Alex Mitchell ajm80@le.ac.uk

Consultant in Liaison Psychiatry & Psycho-oncology

Which is the best screening tool for Post-Stroke Depression:

- Evidence Based Meta-Analysis

RCPsych Symposium 24-June-2010

Contents

Background

Emotional complications of Stroke

Scales and tools

Reexamining the concept of PSD vs PD

1. Background

Stroke: Definition A syndrome characterized by acute onset of a

neurologic deficit that persists for at least 24 hours, reflects focal involvement of the central nervous system, and is the result of a disturbance of the cerebral circulation.

Transient Ischemic Attack (TIA)Reversible Ischemic Neurologic Deficit (RIND)Stroke in Evolution (Progressive Stroke)Completed Stroke

Stroke Subtypes

ICH13%

SAH13%

Lacunar19% Thromboembolic

6%

Cardioembolic14%

Other 3%Unknown

32%

Ischemic 71%

Hemorrhagic 26%

Data from NINCDS Stroke Data Bank: Foulkes et al. Stroke. 1988;19:547.

Healthy life years lost (YLL) to disease Worldwide

Rank Cause % Rank Cause %1 Lower respiratory infections 8.2 1 Ischaemic heart disease 5.92 Diarrhoeal diseases 7.2 2 Major depression 5.73 Perinatal conditions 6.7 3 Road traffic accidents 5.14 Major depression 3.7 4 Cerebrovascular disease 4.45 Ischaemic heart disease 3.4 5 COPD 4.26 Cerebrovascular disease 2.8 6 Lower respiratory infections 3.17 Tuberculosis 2.8 7 Tuberculosis 3.08 Measles 2.7 8 War 3.09 Road traffic accidents 2.5 9 Diarrhoeal diseases 2.710 Congenital abnormalities 2.4 10 HIV 2.6

1990 2020

Global Burden of Disease Study, 1996

Function

DSMV

Not Just

PSD associated with:Poor functional recovery (in 9 of 12 studies)Reduced quality of lifeIncreased cognitive impairmentIncreased mortality

4 studies suggest Rx depression improves stroke outcomes (all since 1998)3 studies suggest Ads post stroke reduce depression

Carson et al, 2000

PSD Prevalence by DSMIV (n=1200)

3614.321.7

401426Vataja et al (2001) Outpatient 275 PSE-DSM-III-R

26 14 40

401426Pohjasvaara et al (1998) Outpatient 277 PSE-

DSM-III-R 26 14 40

381820Castillo et al (1995) Acute hosp 291 PSE-DSM-III

20 18 38

411922Fedoroff et al (1991) Acute hosp 205 PSE-DSM-III

22 19 41

23815Burvill et al (1995) Community 294 PSE-DSM-III

15 8 23

totalMinor %Major %

PSD by Meta-analysis (n=2000)

major depressive disorder 19% (12.1%, 25.6%)

minor depressive disorder 17% (4.1%, 29.4%)

any depressive disorder 30% (20.9%, 39.1%)respectively

2. Antidepressants and treatments

Treatment of Post-Stroke DepressionIndividual Studies

• Placebo Controlled–Lipsey (1984) n = 34–Reding et al (1986) n= 27–Andersen et al (1994) n=66–Grade et al (1998) n = 21

• Head-to-Head–Lauritzen et al (1994) n = 20–Dam et al (1996) n =52–Robinson et al (2000) n = 56–Jorge et al (2003) n=104

Jorge et al (2003) Am J PsychiatryN=104; 9 year follow upNortriptyline, fluoxetine, placebo (RCT)

Prophylactic treatment with SertralinePoulsen, et al, 2003 Stroke Patients Randomly Assigned to 12 Months of Double‐Blind Treatment With Sertraline or Placebo with GDS >16

Antidepressant Prevention – Meta-analysis

Cochrane 2008 Update• Sixteen trials (17 interventions)• N= 1655 participants,• 13x pharmaceutical agents• 4x trials of psychotherapy.

• “some evidence of benefit of pharmacotherapy in terms of a complete remission of depression and a reduction (improvement) in scores on depression rating scales, but there was also evidence of an associated increase in adverse events.

• There was no evidence of benefit of psychotherapy”

3. Scales and Tools of PSD

How

When

By who?

Issues in screening tools for PSD Gaete, et al. 2008)

Are staff always available for “observer” scales

Can all patients “self report?”

Do somatic symptoms contaminate?

Cognitive impairment and delirium (50%; 20% troubling)

Special deficits (5-10%)Speech and language deficitsAnosognosiaVisual impairmentNeglect

Other GuidelinesRCPhycians on the National Clinical Guidelines for Stroke (UK) recommend that ‘patients should be screened for depression within the first month following a

stroke and their mood kept under review’

SIGN 64 recommended“all stroke patients should be screened for mood disturbance. Some form of

screening should occur initially and at three month intervals or key stages of the rehabilitation process and after rehabilitation support has been lost.

Also“All screening measures have limitations (in specificity and sensitivity) so that

some patients. problems will be missed or overestimated. Currentmeasures may include items concerning, for example, activity or concentration, which may be directly affected by stroke. Screening does not constitute a diagnosis of depression and cannot provide insight into the complexity of the individual.s problems.”

