EIP - Critical Appraisal and CrowdCARE - Week 2 Flashcards
Describe the 6 stages of the EBP process.
Assess - a question arises during patient care
Ask - completely articulate all parts of the question
Acquire - conduct a thorough focussed search and select the highest quality evidence
Appraise - evaluate the evidence for validity and clinical applicability
Apply - use clinical expertise to integrate applicable evidence with attention to patient values
Audit - self evaluation of performance
Name 3 barriers to evidence based practice.
Lack of training/skill
Lack of time
Quantity of research evidence
Define critical appraisal.
Determining the level of evidence provided by a study
Describe the evidence hierarchy in order of evidence level (6).
-Systematic review
-Randomised controlled study or prospective cohort study
-Pseudo-randomised controlled study
-Cohort or case control study
-Comparative study without controls
-Case series or cross sectional study
What is a systematic review?
It is a systematic approach to locating, appraising, and synthesising all studies that relate to a particular question. It may include a meta-analysis.
Name three requirements to a good systematic review.
-The research question and inclusion criteria are established before the conduct of the review
-There are at least two independent data extractors and a consensus procedure is used to resolve disagreements
-At least two electronic sources are searched and all searches are supplemented by consulting other sources and by reviewing the references in the studies found
Does a good systematic review use status of publication as an inclusion criteria?
No, it shouldnt exclude based on publication status or language.
Does a good systematic review provide a list of studies?
Yes, a list of included and excluded studies is provided in a good review.
How does a good systematic review treat data in included studies?
Data from the original studies is provided on the participants, interventions and outcomes.
How is the scientific quality of included studies assessed and documented?
The a priori methods of assessment must be provided in a good review
Does a good systematic review consider the scientific quality of included studies in formulating its conclusion?
Yes, the methodological rigour and scientific quality must be considered in its analysis.
How are findings of multiple studies combined in a good systematic review?
A test must be done to ensure the studies are combinable, such as an I^2 test.
How is publication bias addressed in a systematic review?
Funnel plot or Egger regression test
Do systematic reviews need a conflict of interest statement?
Yes, it must be included in both the review and if they are present in the included studies.
What kind of study is the gold standard methodology for intervention studies?
A randomised controlled trial
Define randomised controlled trial.
Participants are randomly divided into separate groups, with the intervention applied to one group, and not the other group.
The only difference between the two groups must be the intervention.
How should participants be allocated to groups in a good randomised controlled trial? What doesnt satisfy a good allocation (3)?
At random. Quasi-randomisation doesnt satisfy this criterion, such as allocation by:
-hospital record number
-birth date
-alternation
When allocations are made in a good randomised controlled trial, should it be concealed or not?
In a good trial, it will be concealed from all individuals involved in the study.
In a randomised controlled trial, how can variable severity of the condition amongst the participants be accounted for (2)?
The trial must include at least one measure of severity and at least one (different) key outcome measure at baseline.
Must participants in a good randomised controlled trial be made blind to the group they are allocated?
Does this apply to the therapist administering the therapy? What about the assessors who measure the outcome?
Yes, a good trial is blinded, both the participants and the therapist.
Assessors must also be blind to the participant’s group.
For a good randomised controlled trial, what percentage of participants must have a measured key outcome?
More than 85% of subjects initally allocated to groups must have a measure of at least one key outcome.
How must the results of between-group statistical comparisons be reported for a good randomised controlled trial?
It must be reported either in the form of a hypothesis test (p-value) or an estimate and its confidence intervals (95% or greater).
How does a good systematic review resolve duplicate study selection and data extraction?
Using at least two independent data extractors and a consensus procedure is sued to resolve disagreements
Does a good systematic review require a comprehensive literature review? Explain.
Yes, it should have at least two electronic sources searched, with the search strategy explained.
All searches should be supplemented by consulting other sources and reviewing the references in the studies found.
