![Page 1: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/1.jpg)
Summarizing the Evidence
Jane Gagliardi, MD, MHS
Megan von Isenburg, MSLS
![Page 2: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/2.jpg)
ASSESS
ASK
ACQUIRE
APPRAISE
APPLY
The 5 A’s
EBMCycle
MUST CONSIDER: - Patient preference- Access to care- Quality of life- Goals of care
WHAT’S GOING ON?- History and Physical- Initial Formulation
PICOTT- Patient /
Population- Intervention- Control- Outcome- Type of Question- Type of Study
LITERATURE SEARCH
VALIDITY CRITERIA- Methods- Results- Sources of Bias- Strength of evidence
The Evidence
Cycle
![Page 3: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/3.jpg)
Objectives
• Recognize and use the evidence cycle
• Define a systematic review
• Define meta-analysis– Discuss what they can and can’t do for you
• Be able to explain:– Heterogeneity
– Weighting
– Publication Bias
• Draw and interpret a forest plot
• Critically appraise a systematic review
• Discuss teaching strategies and decision-points
![Page 4: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/4.jpg)
The Evidence Pyramid
![Page 5: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/5.jpg)
Scenario: Rounding on Gen Med
![Page 6: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/6.jpg)
Why Do We Need EBM? Consider…
• A study of steel mill workers identified by
on-the-job screening program with
hypertension
– ½ sent to primary care providers for specific
treatment of hypertension
– ½ sent to on-the-job-site provider for
algorithm-based treatment of hypertension
Adapted from Virginia Moyer, 1/14/2014
![Page 7: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/7.jpg)
• A study of steel mill workers identified by
on-the-job screening program with
hypertension
• Unexpectedly low rates of treatment by
primary care providers
Adapted from Virginia Moyer, 1/14/2014
Why Do We Need EBM? Consider…
![Page 8: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/8.jpg)
Factors associated with likelihood of
prescription of an antihypertensive agent:
• Diastolic blood pressure
• Age
• Target end-organ damage
•
Adapted from Virginia Moyer, 1/14/2014
Why Do We Need EBM? Consider…
![Page 9: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/9.jpg)
Factors associated with likelihood of
prescription of an antihypertensive agent:
• Diastolic blood pressure
• Age
• Target end-organ damage
• Time since graduation from medical school
Adapted from Virginia Moyer, 1/14/2014
Why Do We Need EBM? Consider…
![Page 10: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/10.jpg)
Scenario: Rounding on Gen Med
![Page 11: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/11.jpg)
Scenario: Pharmacy and Therapeutics
• For as long as anyone can remember, the
formulary has included metoprolol.
• Carvedilol was reviewed for formulary
status in 2003, at which time it was felt
to provide no specific benefit (but cost
more).
![Page 12: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/12.jpg)
• You now are the multidisciplinary health
system Pharmacy and Therapeutics
Committee.
Scenario: Pharmacy and Therapeutics
![Page 13: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/13.jpg)
Scenario: Pharmacy and Therapeutics
• Health system “nonformulary” report
includes large increase in the utilization of
carvedilol, particularly on the Gen Med
services for patients with heart failure.
• The Chair of the committee tasks us with
deciding which beta blocker will be
preferred for formulary use: Metoprolol or
Carvedilol? (Cost is now similar).
![Page 14: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/14.jpg)
A = Metoprolol
B = Carvedilol
What Should the Health System Do?
![Page 15: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/15.jpg)
What is the focused
clinical question? Among patients
with heart failure, is
carvedilol significantly
different than
metoprolol in
preventing all-cause
mortality?
![Page 16: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/16.jpg)
Searching the Literature
![Page 17: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/17.jpg)
Your Task
• The committee needs to decide.
• The Pharmacy administrator has a stack
of journal articles.
• Is there any alternative to looking at each
of 50 randomized controlled trials?
![Page 18: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/18.jpg)
• Answers one focused clinical question
• Summarizes evidence using methods to minimize
the impact of bias
• The statistical method to combine data from
different studies = meta-analysis
• Not all systematic reviews have meta-analysis
(qualitative inferences only)
• Not all meta-analyses combine studies
assembled through a systematic review
What is a Systematic Review?
