Introduction to Systematic Review & Meta-Analysis

a brief introduction to systematic review & meta-analysis

2. Assoc. Prof. Dr.
HasanainFaisalGhazi
•MBChB (University of Baghdad)
Master & PhD (UKM, Malaysia)
•Publication:
89 Full papers (56 in SCOPUS)
H Index = 11
•Published Systematic Review & Meta
Analysis (7)
•Editor in Chief, Global Journal of Public
Health Medicine.
@AP_Dr_Hasanain_Faisal_Ghazi 2

3. INTRODUCTION
Why We Do Sys. Rev. ??????
@AP_Dr.Hasanain_Ghazi

4. Evidence-based medicine
•Istheprocessofsystematicallyreviewing,appraising,andusing
clinicalresearchfindingstoaidthedeliveryofoptimumclinicalcare
topatients.
Clinical
Judgment
Relevant
scientific
evidences
Patient values
& preferences
EBM

5. Definition
•Asystematicreviewsummarisestheresultsofavailablecarefully
designedhealthcarestudies(controlledtrials)andprovidesahigh
levelofevidenceontheeffectivenessofhealthcare
interventions.Judgmentsmaybemadeabouttheevidenceandinform
recommendationsforhealthcare.
Reference:https://consumers.cochrane.org/what-systematic-review
@AP_Dr.Hasanain_Ghazi

6. SR & MA
•Asystematicreviewanswersadefinedresearchquestionby
collectingandsummarisingallempiricalevidencethatfitspre-
specifiedeligibilitycriteria.
•Ameta-analysisistheuseofstatisticalmethodstosummarisethe
resultsofthesestudies.
@AP_Dr.Hasanain_Ghazi

7. WHATISSYSTEMATICREVIEW
•Focus on specific question
•use explicit , pre-planned scientific methods to:
1.Identify
2.Select
3.Appraise
4.Summarize similar but separate studies
@AP_Dr.Hasanain_Ghazi

8. Must Have
•clearlystatedobjectiveswithpre-definedeligibilitycriteriaforstudies
•explicit,reproduciblemethodology
•asystematicsearchthatattemptstoidentifyallstudies
•assessmentofthevalidityofthefindingsoftheincludedstudies(e.g.
riskofbias)
•systematicpresentation,andsynthesis,ofthecharacteristicsand
findingsoftheincludedstudies
@AP_Dr.Hasanain_Ghazi

9. SYSTEMATIC
REVIEW META ANALYSIS
TRADITIONAL
NARRATIVE
REVIEW
@AP_Dr.Hasanain_Ghazi

10. TRADITIONALNARRATIVEREVIEW
•No standard format
•No specific methods for selecting and appraising
•Quantitively synthesis rarely used to integrate information from
multiple studies
@AP_Dr.Hasanain_Ghazi

11. SRSteps
1.Gather your team
2.Develop protocol
3. Data collection, locate and screen
@AP_Dr.Hasanain_Ghazi

12. 4. abstract data , appraise risk of Bias
5. synthesis findings and interpret
6. report writing
7. update
@AP_Dr.Hasanain_Ghazi

13. METAANALYSIS
•Is an OPTIONALcomponent of systematic review.
•Astatisticalanalysiswhichcombinestheresultsofseveral
independentstudiesconsideredbytheanalysttobecombinable
(Huque,1988).
@AP_Dr.Hasanain_Ghazi

14. What meta analysis can help
•Assess strength of evidence
To determine whether an effect exists in a particular direction
•Combine results quantitively
To obtain a single summary result
•Investigate heterogeneity
To examine reasons for different results among studies
@AP_Dr.Hasanain_Ghazi

15. Addressing same question?
•What is the effect of interferon therapy in hepatitis C?
•Size of effect might be higher when participants are older, more
educated or healthier than others
•Different forms of interferon
•Different interferon doses
•Different viral subtypes of hepatitis C
@AP_Dr.Hasanain_Ghazi

16. WHEN TO DO MA??
•Whenmorethan1studyhasestimatedatreatmenteffector
association.
•Whenthedifferenceinstudycharacteristicsareunlikelytoeffect
treatmenteffect.
•Whenthetreatmenteffecthavebeenmeasuredandreportedinsimilar
ways(orwhendataavailable).
@AP_Dr.Hasanain_Ghazi

17. WHO IS DOING SYS. REV.
•Independent authors
•Cochrane collaboration
•Group interested in policy
•Business
@AP_Dr.Hasanain_Ghazi

18. Who is using Sys Rev
•Individual doctors and researchers
•Patient and consumers
•Guidelines producers
•Policy makers
@AP_Dr.Hasanain_Ghazi

19. IOM standers for systematic reviews
•Initiating a systematic review
•Establish a review team
•Manage bias and conflict of interest of the review team
•Ensure stakeholder input
•Manage bias and conflict of interest of those providing input
•Formulate the topic
•Develop the protocol
•Submit the protocol for peer review
•Make protocol publicly available
@AP_Dr.Hasanain_Ghazi

