Step 1: Framing questions for a review
The problems to be addressed by the review should bespecified in the form of clear, unambiguous and structuredquestions before beginning the review work. Once the reviewquestions have been set, modifications to the protocol shouldbe allowed only if alternative ways of defining the populations,interventions, outcomes or study designs become apparent
Step 2: Identifying relevant work
The search for studies should be extensive. Multiple resources(both computerized and printed) should be searched withoutlanguage restrictions. The study selection criteria should flowdirectly from the review questions and be specifiedapriori.Reasons for inclusion and exclusion should be recorded
Step 3: Assessing the quality of studies
Study quality assessment is relevant to every step of a review.Question formulation (Step 1) and study selection criteria (Step2) should describe the minimum acceptable level of design.Selected studies should be subjected to a more refined qualityassessment by use of general critical appraisal guides anddesign-based quality checklists (Step 3). These detailedquality assessments will be used for exploring heterogeneityand informing decisions regarding suitability of meta-analysis(Step 4). In addition they help in assessing the strength ofinferences and making recommendations for future research(Step 5)
Step 4: Summarizing the evidence
Data synthesis consists of tabulation of study characteristics,quality and effects as well as use of statistical methods forexploring differences between studies and combining theireffects (meta-analysis). Exploration of heterogeneity and itssources should be planned in advance (Step 3). If an overallmeta-analysis cannot be done, subgroup meta-analysis maybe feasible
Step 5: Interpreting the findings
The issues highlighted in each of the four steps above shouldbe met. The risk of publication bias and related biases shouldbe explored. Exploration for heterogeneity should helpdetermine whether the overall summary can be trusted, and, ifnot, the effects observed in high-quality studies should beused for generating inferences. Any recommendations shouldbe graded by reference to the strengths and weaknesses of systematic review.
Hierarchies of evidence for questions of therapy, prevention, aetiology or harm:
Level 1a Systematic review (with homogeneity) of randomised controlled trials (RCTs)
Level 1b Individual RCT (with narrow confidence interval)
Level 1c All-or-none studies
Level 2a Systematic review (with homogeneity) of cohort studies
Level 2b Individual cohort study (including low quality RCT; eg <80% follow-up)
Level 2c ‘Outcomes’ research; ecological studies
Level 3a Systematic reviews (with homogeneity) of case-control studies
Level 3b Individual case-control study
Level 4 Case series (and poor quality cohort and case-control studies)
Level 5 Expert opinion without explicit critical appraisal, or based on physiology, bench research on 'first principles'