The Impact of Co-existing Chronic Pain and Mental Health QUERI Conditions:

QUERI Implementation Seminar Series

Matthew J. Bair, MD, MS

Research Scientist, Roudebush VA Center of Excellence on Implementing Evidence Based Practice and Regenstrief Institute, Inc

Assistant Professor of Medicine, IU School of Medicine, Indianapolis

June 28, 2007

OUTLINEThe “Pain Problem” The “Depression Problem” Pain and Depression Dyad SCAMP Study

Baseline Analyses

Questions and Answers

Objectives

To discuss the impact of co-existing chronic pain and depression/anxiety

To introduce a model to assess and treat both chronic pain and depression (anxiety) concurrently

Brief Visits

ComplicatedPatients

Clinical Reminders

Minimal Resources

JCAHO & VHAMandate toManage pain

PoliciesGuidelinesExpectations

Managing PAIN in Primary Care: Issues and Challenges

PAIN CRISES

Pain accounts for 20% of all clinic visits

Analgesics = 12% of all prescriptions (# 2)

$100 billion dollars/yr in health care costs

Excessive surgery (e.g., back pain)

Leading cause of work loss & disability

Leading reason for alternative medicine

Consequences of Under-treatment of Chronic Pain

Physiologic (CV, GI, immune)Psychological (depression, anxiety)Diminished quality of LifeImpairment of activitiesLarge impact on working age adults

Absenteeism, unemployment, and under-employment

VETERAN STORIES “Doc, I hurt all day- 24/7” “Nothing works for my pain” “I can’t do anything because of my pain so I

stay in bed all day” “I can’t deal with this (pain)…it’s depressing” “On a scale of 0 to 10 my pain is a 20! If I

don’t get some relief fast I will blow my head off!”

Prevalence of chronic non-cancer pain in Primary Care

44% (VA); 25 %(university, PCC)

(Reid et al,2002 )

48% VA Primary Care - Palo Alto VA

(Clark, JD, 2002)

• 71 % VA Primary Care – Western New York (Crosby et al 2006)

Pain: 5th Vital Sign in Primary Care and Association with Depression

301 primary care Veteran patientsMean age = 60; 91% men; 85% whiteDepression in 28% (PHQ-9 ≥ 10)Pain in 76%

Mild 21% (score of 1-3) Moderate 31% (score of 4-6) Severe 22% (score of 7-10)

Bair MJ, Williams LS, Kroenke K. J Gen Intern Med 2004;19 (Supplement 1):123.

Pain Severity as Correlate of Depression

Pain Severity Odds Ratio (95% CI) for Depression

MildMild 2.2 (1.1 - 4.4) 2.2 (1.1 - 4.4)

ModerateModerate 5.2 (2.2-12.5)5.2 (2.2-12.5)

SevereSevere 12.0 (4.1-34.4)12.0 (4.1-34.4)

Bair MJ, Williams LS, Kroenke K. J Gen Intern Med 2004;19 (Supplement 1):123.

Global Burden of Disease

1. Ischemic heart disease

2. Major Depression

3. Traffic accidents

4. Cerebrovascular disease

5. COPD

Year 2020

Murray and Lopez, 1996

Depressed Patients Usually Present with Physical Symptoms

69%PresentedONLY With Physical

Symptoms

Other

N = 1146 patients with major depression

1. Simon GE, et al. N. Engl J Med. 1999;341(18):1329-1335.

Unrecognized and Untreated Depression

Interferes with treatment and rehabMay increase pain intensity and

disabilityDecrease pain threshold and toleranceMagnification of medical symptomsLess successful treatment outcomes

PainPain DepressionDepression

RECIPROCAL RELATIONSHIP

Depression and Negative Pain Outcomes

Depression is associated with↑ pain complaints and intensity↑ disability↑ functional limitations↑ utilization (office visits, hospitalizations)↑ costs↑ risk of nonrecovery

Bair MJ, et al. Arch Intern Med. 2003;163(20):2433-2445.

