Treating Late-life Anxiety in Primary Care: Current Status and Future Directions Melinda Stanley, Ph.D. Menninger Department of Psychiatry and Behavioral.

Treating Late-life Anxiety in Primary Care: Current Status and Future Directions

Melinda Stanley, Ph.D.

Menninger Department of Psychiatry and Behavioral Sciences, Baylor College of Medicine

Houston Center for Quality of Care and Utilization Studies, MEDVAMC

Community Prevalence of Anxiety Disorders in Later Life

Overall 6-month rates: 3.5 - 10.2% Most prevalent: GAD (1.9 – 7.3%)

More common than depression

Higher rates among institutionalized elderly

Prevalence of clinically significant anxiety symptoms even higher (20%)

Prevalence of Late-Life Anxiety in Primary Care

Anxiety Disorders

GAD Anxiety Symptoms

General medical samples

11 -15% 3-11% 20%

Patients with chronic

illness20% - 38% Up to

19%16% - 40%

Impact of Anxiety in Later Life

Physical activity & disability

Functional status Over and above

physical health status

Self-perceptions of health

Quality of life and satisfaction Loneliness

Increased mortality

Service utilization and health care costs

Treatment of Late-life Anxiety Most patients seen in primary care

Many remain unrecognized and without treatment

Medication most common modality Benzodiazepines:

• 10-20% (up to 43%)• Efficacy data• Risk-benefit profile limits utility

Antidepressants • Efficacious, but still drawbacks

Cognitive Behavior Therapy (CBT) for Late-Life Anxiety

Acceptance and preference data

Approach is Time-limited Directive Collaborative

Efficacy data Treatment of choice for anxiety in younger adults Older adults with depression

Goals of our Late-Life Anxiety Research Program:

Evaluate efficacy of CBT for late-life GAD Two academic clinical trials (1996, 2003)

Modify and test treatments in primary care Pilot work Ongoing trial Next efforts

Move from efficacy to effectiveness Increasing emphasis on translational value and

sustainability in primary care

Efficacy: Academic Clinical trials of CBT for Late-life GAD CBT includes

Education/awareness Relaxation Cognitive Therapy Exposure

12-15 weeks Control groups:

WL (d = .71) Supportive treatment (d =

.20 - .29)

Response rates Group treatment:

• Post: 28-45%• FU: 50-63%

Individual treatment:• Post: 40-75%• FU: 57-75%

Stanley MA, et al. J Consult Clin Psychol. 2003; Stanley MA, et al. Behav Ther. 1996; Wetherell JL, et al. J Consult Clin Psychol. 2003; Mohlman J, et al. Am J Geriatr Psychiatry. 2003; Mohlman et al., Behav Res Ther 2005.

Limitations of the Data Homogeneous and

non-representative samples Young old (mean age

mid-60’s) Medically healthy Cognitively intact Caucasian (80% – 90%) Well educated (mean

14 years)

Treatment not designed for older adults

Outcomes not optimal

All data from academic clinical settings

First steps toward effectiveness Develop collaborative relationship with primary care

Kelsey-Seybold Clinic Baylor Family Medicine MEDVAMC

Modify intervention model to: Improve outcomes Identify and meet the needs of more representative patient groups

Move toward collaborative care models Younger adult anxiety (Rollman, Craske) Older adult depression (Unutzer, Bartels) Effect sizes = .23 to .57

New Version of CBT for Late-Life GAD in Primary Care

8 – 10 sessions Increased attention to

learning & memory More flexible

treatment: Individual care Vary # and timing of

meetings Home visits Differential emphasis of

coping skills

Wider range of treatment components: Education/awareness Relaxation Changing thoughts Exposure Problem-solving Sleep management

skills

Pilot Work (Stanley et al., 2003, Am J Geriatr Psychiatry)

Open clinical trial, n = 8 Test feasibility Modify treatment based on therapist and patient input

Small clinical trial, n = 12 CBT vs. usual care (UC) Benefits in worry & GAD severity, depression

Ongoing trial: STOP Worry Project(NIMH renewal)

Goal: n = 150 older primary care patients with GAD

Settings Kelsey-Seybold Clinic Baylor Family Medicine

Inclusion criteria GAD MMSE > 24 English-speaking

STOP Worry Project Team

Investigators

Mark Kunik, M.D. Diane Novy, Ph.D. Nancy Wilson, M.A.,

L.M.S.W. Howard Rhoades,

Ph.D. Anthony Greisinger,

Ph.D.

Current Study Staff

Paula Wagener, B.A. Cynthia Kraus, Ph.D. Christopher Griffith,

B.A. Jessica Calleo, M.A. Brandon Weiss

Design Elements

Efficacy Random

assignment Comprehensive

outcome battery Expert clinicians

and independent evaluators

Treatment integrity raters

“Clean” treatment

Effectiveness Recruit & treat in

primary care Less restrictive

inclusion criteria Telephone-based

assessments Control condition =

UC More flexible

treatment

 Stop Worry Design

MMSE

Consent & Screen

Exclude, Refer

Exclude, Refer

3-Month Assessment

Follow-Up Assessments (1 year)

UCCBT

SCID & BaselineAssessment

STOP Worry Design (cont’d)

CBT Conducted in the primary care clinic (or patient’s home) Master’s level clinicians Treatment integrity ratings - excellent

