post-treatment PSS-SR Severity Modeled by Baseline PTSD Severity, Demographics, Treatment Group, Time, Site, and their Interactions

Parameter DF 2 p-value

PSS-SR Baseline 1 71.73 <.001

Race 3 1.34 n.s.

Site 5 24.19 <.001

Age 1 0.68 n.s.

Education 1 4.97 <.05

Time 1 1.08 n.s.

Treatment 1 6.81 <.01

Time*Treatment 1 4.96 <.05

Time* PSS-SR Baseline 1 12.35 <.001

Summary Statistics for Post-treatment PSS-SR Frequency Modeled by Baseline PTSD Frequency, Demographics, Treatment Group, Time, Site, and their Interactions

Parameter DF 2 p-value

PSS-SR Baseline 1 76.74 <.001

Race 3 0.64 n.s.

Site 5 27.60 <.001

Age 1 0.95 n.s.

Education 1 5.07 <.05

Time 1 1.47 n.s.

Treatment 1 8.55 <.01

Time*Treatment 1 5.28 <.05

Time* PSS-SR Baseline1 15.32 <.001

Summary Statistics for Post-treatment Days of Use Modeled by Baseline Days of Use, Demographics, Treatment Group, Time, Site, and their Interactions

Parameter DF 2 p-value

Baseline Days of Use 1 21.66 <.001

Race 3 1.26 n.s.

Site 5 18.37 <.01

Age 1 0.50 n.s.

Education 1 0.95 n.s.

Time 6 11.90 <.10

Treatment 1 0.01 n.s.

Baseline Days of Use*Time 6 20.80 <.01

Summary Statistics for Post-treatment Abstinence Rates Modeled by Baseline Abstinence, Demographics, Treatment Group, Time, Site, and their Interactions

Parameter DF 2 p-value

Baseline Abstinence 1 51.12 <.001

Race 3 2.10 n.s.

Site 5 44.18 <.001

Age 1 0.18 n.s.

Education 1 0.00 n.s.

Time 1 0.02 n.s.

Treatment1 0.12 n.s.