Pain and Negative Depression Outcomes

PAIN ASSOCIATED WITH: depressive symptoms functional limitations unemployment rate frequent use of opioid analgesics frequent pain-related doctor visits worse self-rated health

Von Korff M. Grading the severity of chronic pain. Pain 1992; 50:133-149

Bair MJ, et al. Psychosom Med. 2004;66(1):17-22.

0

1

2

3

4

5

Mild Moderate Severe

Severity of Pain is Associated with Poor Depression Outcome

N=573

Odd

s R

atio

for

Poo

r D

epre

ssio

n R

espo

nse*

* R

elat

ive

to P

atie

nts

With

out P

ain

*

*P<.05 vs patients with no baseline pain

1.5

4.1

ARTIST=A Randomized Trial Investigating SSRI Treatment.**Poor depression treatment response defined as Symptom Checklist-20 >1.3. Pain severity was measured by the SF-36 pain severity item

Baseline Pain Severity

2.0

*

(n=144) (n=170) (n=81)

No effect relative to patients without pain at baseline

What Symptoms are the Most Resistant?

Adapted from: Greco T, et al. J Gen Intern Med. 2004;19(8):813-818.

Impr

ovem

ent

Tre

atm

ent E

ffect

Siz

e

Em

otio

nal

Ph

ysical

ARTIST=A Randomized Trial Investigating SSRI Treatment.

Nonsomatic depressive Sx

Positive well-being

Non-pain somatic Sx

Pain somatic Sx

0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

Baseline 1 Month 3 Months 6 Months 9 Months

N=573

Residual Symptoms Predict Relapse

*Based on Item 13 (general somatic symptoms) of the HAM-D17.

Paykel ES, et al. Psychol Med. 1995;25(6):1171-1180.

0

20

40

60

80

100

% R

elap

se

Patients With ResidualDepressive Symptoms

Patients With No ResidualDepressive Symptoms

25%

76%

94%had

PhysicalSymptoms

© 2

006

Nog

ginS

torm

Lab

s

Integrated Model

Physical Physical SymptomsSymptoms

Psychological Psychological SymptomsSymptoms

PAIN is the most common physical symptom

DEPRESSION most common psychological symptom

Research Spectrum for Pain

Basic •Neurosciences•Genetics•Pharmacology•Imaging

Translational

Clinical Trials

Other Clinical•Epidemiology•Health services

Social•Qualitative•Behavioral•Sociological

YOU YOU ARE ARE

HEREHERE

Primary Care Pain and Depression Trial

SStepped CCare for AAffective disorders and MMusculokeletal

PPain studyFunded by National Institute

of Mental Health-RO1 MH071268-01

SCAMP STUDY TEAM

Kurt Kroenke, MD: Principal Investigator Matt Bair, MD: Co-I (Medical Director) Teresa Damush, PhD: Co-I ( Health psychology) Jason Sutherland, PhD: Co-I (Biostatistics) Shawn Hoke (Project Manager) Carol Kempf, RN and Gloria Nicholas, RN Monica Huffman and Celeste Nicholas Jingwei Wu (Data analyst)

Comorbidity of Pain and Depression Is Common

Reviews have demonstrated a strong association

30-60% overlapCoexisting musculoskeletal pain

with depression is very common

Bair MJ, et al. Arch Intern Med. 2003;163(20):2433-2445.

What is a Stepped-Care Intervention?

Starting with lower intensity, less costly treatments (Step 1)

“Stepping up” to more intensive, costly, or complex treatments In patients that are “poor responders”

Low Back Pain (Von Korff), PGW Syndrome (Engel)

SCAMP DESIGN

PAIN (back or hip/knee)

Stepped Care Usual Care

NONDEPRESSEDDEPRESSED

randomized

(n = 250) (n = 250)

Outcome Assessment at 1, 3, 6, and 12 months

HYPOTHESES Depression/pain care management will, Depression/pain care management will,

compared to usual care:compared to usual care:

Primary Hypothesis Reduce pain and/or depression severity

Secondary Hypotheses Improve health-related quality of life (HRQL),

including work and social functioning Improve pain beliefs/behaviors Be cost-effective in terms of QALYs