![Page 19: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/19.jpg)
What systematic reviews can do for you
• Save time !!!
• Increase power to detect rare events– Obviate need for expensive mega-trials
– Detect harm
• Increase the precision of the estimate of effect
• Enhance the generalizability of the results if samples from different populations are included
• Look for important differences in effectiveness among subgroups of patients
![Page 20: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/20.jpg)
Random Error Systematic Bias
What systematic reviews can’t do for you
![Page 21: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/21.jpg)
Random Error Systematic Bias
What systematic reviews can’t do for you
…they also can’t tell you where TRUTH actually is
![Page 22: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/22.jpg)
Exercise – Creating a Mini Meta-Analysis
![Page 23: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/23.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 24: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/24.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 25: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/25.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 26: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/26.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 27: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/27.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 28: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/28.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 29: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/29.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 30: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/30.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 31: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/31.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 32: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/32.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 33: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/33.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 34: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/34.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 35: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/35.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 36: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/36.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 37: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/37.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 38: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/38.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 39: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/39.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 40: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/40.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 41: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/41.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 42: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/42.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 43: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/43.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 44: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/44.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 45: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/45.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 46: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/46.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 47: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/47.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 48: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/48.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 49: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/49.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 50: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/50.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 51: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/51.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
Secondary analysis of RCT Randomized for CRT-D or ICD, not
M / C
![Page 52: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/52.jpg)
Does Combining Studies Make Sense?
• Determination (before proceeding to math)
of whether or not it makes sense to
combine
• Otherwise known as
The Common Sense Test
![Page 53: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/53.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality
Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
M: 34.4% (21/61)
C: 27.8%(17/61)
RCTC+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
M: 40%(600/1518)
C: 34%(512/1511)
.83 (.74-.93)
RCTA
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
M: 22.1%(196/887)
C: 16.5%(146/887)
HR= .71 unadjHR = .78 adj
Retro.
Cohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
M: 11%(48/438)
C: 10%(104/1077)
HR= .72
2⁰ anal.RCT
Randomized for CRT-D or ICD, not M / C
![Page 54: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/54.jpg)
OutcomeYes
OutcomeNo
ExposureAbsent
TIME
Cohort Study
ExposurePresent
Measure HERE
![Page 55: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/55.jpg)
Does Combining Studies Make Sense?
• Determination (before proceeding to math)
of whether or not it makes sense to
combine
• Otherwise known as
The Common Sense Test
• Meta-analysis of all treatments for all heart disease?
• Meta-analysis of beta blockers for all heart disease?
• Meta-analysis of beta blockers for mortality in CHF?
• Meta-analysis of cohort studies and RCTs involving beta
blockers for mortality in CHF?
![Page 56: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/56.jpg)
We Need a Volunteer
![Page 57: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/57.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
OR0.75 (0.36, 1.58)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
OR0.78(0.68, 0.91)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
HR0.78(0.61, 1.00)
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
OR0.87(0.60, 1.25)
Secondary analysis of RCT
Randomized for CRT-D or ICD, not M / C
RESULTS FOR COMBINING ARE REPORTED BELOW:
![Page 58: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/58.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
OR0.75 (0.36, 1.58)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
OR0.78(0.68, 0.91)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
HR0.78(0.61, 1.00)
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515 who got M or C
Out of 1820 total study pop
Metoprolol succinate
Carvedilol
1. Death
OR0.87(0.60, 1.25)
Secondary analysis of RCT
Randomized for CRT-D or ICD, not M / C
RESULTS FOR COMBINING ARE REPORTED BELOW:
![Page 59: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/59.jpg)
Author/year Population Number of Subjects
Intervention and control
Outcome Measure(s) Results: Mortality Study Methodology
Grade of Evidence (A,B,C,D,F) & Why
Metra, et al2000
Italy Outpatient Ischemic or
nonischemiccardiomyopathy
Class II, III, IV ≥ 6 months
EF ≤ 35% Lasix + ACE
consistent
150
Metoprolol tartrate115±56 mg/d
Carvedilol44±17 mg/d
23±12 months
1. LVEF2. Hemodynamic
variables at rest and peak exercise, exercise tolerance, QOL, NYHA fx class, death, urgent transplantation
OR0.75 (0.36, 1.58)
RCT C+
F: 81%R: 1:1, CA?I: lost 14 in each groupS: seems to beB: yes, pts, cliniciansE: could modify if needed
Poole-Wilson, et al 2003
COMET
Europe Multicenter (341
centers) NYHA II, III, IV EF ≤ 35% diuretics + ACE
3029
Metoprolol tartrate50 mg 2x/day
Carvedilol25 mg 2x/day
58 months mean duration
1. All-cause mortality
OR0.78(0.68, 0.91)
RCT A
F: 10% loss to f/uR: permuted blocks, CAI: yesS: more CABG, ischemia MB: yesE: yes
Delea, et al2005
Constella database USA
CHF + C or M + age ≥ 35
1 outpt pharm claim for ACE + diuretic
? NYHA ?EF Death rate?
