SlideShare a Scribd company logo
1 of 50
Download to read offline
Critique of Therapy
Article – Results
Dr. Majdi N. Al-Jasim
SBFM, ABFM
Consultant Family Physician
PCFCM - AlAhsa
OBJECTIVES
To Discuss the Quantitative Measures Used in Therapy Articles:
▪ P-value
▪ Confidence Interval
▪ Event Rates
▪ Relative Risk
▪ Relative Risk Reduction
▪ Absolute Risk Reduction
▪ Number Needed to Treat or to Harm
Dr. Majdi Al-Jasim; SBFM, ABFM
Validity
Results
Apply it
Is the study done in a correct way?
(Methodology Section)
Are results significant?
(Results Section)
What are benefits and risks?
Is this study generalizable?
Randomized Controlled Trial Critique
Dr. Majdi Al-Jasim; SBFM, ABFM
Dr. Majdi Al-Jasim; SBFM, ABFM
Dr. Majdi Al-Jasim; SBFM, ABFM
P-value
▪ Used to declare Statistical Significant results.
▪ It emphasizes that how the results occur by chance.
▪ The commonly accepted cut point for calling a result
“statistically significant” is p<0.05
Example:
In a multiple choice question with 5 choices, there is a chance by
20% that you can answer the question correctly even if you did
not study well.!!!
So we can say the p-value here is 0.2
Dr. Majdi Al-Jasim; SBFM, ABFM
▪ A range of values that is almost sure to contain the true
population parameter.
▪ 95% CI means: if we repeat the study 100 times, the results in
95 studies will be within the CI range.
Example:
If 95% CI (5 – 13), then we are sure by 95% that if we repeat the
study again and again, the result will be between 5 and 13.
Confidence Interval (CI)
Dr. Majdi Al-Jasim; SBFM, ABFM
▪ A narrow range CI, means more precision; and it means a
large sample size most likely used in that study.
Example:
Confidence Interval (CI)
95% CI (3 – 25)
Wide range CI  Small sample
95% CI (0.7 – 1.2)
Narrow range CI  Large sample
Dr. Majdi Al-Jasim; SBFM, ABFM
▪ In studies that use any ratio in their outcome (like Relative
Risk, Odds Ratio), if CI range crosses the value “1”, then the
result is not statistically significant.
▪ In studies that use difference in their outcome (like weighted
mean difference), if CI range crosses the value “0”, then the
result is not statistically significant.
Confidence Interval (CI)
Dr. Majdi Al-Jasim; SBFM, ABFM
Confidence Interval (CI)
95% CI (0.2 – 0.7)
“Significant”
95% CI (0.9 – 6.7)
“Not significant”
95% CI (6.8 – 9.7)
“Significant”
Dr. Majdi Al-Jasim; SBFM, ABFM
Which of the following 95% CI represent significant results and
large sample size, assuming the result is a ratio:
 95% CI (12 – 105)
Significant but small sample size
 95% CI (0.8 – 1.2)
Not significant but large sample size
 95% CI (0.1 – 0.4)
Significant with large sample size
Confidence Interval (CI)
Dr. Majdi Al-Jasim; SBFM, ABFM
P-value & 95% CI
Dr. Majdi Al-Jasim; SBFM, ABFM
Dr. Majdi Al-Jasim; SBFM, ABFM
Event Rate
The rate of having new outcomes (i.e. incidence) in the group
who took the new treatment (Experiment Event Rate) or the
other group who took old treatment or placebo (Control Event
Rate).
Formula:
Experiment Event Rate (EER) = Events / Group Total
Control Event Rate (CER) = Events / Group Total
Dr. Majdi Al-Jasim; SBFM, ABFM
Event Rate
EER = 4/8
CER = 2/8
Dr. Majdi Al-Jasim; SBFM, ABFM
Relative Risk (RR)
The risk of having the outcome (harm or benefit) among
experiment group to those in control group.
➢ Some authors call it Risk Ratio.
➢ If it is over a period of time, it is called Hazard Ratio (HR).
Formula:
RR = Experiment Event Rate / Control Event Rate
RR = EER / CER
Dr. Majdi Al-Jasim; SBFM, ABFM
EER = 4/8
CER = 2/8
RR = EER/CER
RR = (4/8)/(2/8)
RR = 2
Relative Risk (RR)
Dr. Majdi Al-Jasim; SBFM, ABFM
Interpretation:
▪ Harmful outcome:
➢ > 1  the experiment increases the harm.
➢ < 1  the experiment decreases the harm.
➢ = 1  no difference between experiment or control group.
▪ Beneficial outcome:
➢ > 1  the experiment increases the benefit .
➢ < 1  the experiment decreases the benefit.
➢ = 1  no difference between experiment or control group.
Relative Risk (RR)
Dr. Majdi Al-Jasim; SBFM, ABFM
Kaplan Meier Hazard Function Plot
Relative Risk (RR)
Death from Cardiovascular
causes in Empagliflozin
group is 5 within 4 years.
Death from Cardiovascular
causes in placebo group is 9
within 4 years; p<0.001
Conclusion:
Empagliflozin is better than
placebo in decreasing the
death from cardiovascular
causes within 4 years.
Always…
In Hazard