Which Approach?

Observation

Interview

Visual

Self-Report

DepressionScreeningIn Stroke

DISCS

VA-SES

ET/DT

HAMD-D17

PhysicalGeneral

Signs ofDS

6

Stroke AphasicDepression

Scale21/10

MADRAS10

Trained

ConfidentSkilledClinician

Alone

YALE

SMILEY

Self-Report Scales

Concordance with DSMIV and ICD10

Interview Scales• YALE• “Are you depressed”? PPV 30-50%

Visual-Analogue

Most severe depression

No Depression

The DISCs is displayed on a laminated card. • Each circle is 2 cm in diameter.• The scale measures 15 cm from the centre of

the bottom circle to the centre of the top circle.• A pictorial version also available.

Figure 4: The Depression Intensity Scale Circles (DISCs)

Instructions for administration

• This is a scale to measure depressionPlease point to each of the circles in turn to make sure that you can see them all.

• The black circles show how depressed you feel.

[Indicate the clear circle at the bottom]• The bottom circle shows no depression.

[Indicate the fully shaded circle at the top]• The top circle shows depression as bad as it can

be.

[Pointing at each circle in ascending order]• As you go from the bottom circle to the top, you

can see that depression is becoming more and more severe.

• Which of these circles shows how depressed you feel today?

To the administrator:In your opinion was the person able to understand this scale?

Yes � No �

Comment

0

1

2

3

4

5

6

7

8

9

10 Most severe depression

No depression

The NGRS is displayed on a laminated card It measures 10 cm, with numbered increments every 1 cm

Figure 3: The Numeric Graphic Rating Scale (NGRS)

Instructions for administration

• This is a scale to measure depression.

• Please point to Indicates

• The Highest score [should indicate 10] ………• The Mid-point [should indicate 5]

………• The Lowest score [should indicate 0]

………

• The numbers show how depressed you feel.

[Indicate 0]• The bottom of the scale shows no depression.

[Indicate 10]• The top shows depression as bad as it can be.

[Pointing at each number in ascending order]• As you go from the bottom of the scale to the top,

you can see that depression is becoming more and more severe.

• Which point on the scale shows how depressed you feel today?

To the administratorIn your opinion was the person able to understand this scale?

Yes � No �

Smiley Scale

Smiley Scale

Which Approach Works (Validity)?

Meta-analysis –Meader & Mitchell18 studies possibly eligibility criteria, providing data on 2,045 participants

Level 1 - robust meta-analysis requires 4x studies (STATA)

Level 2 - 3x studies

Non-meta-analytic – 1x study

As of Sept 09HADS-D 6x

HAM-D/HDRS 5x

CES-D 4x

GDS 4x

Zung, BDI fast screen, PHQ-2, PHQ-9 1x

Yale / BDI-II / NGRS / DISCS / SMILEY (1x)

Preliminary Results

Preliminary Results 2

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Pre-test Probability

Post

-test

Pro

babi

lity

HDRS+

HDRS-

Baseline Probability

HADS+

HADS-

GDS30+

GDS30-

CES-D+

CES-D-

Comment: Slide illustrates that there is NO ADVANTAGE to scales without somatic symptoms

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Pre-test Probability

Post

-test

Pro

babi

lity

Smiley+Smiley-Baseline ProbabilityDISCS+DISCS-NGRS+NGRS-Yale+Yale-HDRS+HDRS-HADS+HADS-CES-D+CES-D-

Comment: All methods compared regardless of quality of the study

InterpretationSimple tools work surprisingly well but data is weak

BDI, HADS and GDS are not especially strong

Acceptability is key

4. Comorbid Depression

Why BDI/HADS not successful?

Back to Basics

Lipsey (1986) PSD vs PD

Stroke (n=41) vs MI (n=26) on HAMD17Verhey et al (2009)

loss of interest, psychomotor retardation, and gastro-intestinal complaints more common

Approaches to Somatic Symptoms of DepressionInclusiveUses all of the symptoms of depression, regardless of whether they may or may not be secondary to a physical illness. This approach is used in the Schedule for Affective Disorders and Schizophrenia (SADS) and the Research Diagnostic Criteria.

ExclusiveEliminates somatic symptoms but without substitution. There is concern that this might lower sensitivity. with an increased likelihood of missed cases (false negatives)‏

EtiologicAssesses the origin of each symptom and only counts a symptom ofdepression if it is clearly not the result of the physical illness. This is proposed by the Structured Clinical Interview for DSM and Diagnostic Interview Schedule (DIS), as well as the DSM-III-R/IV).

SubstitutiveAssumes somatic symptoms are a contaminant and replaces these additional cognitive symptoms. However it is not clear what specific symptoms should be substituted

Medically Unwell Alone

Primary Depression Alone

Secondary Depression

Comment: Slide illustrates concept of phenomenology of depressions in medical disease

FatigueAnorexiaInsomnia

Concentration

Study: Somatic symptoms study

N= 4500; Pooled database study; All comparative studies

Physical illness+comorbid depressionVsPhysical illness aloneVs

Primary depression alone

Medically Unwell

Primary Depression

Secondary Depression

Comment: Slide illustrates actual phenomenology of depressions in medical disease

Weight loss

AgitationRetardation

SummaryQuestions