1774Metoprolol tartrate
Carvedilol
1. Death2. Hospitalization3. Death or
hospitalization
HR0.78(0.61, 1.00)
Retro-spectivecohort
B
Study sponsor helped author
Cohort differences: M older, less males, more
southern, more HTN, more arrhythmias, more renal
Ruwald et al2013
MADIT CRT
Patients enrolled in MADIT-CRT study on either M or C
Europe, US, Canada Class I, II EF ≤ 30% Choice of BB left to
treating physician
1515Metoprolol succinate
Carvedilol
1. Death
OR0.87(0.60, 1.25)
Secondary analysis of RCT
Randomized for CRT-D or ICD, not M / C
ANY OTHER DIFFERENCES TO CONSIDER?
![Page 60: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/60.jpg)
Does Combining Results Make Sense?
![Page 61: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/61.jpg)
Does Combining Results Make Sense?
![Page 62: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/62.jpg)
Does Combining Results Make Sense?
![Page 63: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/63.jpg)
Does Combining Results Make Sense?
![Page 64: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/64.jpg)
Does Combining Results Make Sense?
Statistics:
• Test for Heterogeneity– A statistical test asking if study results are more
different than would be expected by chance alone
• I2 Test– A statistical test that describes the percentage of the
variability due to heterogeneity rather than sampling
error (chance).
![Page 65: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/65.jpg)
Back to Pharmacy and Therapeutics
• Health system “nonformulary” report
includes large increase in the utilization of
carvedilol, particularly on the Gen Med
services for patients with heart failure.
• The Chair of the committee tasks you with
deciding which beta blocker will be
preferred for formulary use: Metoprolol or
Carvedilol? (Cost is now similar).
![Page 66: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/66.jpg)
![Page 67: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/67.jpg)
Systematic Reviews: Validity
1. Did the review explicitly address a
sensible clinical question?
First page, left column before METHODS
![Page 68: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/68.jpg)
Systematic Reviews: Validity
2. Was the
search for
relevant
studies
detailed and
exhaustive?
• “Publication
Bias”
• “Funnel Plots”
First page, METHODS
![Page 69: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/69.jpg)
Systematic Reviews: Validity
2. Was the
search for
relevant
studies
detailed and
exhaustive?
• “Publication
Bias”
• “Funnel Plots”
First page, METHODS
![Page 70: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/70.jpg)
Publication Bias
p.1113, above RESULTS
![Page 71: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/71.jpg)
What is a Funnel Plot Anyway?
A.K.A, “The Results”A.K
.A,
“In
cre
asin
g S
am
ple
Siz
e”
Line of Truth
Remember: The “truth is out there” but we don’t actually know what it is
![Page 72: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/72.jpg)
What is a Funnel Plot Anyway?