Function Plot,
the lower curve
is better than
the upper curve
if the outcome is
harmful and vice
versa for
beneficial
outcome.
Dr. Majdi Al-Jasim; SBFM, ABFM
In an article, the risk of having stroke in group taking aspirin is 0.6 times
compared to the group who takes placebo; 95% CI (0.4 - 0.7).
What do you understand from the result?
Relative Risk (RR)
Dr. Majdi Al-Jasim; SBFM, ABFM
In an article, the risk of having stroke in group taking aspirin is 0.6 times
compared to the group who takes placebo; 95% CI (0.4 - 0.7).
Analysis:
RR = 0.6
We noticed that RR < 1
The outcome is stroke (harmful outcome).
So, the aspirin will decrease the stroke events.
95% CI (0.4 - 0.7) is not crossing the value “1”, so the result is statistically
significant. Also it is narrow which means large sample size used in this
study.
But for how much?
Relative Risk (RR)
Dr. Majdi Al-Jasim; SBFM, ABFM
In an article, the risk of having stroke in group taking aspirin is 0.6 times
compared to the group who takes placebo; 95% CI (0.4 - 0.7).
Here RR is 0.6 which means the risk of stroke in aspirin group is 60%
of that in control group.
In other ward, if control group has risk to develop stroke by 1000
cases in million then the aspirin group will be 60% of that (i.e. 600
cases in million). So the actual decrease is 40% (i.e. 400 cases). That
is why they came up with another measuring parameter called
Relative Risk Reduction (RRR).
Relative Risk (RR)
Dr. Majdi Al-Jasim; SBFM, ABFM
Relative Risk (RR)
▪ Relative Risk (RR) just tell if there is
increase or decrease in the outcome
in experiment group; but it doesn’t
tell for how much.
▪ The actual decrease or increase is
calculated as the rest of RR and so
here is where we need to use
Relative Risk Reduction (RRR).
Dr. Majdi Al-Jasim; SBFM, ABFM
Relative Risk Reduction (RRR)
Used to see how much the experiment treatment is
relatively reducing the chance of having outcome in
treated patient.
➢If it measures a beneficial outcome, it is called in some
articles Relative Benefit Increment (RBI).
Formula:
RRR = 1 - RR
Dr. Majdi Al-Jasim; SBFM, ABFM
Interpretation:
Using experiment treatment will relatively reduce the risk of having
the outcome by (%) compared to control treatment.
Example:
If RRR = 70% in comparing ACEI vs placebo in decreasing IHD. This
means in a person who is treated with ACEI, his chance of having IHD
will be relatively reduced by 70%.
Relative Risk Reduction (RRR)
Dr. Majdi Al-Jasim; SBFM, ABFM
In an article, the rate of developing IHD in the group using aspirin was
40%, while it was 54% in the control group.
By how much aspirin will relatively reduce the events
of IHD?
Relative Risk Reduction (RRR)
Dr. Majdi Al-Jasim; SBFM, ABFM
In an article, the rate of developing IHD in the group using aspirin was
40%, while it was 54% in the control group.
Relative Risk Reduction (RRR)
Answer:
EER = 40%
CER = 54%
RR = EER / CER = 40/54 = 0.74
RRR = 1 – RR = 1 – 0.74 = 0.26
So RRR = 0.26 = 26%
So using aspirin will relatively reduce the IHD events by 26%.
Dr. Majdi Al-Jasim; SBFM, ABFM
Relative Risk Reduction (RRR)
Same!!
Dr. Majdi Al-Jasim; SBFM, ABFM
▪ The problem with RRR is that it is bonded to RR.
▪ You can assume wrongly that the treatment is effective
in decreasing the risk of a disease whether it decreases
the risk from 4 per 1000,000 to 2 per 1000,000 or from
4 per 10 to 2 per 10 (both will have RRR = 50%).
▪ Here is where they came up with Absolute Risk
Reduction (ARR).
Relative Risk Reduction (RRR)
Dr. Majdi Al-Jasim; SBFM, ABFM
Absolute Risk Reduction (ARR)
Used to see the magnitude of benefit between experiment
treatment and control treatment.
➢ Some authors call it Attributable Risk or Risk Difference.
➢ If it measures a beneficial outcome, it is called in some articles
Absolute Benefit Increment (ABI).
Formula:
ARR = CER – EER
Dr. Majdi Al-Jasim; SBFM, ABFM
Interpretation:
if 100 patients were treated with experiment treatment, (x)
cases of outcome can be prevented; or using the experiment
treatment will absolutely reduce the outcome by (%).
Example:
If ARR = 15% in comparing ACEI vs placebo in decreasing IHD. This
means if 100 patients were treated with ACEI, 15 cases of IHD can
be prevented compared to placebo.
Absolute Risk Reduction (ARR)
Dr. Majdi Al-Jasim; SBFM, ABFM
In an article, the rate of developing IHD in the group using aspirin was
