Using evidence-based clinical practice guidelines: Examples from the ACCP Antithrombotic and Thrombolytic Therapy Conference Holger Schünemann, MD, PhD

Using evidence-based clinical practice guidelines:

Examples from the ACCP Antithrombotic and Thrombolytic Therapy Conference

Holger Schünemann, MD, PhDDeborah Cook, MD, MSc

Roman Jaeschke, MD, MScJanek Brozek, MD

Gordon Guyatt, MD, MSc

Where would you prefer to live?

← Option 1

Option 2 →

← Option 1 (pink card)

Option 2 → (green card)

Intro: Clinical practice guidelines

What makes guidelines evidence based in 2005?

Strong vs. weak recommendation

High vs. low quality evidence

Grading system

Today’s talk

Intro: Clinical Practice Guidelines1

Clinical Practice Guidelines

Systematically developed statements to assist practitioner and patient decisions about appropriate

health care for specific clinical circumstances

Users’ Guide to the Medical Literature, 2002

Why do clinicians need guidelines?

• Rising Healthcare Cost

• Increasing demand for care

• More expensive technologies

• Variations in service delivery among:

Providers, hospitals and geographical regions*

Assumption that this variation is a result of inappropriate (too much/too little) use of services

*BMJ 1999;318: 527

Clinical Practice Guidelines

…are a result of the desire:

• of healthcare workers to offer and of patients to receive the best possible care

• to make care more efficient and consistent by bridging the gap between what clinicians do and what the evidence shows

The leaky pipeline from research to practice

If 80% achieved at each stage then0.8 x 0.8 x 0.8 x 0.8 x 0.8 x 0.8 x 0.8 = 0.21

Glasziou and Haynes, ACP JC; 2005: 7-9

Aware Accept Target Doable Recall Agree Done

ValidResearch

Do you use guidelines in your practice?

Where do you get your guidelines from?

www.guidelines.gov

“Practice guidelines … have been demonstrated to improve patient

outcomes and lower cost”

S. Weingarten. Hospital Medicine 2005

…be based on sound scientific evidence and implemented in

an effective manner

What makes Guidelines Evidence-Based in 2005?2

First issue of ACCP guidelines in 1986 (CHEST)Initially aimed at consensus Methodologists involved since beginning Now formally convening every 2 to 3 years~200.000 copies in 2001Seventh conference held in 200387 panel members, 22 chapters Across subspecialties565 recommendations, 230 newTranslated: Polish, Spanish, Italian, French

Background: ACCP Antithrombotic and Thrombolytic Therapy Guidelines

Evidence – recommendation: transparent link

Explicit inclusion criteria Comprehensive search

Standardized consideration of study quality

Conduct/use meta-analysis

Grade recommendations

Acknowledge values and preferences underlying recommendations


Schünemann et al. Chest 2004

A bit more practice using the voting instrument….



Remember

You are hiking.

Which of the following animals would you prefer to encounter?



You are buying an ice cream.

Which flavor do you prefer?



Chocolate

Strawberry

You are buying a new car.

Which one would you buy?



Yellow fox

Red Ferrari

What determines your choices?

• pleasure

• social responsibilities

•Risk taking

•Life crisis

•Resources

•Safety

•Past experiences

•Expectations

•Ongoing cost/inconvenience

• impuls control/politics

Case scenario and clinical question

75 year old men with history of hypertension presents to the ED with right upper extremity weakness and slurred speech for approximately two hours earlier in the day. Workup is negative. The symptoms are now resolved. Antihypertensive therapy is initiated.

Which antithrombotic treatment would you recommend?

In elderly men with TIA and hypertension, do antiplatelet agents compared to no antiplatelet agents reduce recurrent strokes?

ACCP Example: Stroke prevention

In patients with history of non-cardioembolic stroke or TIA…, we recommend treatment with an antiplatelet agent (Grade 1A). Aspirin, aspirin + XR dipyridamole or clopidogrel are all acceptable options for initial therapy.

