Professional Documents
Culture Documents
How To Do A Systematic Review: A Guide For HDR Students and EMCR's
How To Do A Systematic Review: A Guide For HDR Students and EMCR's
How to do a
systematic review: a
guide for HDR
students and EMCR’s
Associate Professor
Susan Torres
storres@deakin.edu.au
Deakin Lab for Meta‐analysis
Research (DeLMAR)
Deakin University CRICOS Provider Code: 00113B
Deakin Lab for Meta Analysis
Research (DeLMAR)
Vision: To become the leader in evidence
based meta‐analysis approaches within
Australia
https://www.deakin.edu.au/business/research/delmar
The effect of diet combined with exercise on health‐
related quality of life: a systematic review and meta‐
analysis of randomized controlled trials
Systematic review Finding/critically
appraising the evidence
Meta‐analysis
Analysing the evidence
1
12‐February‐2020
Outline
1. Different types of reviews
2. Why would you do a systematic review
3. Registering your review
4. Developing your review question
5. Developing the search terms and search
strategy
6. Selecting your databases
7. Conducting your searches
8. Screening your studies
Outline
9. Data extraction
10. Assessing risk of bias
11. Synthesising the data
12. Guides for conducting systematic
reviews
13. Systematic review exemplars
14. The outcome of your systematic review
15. Meta‐analysis courses
16. References
1. Different types of reviews
Adapted from: Booth, A. (2016) EVIDENT Guidance for Reviewing the Evidence: a compendium of
methodological literature and websites DOI: 10.13140/RG.2.1.1562.9842
2
12‐February‐2020
Scoping review
To identify types of evidence in a given field
To clarify key concepts/definitions in the literature
Examine how research is conducted on a certain topic
or field
Identify and analyse knowledge gaps
As a precursor to a systematic review
Munn et al. Systematic review or scoping review? Guidance for authors when choosing between a
systematic or scoping review approach. BMC Medical Research Methodology (2018) 18:143
Scoping review
Review protocol
Transparent, peer reviewed search strategy
Include steps to reduce error and increase reliability
(multiple reviewers)
Standardised data extraction forms
Defining a systematic review
“A review of a clearly formulated question that
uses systematic and explicit methods to identify,
select and critically appraise relevant research,
and to collect and analyse data from studies that
are included in the review” (Cochrane
Collaboration, 2003)
3
12‐February‐2020
2. Why would you do a
systematic review?
You have a research question that is best answered by
synthesising existing evidence rather than conducting a
new study
There is no recent review that answers that same
question
To develop high level skills in searching and
synthesising the literature
“High quality literature reviews bring
together, synthesise and critique one or
more literatures to provide an overall
impression of the extent, nature and
quality of evidence in relation to a
particular research question, highlighting
gaps” (Siddaway, 2019)
Time required to complete a
systematic review
Time (in weeks; registered project start to
67.3±31.0 65.8 6–186
publication date)
4
12‐February‐2020
3. Registering your review
Prospero: https://www.crd.york.ac.uk/prospero/
Search for any existing similar systematic reviews
The protocol should specify the following:
Background
Review question
Inclusion and exclusion criteria
Databases and search terms
Participants/population
Intervention/control
Study design (eg RCT, cohort)
Main outcome
Data extraction
Risk of bias (quality) assessment
Strategy for data synthesis
Other organisations to register
your review
Cochrane Library
https://www.cochranelibrary.com/
Joanna Briggs Institute https://joannabriggs.org
Campbell Collaboration
https://www.campbellcollaboration.org/
4. Developing your review question
PICOS
Participants
Intervention
Comparison
Outcome
Study design
5
12‐February‐2020
Participants Intervention
What is the effect of weight loss on psychological
measures of stress in randomized controlled trials
that induced weight loss by dietary restriction
compared to usual diet in overweight and obese
adults?
16
Deakin University CRICOS Provider Code: 00113B
5. Developing your search terms
and search strategy
Spend time developing your search terms and
strategy
Talk to your Faculty librarian about developing
your search strategy
Systematic and systematic‐like review toolkit
http://deakin.libguides.com/c.php?g=638978&p=4473
298
Developing the research question
Developing the search
Selection and screening of articles
Advanced Search Guide‐ Health
http://deakin.libguides.com/ld.php?content_id=29454607
Example search strategy
What is the effect of diet combined with exercise compared to
diet or exercise alone on health‐related quality of life in
community dwelling adults?
quality of life
resistance training
diet
randomised controlled trial
6
12‐February‐2020
19
Deakin University CRICOS Provider Code: 00113B
20
Deakin University CRICOS Provider Code: 00113B
6. Selecting your databases
Consult with your Faculty Librarian
3‐4 databases is usually sufficient
Note that Google Scholar will give different
results to different people with identical
search terms!
This is a big problem for a systematic
review.
