STOP: Skills Training for Osteoarthritis Pain An RCT of a CBT-Based Self-Management Intervention for African-American Seniors with Osteoarthritis

1
STOP Skills Training for Osteoarthritis
PainAn RCT of a CBT-Based Self-Management
Intervention for African-American Seniors with
Osteoarthritis

• University of Michigan at Detroit
• Departments of Rheumatology and Symptom
  Management
• July 22, 2004

2
Study Investigators

• John Samuels, MD, MPH Principal Investigator
• Joan Broderick, PhD Co-Investigator
• Ardith Doorenbos, RN, PhD Co-Investigator
• Kim Innes, MSPH, PhD Co-Investigator
• Shannon Jarrott, PhD Co-Investigator
• David Latini, PhD Co-Investigator
• Suzanne Lechner, PhD Co-Investigator
• Francis Keefe, PhD - Consultant
• Patrick McGrath, PhD - Consultant
• Michael Proschan, PhD Statistical Consultant

3
Community Advisory Board

• Joan Cook, RN, MSN, MPH Senior Public Health
  Nurse, Detroit Department of Public Health
• Rev. Jane Doe - Pastor, Detroit AME Church
• Kathryn Kennedy, MSW Director, Patient
  Education, Detroit Arthritis Foundation
• Rev. John Smith Pastor, Jones Methodist Church
  and Chair, Community Advisory Board
• Ronald Williams Director, Health
  Environmental Resource Center

4
Background

• One out of every 3 adults in the United States
  (70 million) is affected by arthritis.
• Leading cause of disability - 18 of population
• Medical costs for arthritis in 1995 were nearly
  22 billion
• When loss of productivity is added, the cost
  increased to 85 billion.
• OA is the most prevalent type of arthritis.
• CDC

5
Background

• Prevalence of arthritis higher in
• women (37 vs. 28),
• older persons,
• those with less education, and
• white and African-American race.
• Michigan is in the highest prevalence category (gt
  35).
• Rates highest among poor (36) and less educated
  individuals (34).

6
Background

• 27 of adults with arthritis report limitation in
  activity
• communication,
• self-care
• mobility
• Learning
• behavior
• Significant disability, defined as the inability
  to perform two or more personal care activities
  higher among African-American (3.4), Hispanic
  (3.5), poor (5.4), and less educated (3.4).

7
Background

• HRQOL consistently worse for persons with
  arthritis,
• healthy days in the past 30 days, days without
  severe pain,
• ability days (days without activity
  limitation),
• difficulty in performing personal care
  activities.
• Healthy People 2010
• Healthy People 2010 Objective 2-1
• Increase the mean number of days without severe
  pain among adults who have chronic joint
  symptoms.

8
Background

• African-Americans with OA less likely to seek
  joint replacement than whites.
• Dunlop et al, Med Care. 2003 Feb41(2)288-98
• COX inhibitors
• attenuate antihypertensive effects of ACE
  inhibitors
• reduce kidney function
• making non-pharmacologic means of controlling OA
  pain more desirable.
• Izhar et al, Hypertension. 2004 Mar43(3)573-7

9
Pilot Work

• Previous work by team with African-American
  elders shows
• Recruiting procedures worked with this population
• Advisory Committee useful in conceptualizing
  study
• Cultural competence appropriate
• Reading level of materials appropriate (lt 6th
  grade)
• Trend toward decreased observed pain behaviors

10
Study Hypotheses

• A cognitive-behavioral treatment program for low
  SES African-American osteoarthritis patients 65
  years of age or older will lead to significantly
  greater reductions in
• Primary Outcome - Observed pain behavior
• Secondary Outcomes - Self-reported pain on a
  visual analog scale and pain diary, disability
  (physical and psychological), activity level

11
Methods
12
Inclusion Criteria

• African-American
• 65 years old
• Physician-confirmed diagnosis of osteoarthritis
• Willing to be randomized
• Household income less than 2x poverty level
• Able to speak, write, and understand English
• Able to participate in physical activity

• Usual pain 4 out of 10, 4 or more days per week
  over two months under usual care
• Pain impairs functional status
• Pain not better attributed to comorbid condition
  (e.g., HIV, diabetes, cancer, etc.)
• Cognitively intact
• Not on medications for psychosis and no more than
  3 psychotropic medications
• No suicidal intent

13
Enrollment - Sources

• Physical therapist
• OA clinics/support groups
• Community mental health centers
• Community health centers/primary care
• Pharmacies/grocery stores
• Churches

