HTHSCI 3C04 Lecture Notes - Lecture 9: Systematic Review, Campbell Collaboration, Type I And Type Ii Errors
28 Apr 2018
School
Department
Course
Professor
UNIT IX: Systematic Reviews and Meta-analyses
Systematic review: a rigorous summary of all the research evidence that relates to a specific question; the question may pertain to
effectiveness of an intervention, causation, diagnosis or prognosis
- Attempt to overcome possible biases at all stages by following a rigorous methodology for:
o Searching (comprehensive)
▪ Researchers use pre-established criteria for selecting studies (relevant and valid)
▪ 2 or more individuals independently select the studies, compare results, and come to consensus
about selection of the studies
o Research retrieval (including review of study
relevance)
o Appraisal of retrieved research for relevance
and validity (quality of methodology)
o Data extraction
o Data synthesis/analysis
o Interpretation
- Reduces bias by:
o The use of explicit, pre-set criteria to select
studies for inclusion on the basis of relevance
and validity
o By having two or more people independently
make study selection decisions, compare their
decisions, and discuss discrepancies before
moving on to independently extract data from
the studies
o Explicit details of the methods used at every
stage are recorded
- Many incorporate meta-analysis
- Have the potential to overcome several barriers to
utilization of research by clinicians; offers nurses a solution in form of a summary of research-based knowledge on a
topic, which takes into account the validity of the primary research studies
Key computerized databases for systemic reviews
- Cochrane reviews: healthcare and health policy; intervention reviews, diagnostic test accuracy reviews, methodology
reviews, qualitative reviews and prognosis reviews
- DARE (Database of Abstracts of Reviews of Effects): evaluate the effects of healthcare interventions and the delivery
and organization of health services; also contains reviews of housing, transport, and social care where these impact
directly on health or have the potential
- Campbell collaboration: international; effects of social interventions in crime & justice, education, international
development and social welfare
Meta-Analysis: combines results of similar studies quantitatively; produces a summary statistic that represents the effect of the
intervention across different studies; it is much more precise than the effect size in any one contributing study
- Should only proceed from systematic reviews, but not all systematic reviews will include meta-analyses (have to
consider heterogeneity)
- Decision to conduct is based on two criteria:
o Judgement as to whether statistical synthesis is appropriate (does it make clinical and methodological sense
to combine the results); need to examine aspects of the studies such as the population, intervention, and
outcome
o Test of heterogeneity: forest plot; tests whether differences in studies are likely due to chance alone (the more
significant the test, the less likely this is; subgroup analyses often done)
▪ Fixed effects: most often used when no significant heterogeneity exists between studies
• Pool data and examine statistical heterogeneity with Chi square or I-square test
▪ Random effects: most often used where significant heterogeneity exists between studies
• Don’t pool data due to high levels of variability; perform subgroup analysis (high type I
error)
▪ Weighting: how much a particular study contributes to the overall summary statistic; always look at
this to see how much influence individual studies have (ex. does the summary statistic reflect only
a few studies, or many)
Meta-Synthesis: combines the results of qualitative studies (summarizes a small number of interpretations of an event or
phenomenon); by bringing different studies together, strengthens the evidence for an interpretation by discovering common
themes and differences, and building new interpretations of the topic of interest
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Systematic review: a rigorous summary of all the research evidence that relates to a specific question; the question may pertain to effectiveness of an intervention, causation, diagnosis or prognosis. Attempt to overcome possible biases at all stages by following a rigorous methodology for: Searching (comprehensive: researchers use pre-established criteria for selecting studies (relevant and valid) Cochrane reviews: healthcare and health policy; intervention reviews, diagnostic test accuracy reviews, methodology reviews, qualitative reviews and prognosis reviews. Campbell collaboration: international; effects of social interventions in crime & justice, education, international development and social welfare. Meta-analysis: combines results of similar studies quantitatively; produces a summary statistic that represents the effect of the intervention across different studies; it is much more precise than the effect size in any one contributing study. Should only proceed from systematic reviews, but not all systematic reviews will include meta-analyses (have to consider heterogeneity) Decision to conduct is based on two criteria:
