How To Appraise A Clinical Trial Handout

How to Appraise a Clinical Trial
Included below are the questions to ask during Components of a Published Clinical Trial
the appraisal of a clinical trial – a necessary • Title
step during the practice of evidence-based • Authors
medicine. Alongside is displayed relevant key • Introduction: Why we did this
• Methods: What we did
definitions and concepts. • Results: What we observed
Round bullets denote medical student • Discussion: Why we think we
observed this
level questions.
Square bullets denote housestaff level
questions.
Title
Does the title clearly communicate the study’s content?
Is the title free from sensationalism?
Authors
What are the authors’ potential conflicts of interest?
Introduction
Have the authors reviewed relevant background

PICO
literature?
Pfd– Who were the patients?
Is the trial original? Ifd– What was the intervention?
Cfd– What was the control?

What clinical question is the trial addressing?
Ofd– What was the primary
(e.g. PICO)
outcome of interest?
Methods
Subjects
Who were the study subjects?
How were the subjects recruited?
What were the study’s exclusion criteria?
Sample size Study power – The probability the

study will detect a difference
How large was the sample size? between intervention and control
groups, if a true difference exists.
Power is increased by increasing
Was the desired sample size calculated ahead of time? sample size.
Control Types of controls in clinical trials
• Placebo
Was the trial controlled? • A single established
alternative
What was the control? • “Conventional” or “standard”
treatment.
Randomization
Was the trial randomized?

Computerized randomization is by
How was the randomization performed? far the most preferred method of
randomization of patients to
intervention vs. control arms.
Did the trial use allocation concealment?
Blinding
Was the trial blinded?
Single blind trial – A trial in which the subjects are unaware of whether they’ve been allocated to
intervention vs. control.
Double blind trial – A trial in which both the subjects and those tasked with assessing outcomes
are unaware of allocation.
Endpoints
What were the endpoints?
Were they objective or subjective?
Which was the primary endpoint, and which were secondary endpoints?
If objective, were they a “hard” endpoint, or a surrogate endpoint?
Was a composite endpoint used?
Primary endpoint – The main outcome of interest during the trial, which necessarily must be identified
prior to its start. The trial is powered to detect a statistically significant difference in the frequency of the
primary endpoint between the intervention and control arms. As a general rule, trials should only be
considered “positive” if there was a benefit of the intervention over the control in regards to the primary
outcome.
Secondary endpoint – Other outcomes which are measured and analyzed during the trial. Secondary
endpoints may or may not be determined prior to the trial’s start, and the trial may or may not be
adequately powered to identify differences in secondary endpoints.
Surrogate endpoint – An objective measure of treatment effect (e.g. blood pressure, LDL, HbA1c, tumor
size, etc..) that may, but does not necessary, correlate with a hard clinical endpoints (e.g. overall mortality,
disease-specific mortality, incidence of MI, etc…)
Composite endpoint – An endpoint which consists of the occurrence of any among a list of two or more
clinical events. Most composite endpoints are a combination of mortality (either overall mortality or
disease-specific mortality) and one or more less serious but more common disease complications.
Analysis
Where the results analyzed according to intention-to-treat, or per-protocol?
Is the trial a superiority trial, or a non-inferiority trial?
Intention-to-treat Analysis – All enrolled subjects are included in the analysis, and considered to belong
to the study arm to which they were initially allocated. This is almost always the preferred approach to
the analysis of results in a clinical trial.
Per-Protocol Analysis – Only those subjects who completed the study in the arm to which they were
originally assigned are included in the analysis.
Superiority trial – A clinical trial in which researchers are trying to demonstrate that an intervention
works better than the control. Underpowering biases a superiority trial away from a “positive” result.
Non-inferiority trial – A clinical trial in which researchers are trying to demonstrate that an intervention
works at least as well as the control. Underpowering biases a non-inferiority trial towards a “positive”
result.
Miscellaneous
How specifically are the methods described (i.e. Could the trial be accurately reproduced
from the paper alone?)
Were the methods used during the trial representative of common practice?
Results
How good was the randomization?
What are the results?
What was the dropout rate?

Absolute risk reduction is
Do the figures accurately represent the data? what actually matters.
Relative risk reduction is
usually irrelevant.
Are the results clinically significant?
Does the trial emphasize absolute risk reduction or relative Trials stopped early (due
risk reduction? to either notable benefit
or harm) have a tendency
to overestimate that
Was the trial stopped early? effect.
Discussion
Do the conclusions follow logically from the results?
Do the authors make comparisons to the results from similar studies?
Do the authors list limitations of their trial?
Is there evidence of systematic bias?
Selection Bias – Occurs when there is an inherent different in the types of patients who undergo
intervention vs. control, or an inherent different in the patients who participate in the study at all and
those who decline or are excluded from enrollment.
Performance Bias – Occurs when there are systematic differences in the care provided to subjects
aside from the intervention.
Detection Bias (a.k.a. measurement bias) – Occurs when there is a difference in outcome assessment
between the intervention and control arms.
Exclusion Bias – Occurs when some study subjects are excluded from analysis, despite their outcome
data being available to the authors.
Exclusion Bias – Occurs when some study subjects are excluded from analysis, despite their outcome
data being available to the authors.
Attrition Bias – Occurs when outcome data is not available because patients completely dropped out of
the study.
Is the trial valid?
Validity – An absence of systematic bias.
Internal validity – Is present when the study represents truth about the sampled population.
External validity – Is present when the study results can be generalized beyond the sampled
population.
How does the trial fit into the greater body of knowledge related to the clinical question?
▪ Does the trial reaffirm or contradict research and/or conventional wisdom on this or
related clinical questions?
▪ Have the results been duplicated by other trials?
Has any information come to light about the trial since its original publication?
By Eric Strong
Copyright © 2017
This work is made available under the terms of the Creative Commons
Attribution-NonCommerical-NoDerivs 3.0 Unported License
For details, please refer to: creativecommons.org/licenses/by-nc-nd/3.0/deed.en_US

How To Appraise A Clinical Trial Handout

Uploaded by

Copyright:

Available Formats

How To Appraise A Clinical Trial Handout

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

How To Appraise A Clinical Trial Handout

Uploaded by

Copyright:

Available Formats

How to Appraise a Clinical Trial

Does the title clearly communicate the study’s content?

Is the title free from sensationalism?

What are the authors’ potential conflicts of interest?

Have the authors reviewed relevant background

Cfd– What was the control?

Who were the study subjects?

How were the subjects recruited?

What were the study’s exclusion criteria?

Sample size Study power – The probability the

Control Types of controls in clinical trials

Was the trial randomized?

Was the trial blinded?

Were they objective or subjective?

If objective, were they a “hard” endpoint, or a surrogate endpoint?

Was a composite endpoint used?

Where the results analyzed according to intention-to-treat, or per-protocol?

Is the trial a superiority trial, or a non-inferiority trial?

How good was the randomization?

What are the results?

What was the dropout rate?

Do the conclusions follow logically from the results?

Do the authors make comparisons to the results from similar studies?

Do the authors list limitations of their trial?

Is there evidence of systematic bias?

Is the trial valid?

Validity – An absence of systematic bias.

For details, please refer to: creativecommons.org/licenses/by-nc-nd/3.0/deed.en_US

You might also like