20. Formulate the Topic
•Confirm the need for a new review
•Develop an analytic framework
•Use a standard format to articulate each question of interest For
example, PICO(TS)
•State a rationale for each question
•Refine each question
@AP_Dr.Hasanain_Ghazi

21. Formulating the Question in a Systematic
Review
•“The most important maxim for data analysis to heed, and one which
many statisticians have shunned is this: ‘far better an approximate
answer to the right question, which is often vague, than an exact
answer to the wrong question, which can always be made precise.’”
@AP_Dr.Hasanain_Ghazi

22. Ways Similar Studies May Differ
•Patient population
•Inclusion and exclusion criteria
•Exposure definition/intervention composition, dose, timing
•Comparison group/controls
•Management protocol(co interventions) for trials
•Outcome definition(follow-up time, cause-specific mortality)
•Quality of design and execution
•Analysis
@AP_Dr.Hasanain_Ghazi

23. Components of well-structured question
•P patients or population
•I intervention/ exposure
•C comparison groups (s)
•O outcome
@AP_Dr.Hasanain_Ghazi

24. Type of people
•Define condition or disease, including explicit diagnostic criteria
•Population and setting of interest (age, race, sex, community, hospital,
outpatient)
@AP_Dr.Hasanain_Ghazi

25. Type of risk factors
•Timing of exposure
•Route of administration
•Dose of intensity
•Duration of exposure or therapy
@AP_Dr.Hasanain_Ghazi

26. Type of comparison group
•For trials :
•Placebo
•Standard therapy
•No treatment
•For epidemiological STUDIES
•No exposure
•Non-cases (hospital, neighborhood, etc)
@AP_Dr.Hasanain_Ghazi

27. Types of outcome
•Criteria of defining
•Important to consumers, providers
•Unpublished data
@AP_Dr.Hasanain_Ghazi

28. Well constructed and answerable clinical
question
•Forpreschoolchildrenwithmildtomoderatevisualacuity
impairment,areglassesandpatchingeffectiveinimprovingvisual
acuity,comparedwithglassesaloneornotreatment?
@AP_Dr.Hasanain_Ghazi

29. Some groups add TS to PICO
•T= TIMING (Duration of minimum of follow up)
•S= sitting (primary care, specialty, inpatients )
@AP_Dr.Hasanain_Ghazi

30. •EffectivenessbestansweredwithRCT
•HarmbestansweredwithRCTbutusuallymorepracticaltouse
observationaldata(rareevents,occurringafterfollow-upcompletein
anRCT)
@AP_Dr.Hasanain_Ghazi

31. EXAMPLES
@AP_Dr.Hasanain_Ghazi

32. RESEARCH QUESTION
IS Drug therapy associated with long term morbidity and mortality in older
persons with moderate hypertension?
•P = older persons with moderate hypertension
•I= drug therapy
•C= not stated ( presumably any intervention other than named drug therapy
•O= long term morbidity and mortality
@AP_Dr.Hasanain_Ghazi

33. Finding the evidence
@AP_Dr.Hasanain_Ghazi

34. Major databases
•Major bibliographic databases for RCT
•Medline/Pubmed
•EMBASE
•Cochrane central register for controlled trials
•National and regional databases
•Subject specific datbases
•CINAHL
•PsychINFO
•OTSeeker
@AP_Dr.Hasanain_Ghazi

35. Citation databases
•Web of science
•SCOPUS
•Dissertation (ProQuest)
@AP_Dr.Hasanain_Ghazi

36. Register of ongoing trials
•www.clinicaltrials.gov
•www.centerwatch.com
•www.controlles-trials.com
•WHO Portal
@AP_Dr.Hasanain_Ghazi

37. Break into concepts
•P POPULATION
•I INTERVENTION
•C COMPARISON
•O OUTCOME
@AP_Dr.Hasanain_Ghazi

38. @AP_Dr.Hasanain_Ghazi
Patients with choroidal neovascularization associated with age related macular degeneration, do
intravitreal injections of Lucentis , when compared with Avastin , prevent vision loss?
Individuals with choroidal neovascularization associated with
age related macular degeneration
Lucentis (Ranibizumab)
Avastin
Changeinvisualacuity or visual field
P
I
C
O

39. •START WITH GENERAL FORMAT:
•(Population OR synonym #1 OR synonym #2) AND
•(Intervention OR synonym #1 OR synonym #2) AND
•(Comparison OR synonym #1 OR synonym #2) AND
•(Outcome OR synonym #1 OR synonym #2) AND
•Add study type filters
@AP_Dr.Hasanain_Ghazi

40. @AP_Dr.Hasanain_Ghazi

41. Documenting the search
•Bibliographic software management
•Endnote
•Mendely
•QUOSA
@AP_Dr.Hasanain_Ghazi

42. •WHEN DID THE SEARCH
•Month/year
•Where you searched
•Electronic databases
•Hand searches
•Trial registers
•Internet
•bibliographies
@AP_Dr.Hasanain_Ghazi

43. •How you searched
•Strategies
•What you found
•PRISMA flow diagram
@AP_Dr.Hasanain_Ghazi