Clopidogrel: Higher cost

If we had to make a choice between aspirin and clopidogrel, what would that choice be?

Albers et al. Chest 2004

Transparent link between evidence and recommendations

&Explicit inclusion criteria


Table 1 Eligibility Criteria

Section Inclusion Criteria

Population Intervention(s) or Exposure Outcome Methodology

… … … … …

4.1. Patients with unstable

angina, MI, TIA and non-acute stroke

Any antiplatelet agent compared with placebo

or one or more other antiplatelet agents (s);

Death Stroke or recurrent stroke Other vascular events

Randomized controlled trials

4.2 Patients with

cardioembolic stroke Oral anticoaluation

Death Stroke or recurrent stroke

Randomized controlled trials

… … … … …

CAPRIE TrialAspirin vs clopidogrel in patients at risk for

cardiovascular event

19,185 patients, 3 subgroups with > 6,300 patients each (TIA/Stroke; myocardial infarction; peripheral arterial occlusive disease)

Mean duration of follow-up: 1.9 years

Primary outcome: ischemic stroke, myocardial infarction, or vascular death

Clopidogrel versus aspirin in patients at risk of ischaemic events. Lancet 1996

0

2

4

6

8

10

Absolute risk%

Clopidogrel 7.15 5.03 3.71 5.32

Aspirin 7.71 4.84 4.86 5.83

Stroke MI PAOD Total

CAPRIE* trial resultsAbsolute risk

**


NNT 200

CAPRIE* trial resultsRelative risk reduction

7.3

-3.7

23.8

8.7

-30

-20

-10

0

10

20

30

40

Relative risk reduction

%Relative risk

Increase

Clopidogrel better

(Aspirin better)

STROKE MI PAOD Total

p = 0.26 0.66 0.0028 0.043


Which of the following recommendations should one give?

1. Aspirin over clopidogrel in patients with prior history of TIA/Stroke?

OPTION 1 (pink)

2. Clopidogrel over aspirin in patients with prior history of TIA/Stroke?

OPTION 2 (green)

Audience at a prior thrombosis meeting

57% 43%

0%

20%

40%

60%

80%

100%

Aspirin Clopidogrel

Preferred recommendation

Strong vs. weak recommendation3

ACCP Recommendations?Stronger recommendations

strong methodslarge precise effect benefits much greater than downsides, or downsides much

greater than benefitsone size fits allexpect uniform clinician and patient behavior

Grade 1 Weaker recommendations

weaker methods imprecise estimatesmall effectbenefits not clearly greater or smaller than downsidesexpect action to vary

Grade 2

Case scenario

65 year old female with history of hypertension and DM type 2 complaining of chest pain. Diagnosed as unstable angina.

Who would recommend aspirin for our patient?

YES (pink)

No (green)

Strong vs. weak recommendation 4

Evidence weak or strong?Study design

basic

detailed design and execution

Consistency

Directnesssecure generalization?

populations (VKA for patients with A. fib and mitral valve stenosis)

interventions (Aspirin the same as clopidogrel?; LMWH)

outcomes (important versus surrogate outcomes; cholesterol)

comparison (A - C versus A - B & C - B)

Grades of recommendationMethodological quality

Grade A: consistent results from RCTs

Grade B: inconsistent results from RCTs or RCTs with

methodological limitations

Grade C: observational studies

Grade C+: observational studies with very strong

effects or secure generalization from RCTs

Example: Stroke preventionIn patients with history of non-cardioembolic stroke or

TIA…:

we recommend treatment with an antiplatelet agent (Grade 1A). Aspirin, aspirin and XR dipyridamole or clopidogrel are all acceptable options for initial therapy.

…, we suggest use of clopidogrel over aspirin (Grade 2B).

Underlying values and preferences:Underlying values and preferences:

This recommendation places a relatively This recommendation places a relatively high value on a small absolute risk reduction high value on a small absolute risk reduction in stroke rates, and a relatively low value on in stroke rates, and a relatively low value on minimizing drug expendituresminimizing drug expenditures


Example: Acute coronary syndrome

For all patients presenting with NSTE ACS, without a clear allergy to aspirin, we recommend immediate aspirin, 75 to 325 mg po, and then daily, 75 to 162 mg po (Grade 1A).