40%, while it was 54% in the control group.
By how much aspirin will absolutely reduce the
events of IHD?
Absolute Risk Reduction (ARR)
Dr. Majdi Al-Jasim; SBFM, ABFM
In an article, the rate of developing IHD in the group using aspirin was
40%, while it was 54% in the control group.
Absolute Risk Reduction (ARR)
Answer:
EER = 40%
CER = 54%
ARR = CER – EER = 54 – 40 = 14%
So using aspiring will absolutely reduce the IHD by 14%.
That is, if 100 patients took aspirin, 14 cases of IHD can be
prevented.
Have you noticed
that how ARR is
smaller than RRR?
Dr. Majdi Al-Jasim; SBFM, ABFM
Same!!
Be aware from
RRR
Dr. Majdi Al-Jasim; SBFM, ABFM
‫التجارة‬ ‫في‬ ‫غش‬ ‫يصير‬ ‫كيف‬ ‫هل‬ ‫يعني‬!
Dr. Majdi Al-Jasim; SBFM, ABFM
▪ ARR some times becomes not easy to explain for some
readers, especially if we use both ARR and ABI.
▪ Here where they came up with easy way to interpret
the results in weighting the risk versus benefit; that is
Number Needed to Treat (NNT) and Number Needed to
Harm (NNH).
Absolute Risk Reduction (ARR)
Dr. Majdi Al-Jasim; SBFM, ABFM
Widely used to weight risk versus benefit from used
treatment.
Number Needed to Treat / Harm
Benefits
Risks
Dr. Majdi Al-Jasim; SBFM, ABFM
▪ Preventing a harmful outcome:
➢ This is Number Needed to Treat (NNT). It is used to see how
many individuals needed to take the treatment in order to
prevent one bad event.
▪ Causing a harmful outcome:
➢ This is Number needed to harm (NNH). It is used to see how
many individuals needed to take the treatment in order to
develop one bad event.
Formula:
NNT or NNH = 1 / ARR
Number Needed to Treat / Harm
Dr. Majdi Al-Jasim; SBFM, ABFM
In an article, the rate of developing IHD in the group using aspirin was
40%, while it was 54% in the control group.
Number Needed to Treat / Harm
How many patients we need to treat with aspirin in
order to prevent one case of IHD?
Dr. Majdi Al-Jasim; SBFM, ABFM
Answer:
EER = 40%
CER = 54%
ARR = CER – EER = (54 – 40) = 14%
So ARR = 14% = 0.14
NNT = 1 / ARR
NNT = 1 / 0.14
NNT = 7.14
So we need to treat 7 patients with aspirin in order to prevent
one IHD event.
Number Needed to Treat / Harm
Calculation tip:
If you use ARR in (%), then
calculate NNT as
following:
NNT = 100 / ARR (%)
Dr. Majdi Al-Jasim; SBFM, ABFM
In the same article, the rate of developing hemorrhagic stroke in the group
using aspirin was 35%, while it was 19% in the control group.
Number Needed to Treat / Harm
How many patients we need to treat with aspirin in
order to cause one case of hemorrhagic stroke?
Dr. Majdi Al-Jasim; SBFM, ABFM
Answer:
EER = 35%
CER = 19%
ARR = EER – CER = (35 – 19) = 16%
So ARR = 16% = 0.16
NNT = 1 / ARR
NNT = 1 / 0.16
NNH = 6.25
So we need to treat 6 patients with aspirin in order to cause one
hemorrhagic stroke event.
Number Needed to Treat / Harm
Calculation tip:
If you use ARR in (%), then
calculate NNH as
following:
NNH = 100 / ARR (%)
Dr. Majdi Al-Jasim; SBFM, ABFM
Is it NNT or NNH that we need to use in this question?
Number Needed to Treat / Harm
Dr. Majdi Al-Jasim; SBFM, ABFM
Is it NNT or NNH that we need to use in this question?
If the bad events rate in control group are more than that in experiment
group (i.e. CER > EER), then it is NNT and vice versa for good events.
Example:
Number Needed to Treat / Harm
Bad
events
EER = 10.5%
CER = 12.1%
The experiment treatment causes less harmDr. Majdi Al-Jasim; SBFM, ABFM
Is it NNT or NNH that we need to use in this question?
If the bad events rate in experiment group are more than that in control
group (i.e. EER > CER), then it is NNH and vice versa for good events.
Example:
Number Needed to Treat / Harm
Bad
events
EER = 38.3%
CER = 37.1%
The experiment treatment causes more harm
Dr. Majdi Al-Jasim; SBFM, ABFM
Summary
Check the significance and how the results happen by
chance
Compare the risk in experiment versus control group
How much the risk is relatively reduced in treated
patient
Absolute reduced risk in 100 treated patients
Dr. Majdi Al-Jasim; SBFM, ABFM
The number needed to prevent one harm event
The number needed to cause one harm event
We are certain that the repeated results will be within
this range
Summary
Dr. Majdi Al-Jasim; SBFM, ABFM
Dr. Majdi Al-Jasim; SBFM, ABFM
For more EBM interactive lectures, I highly recommend
you visiting Prof Terry Shaneyfelt YouTube Channel:
https://www.youtube.com/user/UABEBMcourse/videos
Yours;
Dr. Majdi Al-Jasim