STUDY SITES Roudebush VAMC medicine clinics University primary care clinics

Clinical Trial Inclusion Pain located in low back, hip or

kneePersistent pain for > 3 months Brief Pain Inventory score of 5

(moderate pain severity)Moderate depression (PHQ-9 10)

Exclusion Criteria Non-English speaking Moderately severe cognitive impairment Bipolar disorder or schizophrenia Current disability claim being adjudicated for pain Tried to cut down on drugs or alcohol in the past

year Currently pregnant or planning to become pregnantAnticipated life expectancy ≤ 12 months

Cohort InclusionHad to have a PHQ-9 depression

score < 8 Identical inclusion/exclusion criteria

to participants in trialTo elucidate frequency & predictors

of incident depression in patients with musculoskeletal pain

Step 1 -- Pharmacotherapy

WHENWHEN WHEREWHERE WHAT (Treatment Action)WHAT (Treatment Action)

BaselineBaseline Clinic Antidepressant startedAntidepressant started

1 wk1 wk Phone Check adherence & side effectsCheck adherence & side effects

3 wk3 wk Phone Adjust dose if neededAdjust dose if needed

6 wk6 wk Clinic Change antidepressant if neededChange antidepressant if needed

9 wk9 wk Phone Adjust dose if neededAdjust dose if needed

12 wk12 wk Clinic Decide if step 2 is warrantedDecide if step 2 is warranted

Antidepressant SelectionVenlafaxineFluoxetineSertralineCitalopramBuproprionMirtazepineNortriptyline

Step 2 – Pain Self-Management

WHENWHEN WHEREWHERE WHAT (Treatment Action)WHAT (Treatment Action)

12 wk12 wk Clinic PSMP – Session 1PSMP – Session 1

14 wk14 wk Phone PSMP – Session 2PSMP – Session 2

16 wk16 wk Clinic PSMP – Session 3PSMP – Session 3

18 wk18 wk Phone PSMP – Session 4PSMP – Session 4

20 wk20 wk Clinic PSMP – Session 5PSMP – Session 5

22 wk22 wk Phone PSMP – Session 6PSMP – Session 6

24 wk24 wk Clinic Close Phase 2. Phone q 3 mo.Close Phase 2. Phone q 3 mo.

Pain Self-Management Program (example components)

Education – pain; vocabulary; red flags;Education – pain; vocabulary; red flags; Identifying /modifying fears and beliefsIdentifying /modifying fears and beliefs Goal-setting and problem-solvingGoal-setting and problem-solving Exercise – strengthening; aerobic; etc.Exercise – strengthening; aerobic; etc. Relaxation; deep-breathing;Relaxation; deep-breathing; Handling pain flare-ups Handling pain flare-ups Working with clinicians and employersWorking with clinicians and employers

DETAILS OF TREATMENT

All aspects of intervention delivered by nurse case manager

Weekly case management meetingsRegular contacts with participants to

monitor depression/pain, response to treatment, introduction of self-management strategies

SCAMP CONCEPTUAL MODEL

Anti-depressant

Pain Self-management

Depression severity

Pain severity

• Impaired Function/QoL

• Increased Health Costs

• Demographics

• Other Psych.-- Anxiety-- Stressors

• Pain-- Coping-- Beliefs

COVARIATES

+

+

− −

MEASURESBrief Pain InventorySCL-20 depression scaleHRQL: -- generic (SF-36)

-- pain-specific (Roland)Other pain (coping, beliefs, self-mgmt)Other psych (anxiety, somatization)Health care utilization (costs)

BASELINE DATA

Baseline Characteristic

Stepped Care

(N=123)

Usual Care

(N=127)

Non depressed

(N=250)

Mean (SD) age, yr 55.2 (12.6) 55.8 (11.0) 62.5 (14.1)

Women, n (%) 69 (56.1%) 63 (49.6%) 127 (50.8%)

Race, n (%)

White 75 (61.0%) 76 (59.8%) 140 (56.0%)

Black 42 (34.1%) 49 (38.6.7%) 100 (40.2%)

Married, n (%) 48 (39.0%) 44 (34.7%) 97 (38.8%)