• Beauty shops
• Barber shops
• Senior centers
• Arthritis Foundation or other CBO
• Home health agencies
• Nursing homes
• Bus advertising
• Meal programs

14
Enrollment - Approaches

• Advertising newspaper, media coverage,
  newsletters, Flyers, coupon packs, radio
• Talks churches, senior centers
• Alternative medicine practitioners
• Endorsements by community leaders
• Minority recruiters from the community
• HIPAA-compliant

15
Cultural and linguistic considerations

• Community-based organizations (CBO)
• Established relationships with CBO serving
  African-Americans in Detroit
• Community leaders helped formulate community
  approach
• Community advisory board members will publicly
  endorse program and assist us in recruiting from
  their organizations.
• Recruiting materials
• Images and examples designed to appeal to
  African-Americans using guidelines developed by
  Guidry et al, 1996.
• African-Americans are over-represented among
  persons with low health literacy e.g., Bennett,
  JCO, 1998. Materials will be written at the 6th
  grade reading level or below, with pictographs
  used where appropriate.
• Teaching techniques for low literacy groups
  outlined by Doak Doak will be used in the
  intervention (e..g, teach-back).

16
Procedures
17
Eligibility assessment

• Screening phone call
• Ascertain caller meets criteria (pain, age,
  ethnicity, income)
• Screening visit
• Rheumatologist examines potential participant
  (physical exam, medical history, X-ray, current
  meds)

18
Randomization

• Blocked randomization scheme
• Recruit 20 people in a block
• Randomize into two equal groups of 10
• Obtain randomization codes from outside lab
• Randomly assign the 2 groups to condition to
  further avoid bias

19
Baseline assessment

• Within 2 weeks of group start
• Gathered during lab visit (paper-and-pencil
  measures, videotaped behavioral pain assessments,
  etc.)
• Assessors blinded to treatment condition.

20
Follow-up assessment visits

• Process measures after each group
• Pain Behavior Observation Scale
• Self-report measures
• Assessors blinded to treatment condition

21
STOP Intervention Format and Structure

• Closed, structured group intervention (Keefe et
  al., 1990)
• 10 consecutive 2.5-hours weekly sessions
• 45-minute relaxation component,
• 90-minute cognitive-behavioral stress management
  component, and
• 15-minute break.
• Groups of 8-10 participants and two
  co-therapists.
• Therapists who administer the intervention will
  not assess participants
• Multi-modal to enhance treatment efficacy
  (Powell, 2004)
• Std care control with frequency contact

22
Issues and Techniques
Osteoarthritis Issue Intervention Technique
Patient Beliefs Information Provision, Cognitive Restructuring
Treatment Side Effects Coping Skills Training, Relaxation Training
Doctor-Patient Relationship Assertiveness Training, Communication Skills Training
Chronic Pain Concerns Cognitive Restructuring, Coping Skills Training, Information Provision
Spouse/Partner Relations Communication Skills Training, Assertiveness Training
Social Isolation Social Support, Group-based Support
Health Behaviors Self-Awareness, Information Provision, Anger Management, Assertiveness
23
Intervention Sessions
Week Relaxation/Exercise Session Content
1 Lemon Visualization/Diaphragmatic Breathing My Health Body, Awareness
2 Diaphragmatic Breathing/Beach Scene Imagery Awareness/Health Behaviors
3 Sunlight Imagery Automatic Thoughts
4 Color Imagery Distortions, Thought Replacement
5 Meditation Cognitive Restructuring
6 Progressive Muscle Relaxation Effective Coping I
7 Stretching for Older Persons with Osteoarthritis Effective Coping II
8 Tai Chi for Older Persons with Osteoarthritis Anger Management
9 Adapted Yoga Poses for Older Persons with Osteoarthritis Assertion Training
10 Participant Choice Social Support Program Review
24
Intervention Fidelity

• Staff Training
• Facilitators Trained Masters and Ph.D.-level
  clinical health psychologists
• Each leader completes a 10-wk, 20-hour training
  seminar based on the Therapist Training Manual
• Adherence
• Monitor cognitive restructuring skills
  acquisition
• Assess adherence to take-home assignments
• Observational assessment of relaxation skills
• Participant diaries

25
Measures
26
Demographic Variables

• age
• education
• Hollingshead Occupational Index (Hollingshead
  Redlich, 1958) for socioeconomic status (SES)
• religious affiliation
• living situation
• relationship status
• Ethnic identity and cultural values (MEIM
  Phinney, 1992)

27
Clinical/Medical Information

• severity
• date of diagnosis
• treatments prescribed
• Comorbidity (15-item Charlson Scale Charlson et
  al., 1987)
• total number of OTC substances, as well as other
  services obtained

28
Primary Outcome Measure

• Observed pain behavior (Keefe et al., 1990)
• A standardized, videotaped observation protocol.
  Patients engage in 1- and 2-minute periods of
  sitting, walking, standing, and reclining.
  Treatment-blind observers code 5 pain behavior
  categories
• Guarding
• Active rubbing of the knee
• unloading of the joint
• rigidity
• joint flexing
• Diary
• Actigraph

29
Secondary Outcome Measures

• End-of-day pain diary Interactive Voice
  Recording (IVR) instrument will be used to
  collect daily ratings of pain levels on 10-point
  numerical scales.
• Coping Strategies Questionnaire
• Arthritis Self-Efficacy Scale Arthritis Impact
  Measurement Scale (AIMS)
• Stress Management Skills-Measure of Current
  Status (MOCS)
• Measure of physical activity level 7-day
  actigraph

30
Process Measures

• Group Relationship Questionnaire (Schneiderman,
  Antoni, Ironson, 1985)
• Profile of Mood States (POMS McNair, Lorr,
  Droppelman, 1971)
• Group Climate Questionnaire (Schneiderman,
  Antoni, Ironson, 1985)
• Facilitator Debrief Measures
• Facilitator Effectiveness Measures

31
Human Participants

• IRB will approve protocol
• Study staff have completed human subjects and
  HIPAA compliance training
• DSMB will be established
• Adverse events will be tracked
• Additional contact time will be provided PRN

32
Analysis
33
Data Management

• Data entered on scannable forms. Study staff scan
  forms daily and resolve discrepancies as needed.
• Scanned data converted to SAS data sets and
  transmitted to study programmer electronically.
• QA/QC reports run on each batch of data to check
  for out-of-range values, inconsistent values,
  etc. Errors resolved with data entry staff as
  needed.
• Complete data files analyzed by study programmer
  under the direction of Dr. Proschan.

34
Sample Size and Power

• Power analysis
• For 2 group mixed model analysis
• Account for possible effect of group membership
• Using data from previous research, pilot study
• Assume SD of change of 4.6
• Observed values 3.2-4.6 (4.1 in pilot study)
• Assume intraclass correlation coefficient of 0.2
• Observed value 0.16 (pilot study)

35
Sample Size and Power

• To detect a change score difference of 2 points
  in observed pain behavior
• At alpha 0.05 and Power90
• Require total sample size of 400
• 200 treatment (20 groups of 10)
• 200 controls

36
Analysis Baseline variables

• Descriptive statistics baseline characteristics
  of treatment vs.control, attriters vs completers
• Demographics
• OA characteristics
• BMI, comorbidities, activity level
• Differences assessed using
• Chi-square (categorical)
• T-tests (continuous, normal distribution)
• Mann-Whitney U test (ordinal, continuous with
  evidence of skewing)
• Variables differing between tx and controlsgt
  covariates in regression models

37
Primary Analysis

• Intent to Treat
• Mixed Model Linear Regression
• Primary Outcome Observed Pain Behavior
• Random effect group membership
• Fixed effects-
• Primary predictor Treatment
• Covariates Age, gender, other
• Missing data Sensitivity analysis using multiple
  imputation methods (Little Rubin, 2001)

38
Secondary Analyses

• Separate mixed model analyses
• Evaluate effect of intervention (tx) on change
  from baseline in
• Observed pain behavior at 3 mo. post-tx follow-up
• Avg. self-reported pain (pain diary)10 weeks 3
  mo. post-tx follow-up
• AIMS score-pain 10 weeks and 3 mo. post-tx
  follow-up
• AIMS score-psychological disability 10 weeks and
  3 mo. post-tx follow-up
• AIMS score-physical disability 10 weeks and 3
  mo. post-tx follow-up
• Average activity level (actigraphy)10 weeks and
  3 mo. post-tx follow-up
• Fixed effects Tx, covariates
• Random effect Group

39
Thanks for your attention

• Grant award notices gladly accepted

40
Health Behaviors

• Alcohol Use Interview (Warheit, 1987)
• 7-item dietary measure (Thompson et al., 1999).
• Pittsburgh Sleep Quality Index (PSQI Buysse et
  al., 1989)
• CHAMPS survey of physical activity (Stewart et
  al., 2001)