7
12‐February‐2020
Grey literature
Theses, government reports, white papers,
unpublished trials
PsycEXTRA, Informit
Hand searching
Review contents page of relevant journals
Review reference list of all articles included
in your final search
Search for articles which have cited articles
included in your final search (Scopus)
8
12‐February‐2020
25
Deakin University CRICOS Provider Code: 00113B
7. Conducting your search
Register and save your searches in the
databases
Export to separate Endnote
libraries/separate folders in one library
Combine all the articles in one Endnote
library
Remove all duplicates
Document your process
Register and save your searches
27
Deakin University CRICOS Provider Code: 00113B
9
12‐February‐2020
Documenting your searches
Diet, exercise and QOL: October 2019
Medline Complete: 1696
PsycINFO: 195
Embase: 1709
Total= 3600
Total (duplicates removed)= 3250
28
A Note on Documentation
Everyone has had a computer disk crash at some point
Keep a running record of every step of your search process
in at least four places
1. On your hard drive
2. On your web based email
3. On a backup drive that is NOT in your office
4. On paper
29
Deleting duplicates in Endnote
Diet, exercise and QOL: October 2019
Medline Complete: 1696
PsycINFO: 195
Embase: 1709
Total= 3600
Total (duplicates removed)= 3110
30
10
12‐February‐2020
References/ Find duplicates/
Cancel then send references to Trash
31
Deakin University CRICOS Provider Code: 00113B
References/ Find duplicates/
32
Deakin University CRICOS Provider Code: 00113B
Deleting duplicates in Endnote
Manually screen for duplicates 2 times
You will still miss duplicates!
Ballin Marcel
Marcel Ballin
33
11
12‐February‐2020
8. Screening your studies
This is the most time consuming part of the
systematic review
Different tools to screen articles
Rayyan https://rayyan.qcri.org
Covidence https://www.covidence.org/reviews/active
remove duplicates
title/abstract screening 2 People
full text screening 2 People‐ with reason for
prisma flow chart excluding paper
data extraction
risk of bias
Demonstration of Covidence
https://www.youtube.com/watch?v=G4JEQ_curT4
9. Data extraction
Use a standard data extraction process for
all studies
Duplicate extraction of key study data
where possible with a second researcher
Record source location for all extracted
data
May need to contact authors for data
12
12‐February‐2020
37
Deakin University CRICOS Provider Code: 00113B
38
Deakin University CRICOS Provider Code: 00113B
10. Assessing risk of bias
Risk of bias assessment determines overall
quality of studies in your review
There are widely used tools for randomised
and quasi‐experimental designs (Cochrane
Risk of Bias Tool and RoB 2)
https://sites.google.com/site/riskofbiastool/welcome/rob‐2‐0‐
tool/current‐version‐of‐rob‐2
There is less consensus about appropriate
tools for observational studies
13
12‐February‐2020
https://methods.cochrane.org/bias/sites/methods.cochrane.org.bias/files/public/uploads/6.%20Assessing%20risk%20of%20
bias%20in%20included%20studies%20v1.0%20Standard%20author%20slides.pdf
Intervention studies
Random sequence generation
Allocation concealment
Blinding of participants and personnel
Blinding of outcome assessment
Incomplete outcome data
Selective reporting
Overall Risk of Bias
Random sequence generation
• occurs at the start of trial before
allocation of participants
• avoids differences between groups
• Low risk: computer random number
generator
14
12‐February‐2020
Allocation concealment
• occurs at the start of trial during
allocation of participants
• avoids selection bias
• Low risk: sequentially numbered,
sealed envelopes
Blinding of participants and personnel
• avoids different treatment of groups
• Low risk: no/incomplete blinding but
outcome unlikely to be influenced
Blinding of outcome assessment
• avoids measurement of outcome
affected by knowledge of the
intervention received
• Low risk: no blinding, but
measurement unlikely to be
influenced
15
12‐February‐2020
Incomplete outcome data
• complete outcome data for all
participants is not available
• Low risk: no missing data, missing
data balanced across groups and
reasons similar
Selective reporting
• can lead to reporting bias
• statistically significant results more
likely to be reported
• Low risk: protocol is available and all
outcomes of interests to the review
are reported in the pre‐specified way
Cochrane handbook: Chapter 8
https://training.cochrane.org/handbook/current
Cochrane training: Assessing risk of bias
in included studies
https://bit.ly/317I2zu
https://methods.cochrane.org/bias/sites/methods.cochrane.
org.bias/files/public/uploads/6.%20Assessing%20risk%20of%
20bias%20in%20included%20studies%20v1.0%20Standard%2
0author%20slides.pdf
48
Deakin University CRICOS Provider Code: 00113B
16
12‐February‐2020
Observational studies
Quality assessment tool for observational cohort
and cross‐sectional studies
https://www.nhlbi.nih.gov/health‐topics/study‐quality‐
assessment‐tools
11. Synthesising the data
There are different approaches
Results
Study selection flow chart
General characteristics of the included
studies
eg sample size, age of participants, country of origin
Tools to measure exposure eg diet
Tools to measure outcome eg stress
51 http://www.prisma‐statement.org/ Template for flow chart
Deakin University CRICOS Provider Code: 00113B
17
12‐February‐2020
12. Guides for conducting a
systematic review
PRISMA
http://www.prisma‐statement.org/
Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis JP, et al.