More Related Content

What's hot

How to read a forest plot?
How to read a forest plot?How to read a forest plot?
How to read a forest plot?
Samir Haffar
 
Study Designs_YL
Study Designs_YLStudy Designs_YL
Study Designs_YL
Yvonne Lee
 

What's hot (20)

4. Calculate samplesize for cross-sectional studies
4. Calculate samplesize for cross-sectional studies4. Calculate samplesize for cross-sectional studies
4. Calculate samplesize for cross-sectional studies
 
PICO Research Question
PICO Research QuestionPICO Research Question
PICO Research Question
 
Odds ratio
Odds ratioOdds ratio
Odds ratio
 
Cohort Study
Cohort Study Cohort Study
Cohort Study
 
How to read a forest plot?
How to read a forest plot?How to read a forest plot?
How to read a forest plot?
 
Methods of Randomization
Methods of RandomizationMethods of Randomization
Methods of Randomization
 
CASE CONTROL STUDY
CASE CONTROL STUDYCASE CONTROL STUDY
CASE CONTROL STUDY
 
Odds ratios (Basic concepts)
Odds ratios (Basic concepts)Odds ratios (Basic concepts)
Odds ratios (Basic concepts)
 
Type of randomization
Type of randomizationType of randomization
Type of randomization
 
Screening and diagnostic tests
Screening and diagnostic testsScreening and diagnostic tests
Screening and diagnostic tests
 
The odds ratio
The odds ratioThe odds ratio
The odds ratio
 
Estimating risk
Estimating riskEstimating risk
Estimating risk
 
Randomized Controlled Trial
Randomized Controlled TrialRandomized Controlled Trial
Randomized Controlled Trial
 
Sample size calculation - a brief overview
Sample size calculation - a brief overviewSample size calculation - a brief overview
Sample size calculation - a brief overview
 
Bias and confounding
Bias and confoundingBias and confounding
Bias and confounding
 
6. Calculate samplesize for cohort studies
6. Calculate samplesize for cohort studies6. Calculate samplesize for cohort studies
6. Calculate samplesize for cohort studies
 
Study Designs_YL
Study Designs_YLStudy Designs_YL
Study Designs_YL
 
Evaluating a diagnostic test presentation www.eyenirvaan.com - part 1
Evaluating a diagnostic test presentation www.eyenirvaan.com - part 1Evaluating a diagnostic test presentation www.eyenirvaan.com - part 1
Evaluating a diagnostic test presentation www.eyenirvaan.com - part 1
 
Bias in Research
Bias in ResearchBias in Research
Bias in Research
 
Case control study
Case control studyCase control study
Case control study
 

Similar to RCT Critical Appraisal - Results

Consider the following hypothet-ical scenario and results .docx
Consider the following hypothet-ical scenario and results .docxConsider the following hypothet-ical scenario and results .docx
Consider the following hypothet-ical scenario and results .docx
donnajames55
 
Why to know statistics
Why to know statisticsWhy to know statistics
Why to know statistics
Hesham Gaber
 
Common measures and statistics in epidemiological literature
Common measures and statistics in epidemiological literatureCommon measures and statistics in epidemiological literature
Common measures and statistics in epidemiological literature
Kadium
 
Statistics tests and Probablity
Statistics tests and ProbablityStatistics tests and Probablity
Statistics tests and Probablity
Abdul Wasay Baloch
 

Similar to RCT Critical Appraisal - Results (20)

Why to know statistics
Why to know statisticsWhy to know statistics
Why to know statistics
 
Risk assessment
Risk assessmentRisk assessment
Risk assessment
 
Consider the following hypothet-ical scenario and results .docx
Consider the following hypothet-ical scenario and results .docxConsider the following hypothet-ical scenario and results .docx
Consider the following hypothet-ical scenario and results .docx
 
Why to know statistics
Why to know statisticsWhy to know statistics
Why to know statistics
 
Single sample z test - explain (final)
Single sample z test - explain (final)Single sample z test - explain (final)
Single sample z test - explain (final)
 
Can Personalized Medicine Save the Health Care System?
Can Personalized Medicine Save the Health Care System?Can Personalized Medicine Save the Health Care System?
Can Personalized Medicine Save the Health Care System?
 
Quantitative critical appraisal october 2015
Quantitative critical appraisal october 2015Quantitative critical appraisal october 2015
Quantitative critical appraisal october 2015
 
NNT: Number Needed to Treat
NNT: Number Needed to TreatNNT: Number Needed to Treat
NNT: Number Needed to Treat
 
Biostatistics
BiostatisticsBiostatistics
Biostatistics
 
Common measures of association in medical research (UPDATED) 2013
Common measures of association in medical research (UPDATED) 2013Common measures of association in medical research (UPDATED) 2013
Common measures of association in medical research (UPDATED) 2013
 
Risky Business: Risk communicat ion in the provider-patient encounter
Risky Business: Risk communicat ion in the provider-patient encounterRisky Business: Risk communicat ion in the provider-patient encounter
Risky Business: Risk communicat ion in the provider-patient encounter
 
Critical Appriaisal Skills Basic 1 | May 4th 2011
Critical Appriaisal Skills Basic 1 | May 4th 2011Critical Appriaisal Skills Basic 1 | May 4th 2011
Critical Appriaisal Skills Basic 1 | May 4th 2011
 
Common measures and statistics in epidemiological literature
Common measures and statistics in epidemiological literatureCommon measures and statistics in epidemiological literature
Common measures and statistics in epidemiological literature
 
What is a Single Sample Z Test?
What is a Single Sample Z Test?What is a Single Sample Z Test?
What is a Single Sample Z Test?
 
2014 lab slides_mo_a
2014 lab slides_mo_a2014 lab slides_mo_a
2014 lab slides_mo_a
 
Statistics tests and Probablity
Statistics tests and ProbablityStatistics tests and Probablity
Statistics tests and Probablity
 
ODDS RATIO AND RELATIVE RISK EVALUATION
ODDS RATIO AND RELATIVE RISK EVALUATIONODDS RATIO AND RELATIVE RISK EVALUATION
ODDS RATIO AND RELATIVE RISK EVALUATION
 
AHCJ 2012 Atlanta conf. talk
AHCJ 2012 Atlanta conf. talkAHCJ 2012 Atlanta conf. talk
AHCJ 2012 Atlanta conf. talk
 
Understanding the evidence in pharmacoepidemiology study
Understanding the evidence in pharmacoepidemiology studyUnderstanding the evidence in pharmacoepidemiology study
Understanding the evidence in pharmacoepidemiology study
 
How to do the maths
How to do the mathsHow to do the maths
How to do the maths
 

More from Dr. Majdi Al Jasim

More from Dr. Majdi Al Jasim (12)

Growth Charts Interpretation DrMajdi
Growth Charts Interpretation DrMajdiGrowth Charts Interpretation DrMajdi
Growth Charts Interpretation DrMajdi
 
Insulin therapy in primary health care DrMajdi
Insulin therapy in primary health care DrMajdiInsulin therapy in primary health care DrMajdi
Insulin therapy in primary health care DrMajdi
 
DM Complications DrMajdi
DM Complications DrMajdiDM Complications DrMajdi
DM Complications DrMajdi
 