Mean (SD) no. of medical diseases

2.7 (1.6) 2.6 (1.4) 2.6 (1.4)

Clinical site, n (%)

University clinics 73 (59.3%) 75 (59.1%) 152 (60.8%)

Veteran administration (VA)

50 (40.7%) 52 (40.9%) 99 (39.2%)

Baseline Characteristics SCAMP Participants

0

0.5

1

1.5

2

2.5

Baseline 12 wk 20 wk 32 wk 40 wk

Eff

ec

t S

ize

(n=86) (n=79) (n=54) (n=53) (n=45)

Depression (PHQ-9)

Pain (BPI)

Response of pain and depression in SCAMP Trial during Phase 1 (optimized antidepressant therapy) and Phase 2 (pain self-management)

End of Phase 1

End of Phase 2

Impact of Depression and Anxiety Alone and in Combination among Primary Care Patients

with Chronic Musculoskeletal Pain

Baseline data analysis

BACKGROUNDIndividually, depression and anxiety

are strongly associated with chronic pain.

Little is known how psychiatric comorbidity affects patients with pain.

Symptom Triangle of Reciprocating Adverse Effects

Depression

Anxiety

Pain

–––– ––––

––––

STUDY OBJECTIVEAmong patients w/ chronic pain:Individual and combined impact of

depression and/or anxietyPain intensityPain interferenceDisability daysHealth-related quality of life (HRQL)

METHODSBaseline analysis of SCAMP data4 cohorts identified

Pain only Pain and Anxiety Pain and Depression Pain, Anxiety, and Depression

ANALYSESANOVA models to compare

baseline differences four groupspain intensity/interferenceDisability daysHRQL

ANALYSESMANOVA to model pain severity

and pain interference concurrentlyInteraction testingCovariates:

Sociodemographics, medical comorbidity, study site, and pain location

RESULTS

3

4

5

6

7

8

Pain Only

Pai

n S

core

Pain & Anxiety

Pain & Depression

Pain, Anxiety, & Depression

BPI Interference

BPI Severity

Patients with concomitant pain, depression, and anxiety had more severe pain

Psychiatric Comorbidity & Disability in 500 pts with musculoskeletal pain

Depression and/or Anxiety

%

Roland Disability Score

Disability Days past 3 mo

Pain only 54% 12.1 18

Pain & Anxiety 7% 15.5 33

Pain & Depression 15% 17.9 37

Pain, Anxiety, & Depression 25% 18.1 42

Bair, Damush, Wu, Sutherland, Kroenke (abstract at SGIM meeting, 2007)Bair, Damush, Wu, Sutherland, Kroenke (abstract at SGIM meeting, 2007)

20

40

60

80

Pain only

SF

-36

Sca

le

Pain & Anxiety

Pain & Depression

Pain, Anxiety, &

Depression

Social

Vitality

General

Pain

Poorer HRQL seen in those with pain and psychiatric comorbidity

MANOVA Model Predicting Pain Severity and Interference

Variables F Value P-value

Age 2.24 0.1075

Medical diseases 1.38 0.2525

Gender 0.49 0.6148

Education 3.80 0.0231

Employment 4.13 0.0026

Marital 0.04 0.9595

Pain location 1.82 0.1630

Clinic site 6.52 0.0016

Depression * Anxiety 5.88 0.0030

Study StrengthsLarge, multi-site primary care studyAnalytic plan that assessed

independent & combined effects of depression and anxiety on variety of pain outcomes

Use of validated measures with good psychometric properties

Study LimitationsCross-sectional dataSingle academic medical centerSelf-report measures for depression

and anxiety (rather than structured clinical interviews)

CONCLUSIONAmong primary care patients with

chronic musculoskeletal painWe found an independent and

additive associations between depression and/or anxiety ANDPain intensity/interferenceDisability daysHRQL

IMPLICATIONSPsychiatric comorbidity among

chronic pain patients is commonAssociated with poor clinical

outcomes as well as patient sufferingRepresents complex management

challenges for VA providers caring for these patients

QUESTIONS/COMMENTS

mbair@iupui.edu