Standardized consideration of study quality

Conduct/use meta-analysis





The ACCP Antithrombotic Therapy grading system

Clear separation of two issues:

Evidence: weak or strong?

methodological quality of evidence

likelihood of bias

Recommendation: weak or strong?

trade-off between benefits and downsides

Values and preferences

If available, they are integrated into recommendations and described by guideline developers

If unavailable, adequate representation of patients’ or society’s interests is assumed

To increase the likelihood of adequate representation, the process included review of recommendations by research methodologists, practicing generalists and specialists

Grading systemGrade of

recommendation Clarity of risk/benefit Strength of supporting evidence

1A Benefits clearly outweigh risk and burdens, or vice versa

Consistent evidence from well performed randomized, controlled trials or overwhelming evidence of some other form. Further research is unlikely to change our confidence in the estimate of benefit and risk.

1B Benefits clearly outweigh risk and burdens, or vice versa

Evidence from randomized, controlled trials with important limitations (inconsistent results, methodologic flaws, indirect or imprecise), or very strong evidence of some other form. Further research (if performed) is likely to have an impact on our confidence in the estimate of benefit and risk and may change the estimate.

1C Benefits appear to outweigh risk and burdens, or vice versa

Evidence from observational studes, unsystematic clinical experience, or from randomized, controlled trials with serious flaws. Any estimate of effect is uncertain.

2A Benefits closely balanced with risks and burdens

Consistent evidence from well performed randomized, controlled trials or overwhelming evidence of some other form. Further research is unlikely to change our confidence in the estimate of benefit and risk.

2B Benefits closely balanced with risks and burdens; some uncertainty in the estimates of benefits, risks, and burdens

Evidence from randomized, controlled trials with important limitations (inconsistent results, methodologic flaws, indirect or imprecise), or very strong evidence of some other form. Further research (if performed) is likely to have an impact on our confidence in the estimate of benefit and risk and may change the estimate.

2C Uncertainty in the estimates of benefits, risks, and burdens; benefits may be closely balanced with risks and burdens

Evidence from observational studies, unsystematic clinical experience, or from randomized, controlled trials with serious flaws. Any estimate of effect is uncertain.

GRADE

Grades of Recommendation

Assessment, Development and

Evaluation

System adopted by:ACCP

UpToDateUrology associations

Endocrine Society

*Grade Working Group. CMAJ 2003, BMJ 2004, BMC 2004, BMC 2005

Summary

Integration of values and preferences is challenging but critical for clinical practice guideline development and application

High transparency between evidence and recommendations required

GRADE approach to grading quality of evidence and strength of recommendations is gaining acceptance and application

QUESTIONS?

End

Evidence alone does not make decisions

Expert opinion is not evidence – expert opinion is an interpretation of the evidence

Finalization and harmonization of the guidelines

Preliminary versions formulated by authors and presented before and during conference

Controversial recommendations were presented during conference

Editors harmonized the chapters and facilitated discussion of contested recommendations

Limitations of guidelines

Possibility that some authors followed this methodology more closely than others

Possibility of missing relevant studies

No centralization of the methodological evaluation of all studies

Few meta-analysis conducted

Sparse data on patients’ values and preferences and resources utilization

Future directions of ACCP Guidelines

Tackle limitations mentioned above

Perform additional evaluations, supervised and coordinated centrally, of the quality of included trials

Formed “Cost” and “Grading” task forces

Merge with GRADE* approach

*Grading Recommendations Assessment, Development and Evaluation Working Group. BMJ 2004



Standard consideration of study quality

( Conduct/use meta-analysis)



1.1. Patient group/condition, outcome, intervention

1.1. Discussion of eligible evidence answering the question

1.1. Statement of values and preferences if not obvious or particularly pertinent to the recommendation

1.1. Recommendation: Based on (quality) evidence, statement of recommendation with wording related to strength (GRADE STRENGTH/EVIDENCE QUALITY).