Research literature review dr majdi
Research literature review dr majdiResearch literature review dr majdi
Research literature review dr majdi
 
Critical Appraisal of systematic review and meta analysis articles
Critical Appraisal of systematic review and meta analysis articlesCritical Appraisal of systematic review and meta analysis articles
Critical Appraisal of systematic review and meta analysis articles
 
Practice managment dr majdi
Practice managment dr majdiPractice managment dr majdi
Practice managment dr majdi
 
Spss series - data entry and coding
Spss series - data entry and codingSpss series - data entry and coding
Spss series - data entry and coding
 
Questionnaire data collection tool dr majdi
Questionnaire data collection tool dr majdiQuestionnaire data collection tool dr majdi
Questionnaire data collection tool dr majdi
 
Selecting research topic dr majdi
Selecting research topic dr majdiSelecting research topic dr majdi
Selecting research topic dr majdi
 
Research inclusion and exclusion criteria DrMajdi
Research inclusion and exclusion criteria DrMajdiResearch inclusion and exclusion criteria DrMajdi
Research inclusion and exclusion criteria DrMajdi
 
Analyzing data in health care Dr.Majdi
Analyzing data in health care Dr.MajdiAnalyzing data in health care Dr.Majdi
Analyzing data in health care Dr.Majdi
 
سبل الوقاية من فيروس كورونا
سبل الوقاية من فيروس كوروناسبل الوقاية من فيروس كورونا
سبل الوقاية من فيروس كورونا
 

Recently uploaded

Spellings Wk 4 and Wk 5 for Grade 4 at CAPS
Spellings Wk 4 and Wk 5 for Grade 4 at CAPSSpellings Wk 4 and Wk 5 for Grade 4 at CAPS
Spellings Wk 4 and Wk 5 for Grade 4 at CAPS
AnaAcapella
 

Recently uploaded (20)

REMIFENTANIL: An Ultra short acting opioid.pptx
REMIFENTANIL: An Ultra short acting opioid.pptxREMIFENTANIL: An Ultra short acting opioid.pptx
REMIFENTANIL: An Ultra short acting opioid.pptx
 
Spellings Wk 4 and Wk 5 for Grade 4 at CAPS
Spellings Wk 4 and Wk 5 for Grade 4 at CAPSSpellings Wk 4 and Wk 5 for Grade 4 at CAPS
Spellings Wk 4 and Wk 5 for Grade 4 at CAPS
 
Tatlong Kwento ni Lola basyang-1.pdf arts
Tatlong Kwento ni Lola basyang-1.pdf artsTatlong Kwento ni Lola basyang-1.pdf arts
Tatlong Kwento ni Lola basyang-1.pdf arts
 
How to Add a Tool Tip to a Field in Odoo 17
How to Add a Tool Tip to a Field in Odoo 17How to Add a Tool Tip to a Field in Odoo 17
How to Add a Tool Tip to a Field in Odoo 17
 
NO1 Top Black Magic Specialist In Lahore Black magic In Pakistan Kala Ilam Ex...
NO1 Top Black Magic Specialist In Lahore Black magic In Pakistan Kala Ilam Ex...NO1 Top Black Magic Specialist In Lahore Black magic In Pakistan Kala Ilam Ex...
NO1 Top Black Magic Specialist In Lahore Black magic In Pakistan Kala Ilam Ex...
 
FSB Advising Checklist - Orientation 2024
FSB Advising Checklist - Orientation 2024FSB Advising Checklist - Orientation 2024
FSB Advising Checklist - Orientation 2024
 
How to Manage Call for Tendor in Odoo 17
How to Manage Call for Tendor in Odoo 17How to Manage Call for Tendor in Odoo 17
How to Manage Call for Tendor in Odoo 17
 
VAMOS CUIDAR DO NOSSO PLANETA! .
VAMOS CUIDAR DO NOSSO PLANETA!                    .VAMOS CUIDAR DO NOSSO PLANETA!                    .
VAMOS CUIDAR DO NOSSO PLANETA! .
 
Unit 3 Emotional Intelligence and Spiritual Intelligence.pdf
Unit 3 Emotional Intelligence and Spiritual Intelligence.pdfUnit 3 Emotional Intelligence and Spiritual Intelligence.pdf
Unit 3 Emotional Intelligence and Spiritual Intelligence.pdf
 
Jamworks pilot and AI at Jisc (20/03/2024)
Jamworks pilot and AI at Jisc (20/03/2024)Jamworks pilot and AI at Jisc (20/03/2024)
Jamworks pilot and AI at Jisc (20/03/2024)
 
Sensory_Experience_and_Emotional_Resonance_in_Gabriel_Okaras_The_Piano_and_Th...
Sensory_Experience_and_Emotional_Resonance_in_Gabriel_Okaras_The_Piano_and_Th...Sensory_Experience_and_Emotional_Resonance_in_Gabriel_Okaras_The_Piano_and_Th...
Sensory_Experience_and_Emotional_Resonance_in_Gabriel_Okaras_The_Piano_and_Th...
 
