Lecture 11 - Secondary Research Evidence: An Overview Flashcards
Why do we use secondary evidence?
• Secondary evidence (constructed well) attenuates the biases incurred from single primary studies
– Systematic literature reviews and Meta- analyses use a non-biased, transparent approach to collating and evaluating evidence from primary research
– Secondary evidence produces summary findings ‘on balance’
Literature Review
- A literature review is both a summary and explanation of the complete and current state of knowledge on a topic
- Most deal with a broad range of issues related to a given topic rather than addressing a particular issue in depth
Purpose of the Literature Review
- Easy access to research on a particular topic by summarizing articles or studies that are relevant, meaningful and important into one complete report
- Excellent starting point for researchers beginning to do research in a new area
- Ensures researchers don’t duplicate work
- Provides clues as to where future research is heading or recommend areas on which to focus.
- Highlights key findings and may be useful for background reading and gaining a superficial understanding
- It identifies inconsistencies, gaps and contradictions in the literature
- May provide constructive analysis of the methodologies and approaches of other researchers
Failing of the literature review
• Lack of rigour in the systematic review process
• Lack of clarity on how studies were identified, selected and integrated
• Selective use and reporting of literature to suit the needs of the author(s)
– May be biased
• Despite these limitations, literature reviews continue to be popular
-can’t use literature reviews for foreground qns
Systematic Reviews
• Have replaced many traditional literature reviews as a way of summarizing research evidence
• Must have the same level of rigour to reviewing research evidence as should be used in producing that research evidence in the first place
• Predominantly focused on quantitative research evidence
– Increasingly utilize also qualitative research evidence
• Systematic & comprehensive
– Methods to reduce and ultimately eliminate bias
– Transparent, rigorous and reproducible process
– Integrate / synthesise findings from individual studies
• Higher grade of information
– likely to provide stronger evidence
• Reveal new results or generate new ideas for research
Why do we need systematic reviews?
– Increasing quantity and quality of research evidence
– Increasing access to research evidence • Open access journals
– Increasing need for high level research evidence to inform health care decisions
• A range of health care stakeholders
– Unclear, confusing or contradictory research evidence when viewed individually
• Synthesized evidence may provide a clearer picture
Definition of systematic review
• Systematic reviews, are a form of literature review which is focused on a single issue, and which attempts, through a set procedure, to thoroughly identify, appraise, select and synthesise, the best possible evidence (Cochrane, 2009).
What type of questions do systematic reviews answer?
• Answer specific, often narrow, clinical questions in depth.
– Specific population and setting – Condition of interest
– Exposure to a test or treatment – Specific outcomes
• Very similar to the PICO framework
Steps in a Systematic Review
• Formulate an appropriate review question – A clear and focused question
• Generate protocol (the ‘recipe’ for the review) – A transparent and upfront methodology
• Search for evidence (collate ‘data’)
– Looking for studies to inform the review question
– Databases and other sources of research (such as organizations)
– Grey literature (unpublished research)
• Include studies relevant to the research question
• Critically appraise the included studies Process
– Methodological quality of the study
• Extract data from included studies
– Reported findings relevant to the review question are
extracted
– Data extraction form
• Synthesise the data
– Aggregation of evidence
– Meta-analysis/meta-synthesis
• Summarise the evidence
– Placing the findings in context (eg: quality of evidence base)
• Provide an answer to the clinical question
Meta-analyses
• One product of a systematic review
• Meta-analyses combine the data of several studies that address a set of related research hypotheses
– What did each study find?
• By doing so you can gain the true effect of the treatment (effect size)
– Combining a number of research studies into one large study
Meta-analysis is useful if …
• If the studies report different effects
• If the sample sizes are small and as such there in insufficient power
– Study A – 25; Study B – 20; Study C – 22; Study D – 19; Study E – 29
• Single studies rarely provide definitive conclusions
– You will need a body of evidence
Meta-analysis can only be used when …
• All included studies have the
– Same population
– Use the same intervention administered in the same way
– Measure the same outcomes in the same way
• Homogeneity
– Studies are sufficiently similar to estimate an average effect
Meta-analysis is not possible when…
• Descriptive/narrative synthesis
– When there are two or more studies which test similar interventions, on similar (but not necessarily the same) subject types, reporting any outcome measure
• Multiple ways to do this with no standard, agreed process
• Drawbacks
– Can be subjective
Limitations of systematic review
• Like any research, systematic reviews can be done badly
• Inappropriate aggregation of studies
– different interventions
– sub-groups within samples
• e.g.: sub acute and chronic low back pain
• Out of date
– systematic reviews can become out of date very quickly
Systematic review is not …
• Ideal when the evidence base is limited
– in terms of quantity (emerging area)
– in terms of quality (low level evidence only such pre-post study)
• Always readily available for all types of review questions
– predominant focus on intervention
• Immune to publication bias
– a systematic review may only include research which is published