The PRISMA statement for reporting systematic reviews and meta‐analyses of
studies that evaluate healthcare interventions: explanation and elaboration.
BMJ. 2009;339:b2700.
Moher D, Liberati A, Tetzlaff J, Altman DG, Group P. Preferred reporting items
for systematic reviews and meta‐analyses: the PRISMA statement. BMJ.
2009;339:b2535.
Cochrane collaboration
https://training.cochrane.org/handbook/current
Version 6 October 2019
PRISMA Checklist: How to write
your systematic review
http://www.prisma‐statement.org/
Title: Psychological stress reactivity and future
health and disease outcomes: a systematic review
of prospective evidence
Protocol and registration: This review protocol was
registered with PROSPERO on 19 Dec 2017 (Registration
ID CRD42017084161)
Eligibility criteria: Included studies were conducted in
apparently healthy adult populations. They were
observational, prospective, population‐based cohort studies.
18
12‐February‐2020
Study selection: Our primary search yielded 1719
studies and we identified one additional study via our
included studies.
Risk of bias: For this body of work as a whole, study quality is
generally high and risk of bias is generally low (Figure 3)
Evidence Synthesis Academy
www.evsynthacademy.org
Boland A, Cherry GM, Dickson R.
Doing a systematic review. A
students guide. 1st ed. London:
Sage; 2014.
19
12‐February‐2020
Deakin library
https://www.deakin.edu.au/library/research
13. Systematic review
exemplars
British Medical Journal
https://www.bmj.com/content/bmj/348/bmj.g3253.full.pdf
https://www.bmj.com/content/bmj/364/bmj.k4718.full.pdf
Qualitative systematic review
http://web.a.ebscohost.com/ehost/pdfviewer/pdfviewer?vid=1
&sid=37fd0fd6‐e2de‐4a68‐8fdb‐6692fb98f990%40sdc‐v‐
sessmgr03
New England Journal of Medicine
14. The outcome of your
systematic review
HDR students: Confirmation document
HDR and EMCR: Publication
20
12‐February‐2020
HDR students: confirmation
document
Literature
review Systematic review
background
search strategy
results (Table)
discussion
Aim
Research proposal
Study 1 (Systematic review‐ one page overview)
Study 2 (Original study 1)
Study 3 (Original study 2)
Study 4 (original study 3)
61
Deakin University CRICOS Provider Code: 00113B
HDR students: final thesis
Chapter 1 Literature review
Chapter 2 Systematic review
Chapter 3 Original study 1
Chapter 4 Original study 2
Chapter 5 Original study 3
Chapter 6 General discussion
Outcome of the systematic
review
Decide not to publish
Publish as a systematic review
Develop into a meta‐analysis
21
12‐February‐2020
When does a systematic
review become a meta‐analysis?
Statistical combination of results from two or
more separate studies
When you have similar data across studies
eg mean changes, odds ratios
What is the effect of weight loss on stress?
15. Meta‐analysis courses
Faculty of Health Biostatistics unit: Stata (free)
https://www.deakin.edu.au/students/faculties/faculty‐of‐
health/research/deakin‐biostatistics‐unit
DeLMAR (free)
Cochrane: RevMan (free)
https://training.cochrane.org/online‐learning/core‐software‐cochrane‐
reviews/review‐manager‐revman
Comprehensive Meta‐Analysis: 3 day course ($3500)
and software ($410 2 year licence)
https://www.meta‐analysis‐workshops.com/
R: training? (free)
Available on Deakin Software Centre
Any Questions?
22
12‐February‐2020
16. References
Booth, A. (2016) EVIDENT Guidance for Reviewing the Evidence: a compendium of
methodological literature and websites DOI: 10.13140/RG.2.1.1562.9842
Borah R, Brown AW, Capers PL, et al Analysis of the time and workers needed to
conduct systematic reviews of medical interventions using data from the PROSPERO
registry BMJ Open 2017;7:e012545. doi: 10.1136/bmjopen‐2016‐012545
Siddaway AP et al. How to do a systematic review: a best practice guide for conducting
and reporting narrative reviews, meta‐analyses, and meta‐syntheses. Annu Rev
Psychol 2019 70:747‐70
Cochrane Collab 2003 http://community.Cochrane.org/glossary
23