HMCS Max Bernays Pre-Deployment Brief (May 2024).pptx
HMCS Max Bernays Pre-Deployment Brief (May 2024).pptxHMCS Max Bernays Pre-Deployment Brief (May 2024).pptx
HMCS Max Bernays Pre-Deployment Brief (May 2024).pptx
 
How to setup Pycharm environment for Odoo 17.pptx
How to setup Pycharm environment for Odoo 17.pptxHow to setup Pycharm environment for Odoo 17.pptx
How to setup Pycharm environment for Odoo 17.pptx
 
Economic Importance Of Fungi In Food Additives
Economic Importance Of Fungi In Food AdditivesEconomic Importance Of Fungi In Food Additives
Economic Importance Of Fungi In Food Additives
 
What is 3 Way Matching Process in Odoo 17.pptx
What is 3 Way Matching Process in Odoo 17.pptxWhat is 3 Way Matching Process in Odoo 17.pptx
What is 3 Way Matching Process in Odoo 17.pptx
 
Exploring_the_Narrative_Style_of_Amitav_Ghoshs_Gun_Island.pptx
Exploring_the_Narrative_Style_of_Amitav_Ghoshs_Gun_Island.pptxExploring_the_Narrative_Style_of_Amitav_Ghoshs_Gun_Island.pptx
Exploring_the_Narrative_Style_of_Amitav_Ghoshs_Gun_Island.pptx
 
AIM of Education-Teachers Training-2024.ppt
AIM of Education-Teachers Training-2024.pptAIM of Education-Teachers Training-2024.ppt
AIM of Education-Teachers Training-2024.ppt
 
FICTIONAL SALESMAN/SALESMAN SNSW 2024.pdf
FICTIONAL SALESMAN/SALESMAN SNSW 2024.pdfFICTIONAL SALESMAN/SALESMAN SNSW 2024.pdf
FICTIONAL SALESMAN/SALESMAN SNSW 2024.pdf
 
Accessible Digital Futures project (20/03/2024)
Accessible Digital Futures project (20/03/2024)Accessible Digital Futures project (20/03/2024)
Accessible Digital Futures project (20/03/2024)
 
Wellbeing inclusion and digital dystopias.pptx
Wellbeing inclusion and digital dystopias.pptxWellbeing inclusion and digital dystopias.pptx
Wellbeing inclusion and digital dystopias.pptx
 