44. PRISMA
http://www.prisma-statement.org/
@AP_Dr.Hasanain_Ghazi
Preferred Reporting Items for Systematic Review and Meta-analysis

45. @AP_Dr.Hasanain_Ghazi

46. @AP_Dr.Hasanain_Ghazi

47. Assessing the risk of BIAS
@AP_Dr.Hasanain_Ghazi

48. ELEMENTS OF STUDY QUALITY THAT CAN
BE ASSESSED BY READING STUDY REPORT
•INTERNAL VALIDITY
•EXTERNAL VALIDITY
•RELEVANCE, ORIGINALITY
•ETHICS
@AP_Dr.Hasanain_Ghazi

49. BIAS
•SELECTION BIAS
•INFORMATION BIAS
•ANALYSIS
@AP_Dr.Hasanain_Ghazi

50. META ANALYSIS
@AP_Dr.Hasanain_Ghazi

51. •A meta-analysis is a survey in which the results of the studies included
in the review are statistically similar and are combined and analysed as
if they were one study.
@AP_Dr.Hasanain_Ghazi

52. PLANNING THE ANALYSIS
•Results of meta analysis can be very misleading if suitable attention
has not been given to:
•Formulate the review question
•Specify eligibility criteria
•Identify, select and critically appraise studies
•Collect appropriate data
•Decide what would be meaningful to analyze
@AP_Dr.Hasanain_Ghazi

53. •Systematic review contain analysis of the primary studies
•Qualitative: structured summary, description , and discussion of the
studies characteristics that may affect the cumulative evidence
•Quantitative: involves statistical analysis( meta-analysis)
@AP_Dr.Hasanain_Ghazi

54. A general framework for synthesis
•What is the direction of effect (association )
•What is the size of effect?
•Is the effect consistent across studies ?
•What is the strength of evidence for the effect ?
@AP_Dr.Hasanain_Ghazi

55. What is meta analysis
•Anoptionalcomponentofsystematicreview
•Definition:
•Thestatisticalanalysisofalargecollectionofanalysisresultsfrom
individualstudiesforthepurposeofintegratingthefindings(Glass1976)
•Thestatisticalanalysiswhichcombinestheresultsofseveralindependent
studiesconsideredbytheanalysttobecombinable(Huque,1988)
@AP_Dr.Hasanain_Ghazi

56. Data presentation
@AP_Dr.Hasanain_Ghazi
•FOREST GRAPHS
•FUNNEL PLOTS
•RADIAL PLOTS

57. Reference : Hoffman, J. I. E. (2015).Meta-analysis. Biostatistics for Medical and Biomedical
Practitioners, 645–653.doi:10.1016/b978-0-12-802387-7.00036-

58. •Typicalfigureofresponsetosomenewtreatment.Thedotsarethe
meanlogoddsratio,andthehorizontallinesindicate95%confidence
limits.Figuresintheright-handcolumnaretheweightsassignedto
eachstudy;thestudiesareoftenarrangedindescendingweights.The
solidverticallineat0indicatesnodifferencebetweenthegroups,and
thedashedverticallineindicatesthepooledaverage.Somefigures
includethenumbersofsubjectsineachgroup.
@AP_Dr.Hasanain_Ghazi

59. Note on interpretation
•Theresultsofmeta-analysis(estimateandconfidenceinterval)asany
otherstudymustbeinterpretedinthecontextofaclinicallyimportant
effectsize
•Statisticallysignificantresultmaynotbeclinicallyimportant
•Resultsthatisnotstatisticallysignificantmaystillbecompatiblewith
aclinicallyimportanteffect
@AP_Dr.Hasanain_Ghazi

60. Why we do meta analysis
@AP_Dr.Hasanain_Ghazi
1.Increase power and precision (detect effect as statistically significant)
2. Quantify effect sizes and their uncertainty
Reduce problems of interpretation due to sampling variation
3. Assess homogeneity / heterogenicity of results
4. Answer questions not posed by individual studies

61. CONCLUSION
•Systematic review is essential for healthcare
•Meta analysis is an optional thing in SR
•SR should be done in team
@AP_Dr.Hasanain_Ghazi

62. •CochraneisaBritish
internationalcharitable
organisationformedto
organisemedicalresearch
findingstofacilitate
evidence-basedchoices
abouthealthinterventions
involvinghealth
professionals,patientsand
policymakers.Itincludes
53reviewgroupsthatare
basedatresearch
institutionsworldwide
@AP_Dr.Hasanain_Ghazi

63. @AP_Dr.Hasanain_Ghazi

64. @AP_Dr.Hasanain_Ghazi

65. FURTHER READINGS
@AP_Dr.Hasanain_Ghazi

66. https://www.nap.edu/
catalog/13059/finding
-what-works-in-health-
care-standards-for-
systematic-reviews
@AP_Dr.Hasanain_Ghazi

67. https://training.cochrane.org/handbook
@AP_Dr.Hasanain_Ghazi

68. @AP_Dr.Hasanain_Ghazi

69. THANK YOU
@AP_Dr.Hasanain_Ghazi