A.K.A, “The Results”A.K
.A,
“In
cre
asin
g S
am
ple
Siz
e”
Line of Truth
![Page 73: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/73.jpg)
Publication Bias
p.1114, bottom left column
![Page 74: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/74.jpg)
Systematic Reviews: Validity
3. Were selection
and assessment
of studies
reproducible?
First page, right column, last 3 paragraphs
![Page 75: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/75.jpg)
Data Table – p 1112
![Page 76: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/76.jpg)
Systematic Reviews: Validity
4. Did the review address possible explanations of
between-study differences in results?
Discussion section
![Page 77: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/77.jpg)
Systematic Reviews: Validity
5. Did the review present results that are ready for
clinical application?
Third page
![Page 78: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/78.jpg)
Systematic Reviews: Validity
5. Did the review provide a rating for confidence in
effect estimates or provide information needed to
evaluate the confidence?
Third page
![Page 79: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/79.jpg)
Data Table – p 1112
![Page 80: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/80.jpg)
Systematic Review: Results
1. Were the results similar from study to
study?
• Clinical assessment of heterogeneity in
population, intervention, outcomes
• “Eyeball” test in Forest Plot
• Statistical test or I2
![Page 81: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/81.jpg)
Eyeball Test for Heterogeneity
Statistical Tests for Heterogeneity
![Page 82: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/82.jpg)
Eyeball Test for Heterogeneity
Statistical Tests for Heterogeneity
![Page 83: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/83.jpg)
What are the Results?
2. What are the overall results of the
review?
– Forest plots and tables
3. How confident are you in the estimates?
– Confidence intervals
![Page 84: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/84.jpg)
![Page 85: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/85.jpg)
A = Metoprolol
B = Carvedilol
What Should the Health System Do?
![Page 86: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/86.jpg)
RECAP
![Page 87: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/87.jpg)
How can you tell if an article is a
“systematic review” rather than a
“general review” article?
A. Top journals only publish systematic reviews
B. It will cover all known information about the topic
(diagnosis, prognosis, treatment, etc)
C. It has a Methods Section
D. B and C
![Page 88: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/88.jpg)
From which types of studies is it
possible to combine data (do a
“meta-analysis”)?
A. Randomized trials only
B. Randomized trials and cohort studies
C. Randomized trials, cohort studies, and case-control
studies
D. Randomized trials, cohorts studies, case-control
studies, and case series
![Page 89: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/89.jpg)
A. Deciding whether it makes sense to combine them
based on your clinical knowledge
B. Seeing if a statistical test for heterogeneity among the
results of the studies is non-significant, or an I2 statistic
is <20%
C. Looking for overlapping confidence intervals on a forest
plot
D. A, B and C
You should assess for
heterogeneity between studies in a
systematic review by:
![Page 90: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/90.jpg)
“Weighting” refers to:
A. Eating too many snacks during your EBM workshop
B. A mathematical adjustment which makes larger studies
contribute more to the combined result than smaller
ones
C. A mathematical adjustment which makes better quality
studies contribute more to the combined result than
smaller ones
D. B or C
![Page 91: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/91.jpg)
Objectives
• Define a systematic review
• Define meta-analysis– Discuss what they can and can’t do for you
• Be able to explain:– Heterogeneity
– Weighting
– Publication Bias
• Draw and interpret a forest plot
• Critically appraise a systematic review
![Page 92: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/92.jpg)
Teaching Decisions
• Planning
• Process vs Content
• Working With What You’ve Got– (Play To Your Strengths)
– Homo ridiculousness vs. Homo seriousness
• Knowing Your Audience
• To Mark or Not to Mark? – That is the Question
• Triage
![Page 93: Summarizing the Evidence - Sites@Dukesites.duke.edu/ebmworkshop/files/2012/06/Systematic-Reviews-1.pdfSummarizing the Evidence Jane Gagliardi, MD, MHS ... Why Do We Need EBM? ... Scenario:](https://reader034.vdocument.in/reader034/viewer/2022051722/5aa009287f8b9a0d158d9be0/html5/thumbnails/93.jpg)
Thanks!
• Evaluations
• https://www.youtube.com/watch?v=FqQ-
JuRDkl8