RCT Critical Appraisal - Results

  • 1. Critique of Therapy Article – Results Dr. Majdi N. Al-Jasim SBFM, ABFM Consultant Family Physician PCFCM - AlAhsa
  • 2. OBJECTIVES To Discuss the Quantitative Measures Used in Therapy Articles: ▪ P-value ▪ Confidence Interval ▪ Event Rates ▪ Relative Risk ▪ Relative Risk Reduction ▪ Absolute Risk Reduction ▪ Number Needed to Treat or to Harm Dr. Majdi Al-Jasim; SBFM, ABFM
  • 3. Validity Results Apply it Is the study done in a correct way? (Methodology Section) Are results significant? (Results Section) What are benefits and risks? Is this study generalizable? Randomized Controlled Trial Critique Dr. Majdi Al-Jasim; SBFM, ABFM
  • 4. Dr. Majdi Al-Jasim; SBFM, ABFM
  • 5. Dr. Majdi Al-Jasim; SBFM, ABFM
  • 6. P-value ▪ Used to declare Statistical Significant results. ▪ It emphasizes that how the results occur by chance. ▪ The commonly accepted cut point for calling a result “statistically significant” is p<0.05 Example: In a multiple choice question with 5 choices, there is a chance by 20% that you can answer the question correctly even if you did not study well.!!! So we can say the p-value here is 0.2 Dr. Majdi Al-Jasim; SBFM, ABFM
  • 7. ▪ A range of values that is almost sure to contain the true population parameter. ▪ 95% CI means: if we repeat the study 100 times, the results in 95 studies will be within the CI range. Example: If 95% CI (5 – 13), then we are sure by 95% that if we repeat the study again and again, the result will be between 5 and 13. Confidence Interval (CI) Dr. Majdi Al-Jasim; SBFM, ABFM
  • 8. ▪ A narrow range CI, means more precision; and it means a large sample size most likely used in that study. Example: Confidence Interval (CI) 95% CI (3 – 25) Wide range CI  Small sample 95% CI (0.7 – 1.2) Narrow range CI  Large sample Dr. Majdi Al-Jasim; SBFM, ABFM
  • 9. ▪ In studies that use any ratio in their outcome (like Relative Risk, Odds Ratio), if CI range crosses the value “1”, then the result is not statistically significant. ▪ In studies that use difference in their outcome (like weighted mean difference), if CI range crosses the value “0”, then the result is not statistically significant. Confidence Interval (CI) Dr. Majdi Al-Jasim; SBFM, ABFM
  • 10. Confidence Interval (CI) 95% CI (0.2 – 0.7) “Significant” 95% CI (0.9 – 6.7) “Not significant” 95% CI (6.8 – 9.7) “Significant” Dr. Majdi Al-Jasim; SBFM, ABFM
  • 11. Which of the following 95% CI represent significant results and large sample size, assuming the result is a ratio:  95% CI (12 – 105) Significant but small sample size  95% CI (0.8 – 1.2) Not significant but large sample size  95% CI (0.1 – 0.4) Significant with large sample size Confidence Interval (CI) Dr. Majdi Al-Jasim; SBFM, ABFM
  • 12. P-value & 95% CI Dr. Majdi Al-Jasim; SBFM, ABFM
  • 13. Dr. Majdi Al-Jasim; SBFM, ABFM
  • 14. Event Rate The rate of having new outcomes (i.e. incidence) in the group who took the new treatment (Experiment Event Rate) or the other group who took old treatment or placebo (Control Event Rate). Formula: Experiment Event Rate (EER) = Events / Group Total Control Event Rate (CER) = Events / Group Total Dr. Majdi Al-Jasim; SBFM, ABFM
  • 15. Event Rate EER = 4/8 CER = 2/8 Dr. Majdi Al-Jasim; SBFM, ABFM
  • 16. Relative Risk (RR) The risk of having the outcome (harm or benefit) among experiment group to those in control group. ➢ Some authors call it Risk Ratio. ➢ If it is over a period of time, it is called Hazard Ratio (HR). Formula: RR = Experiment Event Rate / Control Event Rate RR = EER / CER Dr. Majdi Al-Jasim; SBFM, ABFM
  • 17. EER = 4/8 CER = 2/8 RR = EER/CER RR = (4/8)/(2/8) RR = 2 Relative Risk (RR) Dr. Majdi Al-Jasim; SBFM, ABFM
  • 18. Interpretation: ▪ Harmful outcome: ➢ > 1  the experiment increases the harm. ➢ < 1  the experiment decreases the harm. ➢ = 1  no difference between experiment or control group. ▪ Beneficial outcome: ➢ > 1  the experiment increases the benefit . ➢ < 1  the experiment decreases the benefit. ➢ = 1  no difference between experiment or control group. Relative Risk (RR) Dr. Majdi Al-Jasim; SBFM, ABFM
  • 19. Kaplan Meier Hazard Function Plot Relative Risk (RR) Death from Cardiovascular causes in Empagliflozin group is 5 within 4 years. Death from Cardiovascular causes in placebo group is 9 within 4 years; p<0.001 Conclusion: Empagliflozin is better than placebo in decreasing the death from cardiovascular causes within 4 years. Always… In Hazard Function Plot, the lower curve is better than the upper curve if the outcome is harmful and vice versa for beneficial outcome. Dr. Majdi Al-Jasim; SBFM, ABFM
  • 20. In an article, the risk of having stroke in group taking aspirin is 0.6 times compared to the group who takes placebo; 95% CI (0.4 - 0.7). What do you understand from the result? Relative Risk (RR) Dr. Majdi Al-Jasim; SBFM, ABFM
  • 21. In an article, the risk of having stroke in group taking aspirin is 0.6 times compared to the group who takes placebo; 95% CI (0.4 - 0.7). Analysis: RR = 0.6 We noticed that RR < 1 The outcome is stroke (harmful outcome). So, the aspirin will decrease the stroke events. 95% CI (0.4 - 0.7) is not crossing the value “1”, so the result is statistically significant. Also it is narrow which means large sample size used in this study. But for how much? Relative Risk (RR) Dr. Majdi Al-Jasim; SBFM, ABFM
  • 22. In an article, the risk of having stroke in group taking aspirin is 0.6 times compared to the group who takes placebo; 95% CI (0.4 - 0.7). Here RR is 0.6 which means the risk of stroke in aspirin group is 60% of that in control group. In other ward, if control group has risk to develop stroke by 1000 cases in million then the aspirin group will be 60% of that (i.e. 600 cases in million). So the actual decrease is 40% (i.e. 400 cases). That is why they came up with another measuring parameter called Relative Risk Reduction (RRR). Relative Risk (RR) Dr. Majdi Al-Jasim; SBFM, ABFM
  • 23. Relative Risk (RR) ▪ Relative Risk (RR) just tell if there is increase or decrease in the outcome in experiment group; but it doesn’t tell for how much. ▪ The actual decrease or increase is calculated as the rest of RR and so here is where we need to use Relative Risk Reduction (RRR). Dr. Majdi Al-Jasim; SBFM, ABFM
  • 24. Relative Risk Reduction (RRR) Used to see how much the experiment treatment is relatively reducing the chance of having outcome in treated patient. ➢If it measures a beneficial outcome, it is called in some articles Relative Benefit Increment (RBI). Formula: RRR = 1 - RR Dr. Majdi Al-Jasim; SBFM, ABFM