UC Biweekly telephone contact

Assessments (3 mo. & 1-year FU) Telephone-based Independent evaluators

Outcome Measures

Worry GAD Severity (GADSS) Penn State Worry

Questionnaire (PSWQ)

Anxiety Beck Anxiety

Inventory (BAI) SIGH-A (Hamilton)

Depression Beck Depression

Inventory (BDI)

Quality of Life Quality of Life

Inventory (QOLI) SF – 12

Health Service Use

Recruitment Strategies Timeline: 3/04 to 8/06

2 2-month interruptions Total 26 months

Self-referral Brochures: Waiting rooms, newsletters, Information

Center Letters to patients

Provider referral Presentations, meetings E-mail reminders

Recruitment Summary 968 patients referred

381 signed consent, 313 assessed

165 excluded No GAD, cognitive impairment 11 non-study clinical training cases

148 included (99% of target!) Approx. 6 per month 95% self-referred

Sample Characteristics (n = 148)

Demographics 78% women Mean age = 67 years (SD = 5.8) Mean education = 16 years (SD = 3.0) 17% African American; 9% Hispanic 50% using psychotropic medication

Coexistent psychiatric diagnosis 69% at least 1 45% Depression (MDD, Dysthymia, NOS) 26% Specific Phobia 14% Social Phobia

Treatment Progress (1/07)

CBT UC

Randomized 70 64

ITT 61 57

Completed 42 37

Overall attrition = 27%[9% before randomization; 12% before post-treatment;

6% during follow-up]Higher attrition in UC (26%) vs. CBT (13%)

ITT Analysis: PSWQ change scores

0

1

2

3

4

5

6

7

8

9

3-months 6-months

CBT

UC

3-month d = .64 (.001); 6-month d = .73 (.001)

ITT Analysis: SIGH-A change scores

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

3-months 6-months

CBT

UC

3-month d = .28 (ns); 6-month d = .46 (.02)

ITT Analysis: GADSS change scores

0

0.5

1

1.5

2

2.5

3-months 6-months

CBT

UC

3-month d = .35 (.06); 6-month d = .35 (.06)

Summary & Limitations

CBT is moderately effective relative to UC Effect sizes equivalent to other collaborative care

studies (.28 - .73 vs. .23 - .57) Increasing effects over time

Low attrition in CBT (13%) Translational value limited

Homogeneous, non-representative sample Limited generalizability of treatment to real-world care

(expert providers, long sessions, no attention to patient preference)

Future Directions: Peaceful Living Project

Increase emphasis on effectiveness and collaborative care models Examine effectiveness of CBT delivered by

non-expert providers Enhance communication and collaboration

with primary care via use of EMR Modify CBT to increase flexibility and enhance

translational value/sustainability Implement intervention in more diverse

medical setting (BCM, MEDVAMC)

Peaceful Living Team

Academic investigators

Mark Kunik, M.D. Nancy Wilson, M.A.,

L.M.S.W. Jeff Cully, Ph.D. Louise Quijano, Ph.D.,

L.C.S.W. Michael Kallen, Ph.D.,

M.P.H.

Primary care investigators

Michael Crouch, M.D. (BCM)

Victor Narcisse, M.D. (BCM)

Stinson Tillerson, M.D. (MEDVAMC)

Peaceful Living: CBT Modifications

Modular treatment Integrate telephone-based service

delivery Incorporate patient preference Briefer sessions and simplified procedures BA module to target coexistent

depression Extended duration of treatment, with

intensive telephone booster phase

Treatment Components

Core modules Education &

awareness Motivational

exercise Breathing Coping self-

statements

Elective modules Relaxation Cognitive therapy Sleep skills Problem-solving Behavioral

activation Exposure

Treatment Process

First 3 months Sessions 1-3: Core modules Sessions 6-9: Elective modules

• Patient preference: Content, mode Session 10: Review 26% telephone sessions

Second 3 months Intensive telephone follow-up

Peaceful Living: Open Trial Study sites

BCM Family Medicine MEDVAMC

New recruitment procedures Identify potential patients through EMR PCP sends letter of invitation

Recruitment success More physician-referred patients (44% vs. 5% in STOP Worry) Greater diversity (53% women; 67% Caucasian) Higher baseline severity (GADSS, SIGH-A)

Peaceful Living: Open Trial

Treatment process Group assignment:

• Anxiety Clinic Specialists (MA-level or above) • Counselor (BA-level or below)

Adherence/competence equivalent in 2 groups

Treatment progress 16 patients included CBT-ACS (n = 9); CBT-Counselor (n = 7) Completers: 3-months (n = 10); 6-months (n = 8) Within-subject d = .43 to 1.39

SIGH-A: STOP Worry vs. PLWithin CBT effect sizes

0

0.2

0.4

0.6

0.8

1

1.2

1.4

3-months 6-months

STOP WorryPeaceful Living

GADSS: STOP Worry vs. PLWithin CBT effect sizes

0

0.2

0.4

0.6

0.8

1

1.2

3-months 6-months

STOP WorryPeaceful Living

PSWQ: STOP Worry vs. PLWithin CBT effect sizes

0

0.2

0.4

0.6

0.8

1

1.2

1.4

3-months 6-months

STOP WorryPeaceful Living

Peaceful Living: Summary Effects generally are comparable to or

greater than STOP Worry Exception: PSWQ at 6-months

Patients are more diverse and have more severe symptoms

Intervention model is more sustainable

NIMH should give us 5 more years!