Summary of Recommendations

What we have achieved

Long Distance Travel

For long-distance travelers with other risk factors for VTE, we recommend the general strategies listed above. If active prophylaxis is considered, because of perceived increased risk of venous thrombosis, we suggest … single prophylactic dose of LMWH, injected prior to departure (Grade 2B).

Geerts et al. Chest 2004

Chronic limb ischemia

We recommend clopidogrel in comparison to no antiplatelet therapy (Grade 1C+), but suggest that aspirin be used instead of clopidogrel (Grade 2A).

Underlying values and preferences: This recommendation places a relatively high value on avoiding large expenditures to achieve small reductions in vascular events.

We recommend clopidogrel over ticlopidine (Grade 1C+)


Explicit inclusion criteria Comprehensive search Standard consideration of

study quality( Conduct/ use meta-analysis) Grade recommendations Acknowledge values and

preferences underlying recommendations

Knee Arthroscopy

For patients undergoing arthroscopic knee surgery we recommend against routine thromboprophylaxis, other than early mobilization (Grade 2B).

For patients undergoing arthroscopic knee surgery and who are at higher than usual risk, based on pre-existing VTE risk factors or following a prolonged or complicated procedure, we suggest thromboprophylaxis with LMWH (Grade 2B).

Geerts et al. Chest 2004

Why Grade Recommendations?

Strong recommendationone size fits all

expect uniform clinician and patient behavior

Weaker recommendation expect action to vary

Factors that influence the strength of the recommendation

Issue Example

Evidence for less serious event than one hopes to prevent

Preventing post-phlebitic syndrome with thrombolytic therapy in DVT rather than death from PE.

Smaller Treatment Effect

Clopidogrel versus aspirin leads to a smaller stroke reduction in TIA (8.7%% RRR) than anticoagulation versus placebo in AF (68% RRR)

Imprecise Estimate of Treatment Effect

ASA versus placebo in AF has a wider confidence interval than ASA for stroke prevention in patients with TIA

Lower Risk of Target Event

Some surgical patients are at very low risk of post-operative DVT and PE while others surgical patients have considerably higher rates of DVT and PE

Higher Risk of Therapy

ASA and clopidogrel in acute coronary syndromes have a higher risk for bleeding than ASA alone

Higher Costs

TPA has much higher cost than streptokinase in acute MI

Varying Values

Most young, healthy people will put a high value on prolonging their lives (and thus incur suffering to do so); the elderly and infirm are likely to vary in the value they place on prolonging their lives (and may vary in the suffering they are ready to experience to do so).

Quality of evidenceThe extent to which one can be confident that an estimate of effect or association is correct. This depends on the:

study design (e.g. RCT, cohort study, case series)

study quality (protection against bias; e.g. concealment of allocation,blinding, follow-up)

consistency of results

directness of the evidence including the

populations (those of interest versus similar; for example, older, sicker or more co-morbidity)

interventions (those of interest versus similar; for example, drugs within the same class)

outcomes (important versus surrogate outcomes)



• Evidence for less serious event than one hopes to prevent

• Smaller Treatment Effect

• Imprecise Estimate of Treatment Effect

• Low Risk of Target Event

• Higher Risk of Therapy

• Higher Costs

• Varying Values

• Higher Burden of Therapy


Peripheral arterial occlusive disease

We recommend lifelong aspirin therapy (75 - 162 mg/d) in comparison to no antiplatelet therapy in both patients with clinically manifest coronary or cerebrovascular disease (Grade 1A) and those without clinically manifest coronary or cerebrovascular disease (Grade 1C+).

We recommend clopidogrel in comparison to no antiplatelet therapy (Grade 1C+).

Which of the following options would you recommend?

1. Aspirin over clopidogrel in patients with PAOD?

OPTION 1

2. Clopidogrel over aspirin in patients with PAOD?

OPTION 2

PAOD

In patients with PAOD we suggest that aspirin be used instead of clopidogrel (Grade 2A).