  • 25. Interpretation: Using experiment treatment will relatively reduce the risk of having the outcome by (%) compared to control treatment. Example: If RRR = 70% in comparing ACEI vs placebo in decreasing IHD. This means in a person who is treated with ACEI, his chance of having IHD will be relatively reduced by 70%. Relative Risk Reduction (RRR) Dr. Majdi Al-Jasim; SBFM, ABFM
  • 26. In an article, the rate of developing IHD in the group using aspirin was 40%, while it was 54% in the control group. By how much aspirin will relatively reduce the events of IHD? Relative Risk Reduction (RRR) Dr. Majdi Al-Jasim; SBFM, ABFM
  • 27. In an article, the rate of developing IHD in the group using aspirin was 40%, while it was 54% in the control group. Relative Risk Reduction (RRR) Answer: EER = 40% CER = 54% RR = EER / CER = 40/54 = 0.74 RRR = 1 – RR = 1 – 0.74 = 0.26 So RRR = 0.26 = 26% So using aspirin will relatively reduce the IHD events by 26%. Dr. Majdi Al-Jasim; SBFM, ABFM
  • 28. Relative Risk Reduction (RRR) Same!! Dr. Majdi Al-Jasim; SBFM, ABFM
  • 29. ▪ The problem with RRR is that it is bonded to RR. ▪ You can assume wrongly that the treatment is effective in decreasing the risk of a disease whether it decreases the risk from 4 per 1000,000 to 2 per 1000,000 or from 4 per 10 to 2 per 10 (both will have RRR = 50%). ▪ Here is where they came up with Absolute Risk Reduction (ARR). Relative Risk Reduction (RRR) Dr. Majdi Al-Jasim; SBFM, ABFM
  • 30. Absolute Risk Reduction (ARR) Used to see the magnitude of benefit between experiment treatment and control treatment. ➢ Some authors call it Attributable Risk or Risk Difference. ➢ If it measures a beneficial outcome, it is called in some articles Absolute Benefit Increment (ABI). Formula: ARR = CER – EER Dr. Majdi Al-Jasim; SBFM, ABFM
  • 31. Interpretation: if 100 patients were treated with experiment treatment, (x) cases of outcome can be prevented; or using the experiment treatment will absolutely reduce the outcome by (%). Example: If ARR = 15% in comparing ACEI vs placebo in decreasing IHD. This means if 100 patients were treated with ACEI, 15 cases of IHD can be prevented compared to placebo. Absolute Risk Reduction (ARR) Dr. Majdi Al-Jasim; SBFM, ABFM
  • 32. In an article, the rate of developing IHD in the group using aspirin was 40%, while it was 54% in the control group. By how much aspirin will absolutely reduce the events of IHD? Absolute Risk Reduction (ARR) Dr. Majdi Al-Jasim; SBFM, ABFM
  • 33. In an article, the rate of developing IHD in the group using aspirin was 40%, while it was 54% in the control group. Absolute Risk Reduction (ARR) Answer: EER = 40% CER = 54% ARR = CER – EER = 54 – 40 = 14% So using aspiring will absolutely reduce the IHD by 14%. That is, if 100 patients took aspirin, 14 cases of IHD can be prevented. Have you noticed that how ARR is smaller than RRR? Dr. Majdi Al-Jasim; SBFM, ABFM
  • 34. Same!! Be aware from RRR Dr. Majdi Al-Jasim; SBFM, ABFM
  • 35.
  • 36. ‫التجارة‬ ‫في‬ ‫غش‬ ‫يصير‬ ‫كيف‬ ‫هل‬ ‫يعني‬! Dr. Majdi Al-Jasim; SBFM, ABFM
  • 37. ▪ ARR some times becomes not easy to explain for some readers, especially if we use both ARR and ABI. ▪ Here where they came up with easy way to interpret the results in weighting the risk versus benefit; that is Number Needed to Treat (NNT) and Number Needed to Harm (NNH). Absolute Risk Reduction (ARR) Dr. Majdi Al-Jasim; SBFM, ABFM
  • 38. Widely used to weight risk versus benefit from used treatment. Number Needed to Treat / Harm Benefits Risks Dr. Majdi Al-Jasim; SBFM, ABFM
  • 39. ▪ Preventing a harmful outcome: ➢ This is Number Needed to Treat (NNT). It is used to see how many individuals needed to take the treatment in order to prevent one bad event. ▪ Causing a harmful outcome: ➢ This is Number needed to harm (NNH). It is used to see how many individuals needed to take the treatment in order to develop one bad event. Formula: NNT or NNH = 1 / ARR Number Needed to Treat / Harm Dr. Majdi Al-Jasim; SBFM, ABFM
  • 40. In an article, the rate of developing IHD in the group using aspirin was 40%, while it was 54% in the control group. Number Needed to Treat / Harm How many patients we need to treat with aspirin in order to prevent one case of IHD? Dr. Majdi Al-Jasim; SBFM, ABFM
  • 41. Answer: EER = 40% CER = 54% ARR = CER – EER = (54 – 40) = 14% So ARR = 14% = 0.14 NNT = 1 / ARR NNT = 1 / 0.14 NNT = 7.14 So we need to treat 7 patients with aspirin in order to prevent one IHD event. Number Needed to Treat / Harm Calculation tip: If you use ARR in (%), then calculate NNT as following: NNT = 100 / ARR (%) Dr. Majdi Al-Jasim; SBFM, ABFM
  • 42. In the same article, the rate of developing hemorrhagic stroke in the group using aspirin was 35%, while it was 19% in the control group. Number Needed to Treat / Harm How many patients we need to treat with aspirin in order to cause one case of hemorrhagic stroke? Dr. Majdi Al-Jasim; SBFM, ABFM
  • 43. Answer: EER = 35% CER = 19% ARR = EER – CER = (35 – 19) = 16% So ARR = 16% = 0.16 NNT = 1 / ARR NNT = 1 / 0.16 NNH = 6.25 So we need to treat 6 patients with aspirin in order to cause one hemorrhagic stroke event. Number Needed to Treat / Harm Calculation tip: If you use ARR in (%), then calculate NNH as following: NNH = 100 / ARR (%) Dr. Majdi Al-Jasim; SBFM, ABFM
  • 44. Is it NNT or NNH that we need to use in this question? Number Needed to Treat / Harm Dr. Majdi Al-Jasim; SBFM, ABFM
  • 45. Is it NNT or NNH that we need to use in this question? If the bad events rate in control group are more than that in experiment group (i.e. CER > EER), then it is NNT and vice versa for good events. Example: Number Needed to Treat / Harm Bad events EER = 10.5% CER = 12.1% The experiment treatment causes less harmDr. Majdi Al-Jasim; SBFM, ABFM
  • 46. Is it NNT or NNH that we need to use in this question? If the bad events rate in experiment group are more than that in control group (i.e. EER > CER), then it is NNH and vice versa for good events. Example: Number Needed to Treat / Harm Bad events EER = 38.3% CER = 37.1% The experiment treatment causes more harm Dr. Majdi Al-Jasim; SBFM, ABFM
  • 47. Summary Check the significance and how the results happen by chance Compare the risk in experiment versus control group How much the risk is relatively reduced in treated patient Absolute reduced risk in 100 treated patients Dr. Majdi Al-Jasim; SBFM, ABFM
  • 48. The number needed to prevent one harm event The number needed to cause one harm event We are certain that the repeated results will be within this range Summary Dr. Majdi Al-Jasim; SBFM, ABFM
  • 49. Dr. Majdi Al-Jasim; SBFM, ABFM
  • 50. For more EBM interactive lectures, I highly recommend you visiting Prof Terry Shaneyfelt YouTube Channel: https://www.youtube.com/user/UABEBMcourse/videos Yours; Dr. Majdi Al-Jasim