Underlying values and preferences:

This recommendation places a relatively high value on avoiding large expenditures to achieve small reductions in vascular events.

Evidence weak or strong?study design

basic

detailed design and execution

consistency

directnesssecure generalization?

populations (VKA for patients with A.fib and mitral valve stenosis)

interventions (Aspirin the same as clopidogrel; LMWH)

outcomes (important versus surrogate outcomes; cholesterol)


Why Grade Recommendations?Strong recommendations

strong methods large precise effect few downsides of therapy

Weak recommendationsweak methodsimprecise estimatesmall effectsubstantial downsides


strong methods large precise effect few downsides of therapyone size fits allexpect uniform clinician and patient behavior

Weak recommendationsweak methodsimprecise estimatesmall effectsubstantial downsides


strong methods large precise effect few downsides of therapyone size fits allexpect uniform clinician and patient behavior

Weak recommendationsweak methodsimprecise estimatesmall effectsubstantial downsidesexpect action to vary

Chapter authors

Develop the Clinical Question

Organize by patient groups or conditionsExamples from chapter on Ischemic Stroke

Previous Now

Stroke Prevention Antiplatelet agents Non-cardioembolic stroke Cardioembolic stroke Oral Anticoagulation Cardioembolic stroke Non-cardioembolic stroke

Stroke Prevention Non-cardioembolic stroke Antiplatelet agents Oral Anticoagulation Cardioembolic stroke Oral Anticoagulation Antiplatelet agents

Explicit eligibility criteria

Example: Thrombolysis compared with no thrombolysis for acute stroke

Patients: Patients presenting with acute thrombotic stroke

Intervention: any thrombolytic regimen

Outcome: death, or validated functional status instrument

Methodology: randomized trials

Trombolisi confrontata con non trombolisi per stroke acuto

Questo quesito clinico orienta verso diverse raccomandazione:1.1. tPA per via intra-venosa in caso di stroke ischemico acuto

caratterizzato dalla presenza di sintomi per < 3 ore

1.2. tPA per via intra-venosa per stroke ischemico acuto caratterizzato dalla presenza di sintomi dalle 3 alle 6 ore

1.3. Streptokinase intravenoso in caso di stroke ischemico acuto caratterizzato dalla presenza di sintomi < 3 ore

1.4. Streptokinase intravenoso per stroke ischemico acuto caratterizzato dalla presenza di sintomi dalle 3 alle 6 ore


Explicit inclusion criteria Comprehensive search Standard consideration of

study quality Conduct/ use meta-analysis Grade recommendations Acknowledge values and

preferences underlying recommendations

Role of librarians

Use questions to develop search strategy e.g. identify all search terms (MESH and keywords) for antiplatelet

agents or myocardial infarction

Search:Cochrane database of systematic reviews

Database of Abstracts of Reviews of Effectiveness

Cochrane Register of Controlled Trial

MEDLINE and Embase (1966 - Dec 2002)

ACP Journal Club

Provide search resultsUsed Endnote ® software

e.g. 490 citations on thrombolysis in acute stroke

Chapter authors

Identifying the Clinical Question:Prior experience

Prior recommendations

What matters in clinical practice

The questions:

Identify patients, interventions, and outcomes, but also methodological criteria

Methodological quality

Criteria for baseline risk studies in specific populations:

Cohort studies reporting of at least 200 participants

Control groups of RCTs reporting 200 participants

Focus in similar populations

Sufficient length of follow-up

Less than 20% loss to follow-up

Case scenario

A 67 year old engineer is brought to the ER with tachyarrhythmia and near syncope. An EKG reveals atrial fibrillation. Other workup is negative, but the patients states that he – on and off – felt his heart racing for several days. Together with your team you diagnose the patient with lone atrial fibrillation.


Schünemann HJ et al. Chest 2004

Documents

Using evidence-based clinical practice guidelines: Examples from the ACCP Antithrombotic and Thrombolytic Therapy Conference Holger Schünemann, MD, PhD