"A systematic review attempts to collate all empirical evidence that fits pre-specified eligibility criteria to answer a specific research question. It uses explicit, systematic methods that are selected to minimize bias, thus providing reliable findings from which conclusions can be drawn and decisions made... Meta-analysis is the use of statistical methods to summarize and combine the results of independent studies." - Cochrane Collaboration
I. Before carrying out your systematic review, search the literature (in PubMed, Scopus, Google Scholar, etc.) for any systematic reviews that have already been published on your topic or related to your topic. The benefits are two-fold:
(1) helps to ensure that the work has not already been done (if it has, you can re-formulate your topic)
(2) provides examples of search strategies used for your topic or for parts of your topic.
NOTE: You should also search for other types of studies on your topic to make sure there is enough literature to conduct a systematic review. For example, for intervention or therapy studies, for which randomized controlled trials are the gold standard in terms of evidence, you want to find at least a few RCT's on your topic to make sure a full systematic review is feasible.
II. It is strongly recommended that you register your review with PROSPERO to help make sure the same search question wasn't / isn't being investigated by others and that no one else will start a systematic review on this topic while you're conducting one.
III. As obvious as it sounds, a systematic review / meta-analysis must be conducted in a systematic manner, and a well-defined and well-constructed protocol will get you started on the right foot! Make sure to consult at least one set of guidelines (e.g. PRISMA, Cochrane) to help develop your protocol:
Construct a narrow foreground question using the PICO format (worksheet available in right-hand column). Or, use the following formula as a template:
In Patients [include any significant demographics] with [specify Problem],
does [specify Intervention] or [specify Comparison, if any]
affect [specify Outcome]?
For example: In patients with type 2 diabetes, does a low-carbohydrate diet or a low-fat diet improve metabolic control and quality of life?
The type of studies you should include in your literature search will depend on the type of question being asked:
|Question Type||Suggested Study Types|
|Diagnosis||RCTs > prospective studies|
|Therapy||RCTs > cohort studies > case-control studies|
|Etiology||RCTs > cohort studies > case-control studies|
|Prognosis||cohort studies > case-control studies|
Use these forms to better define your own research question and search strategy.
A protocol checklist and flow diagram from PRISMA.
Contact a librarian if you would like some help searching for information on your topic or tracking down the full text of an article. That's why we're here!
NOTE: Step 14 is for researchers performing meta-analyses only.
Tsafnet G, Glasziou P, Choong MK, et al. Systematic review automation technologies. Systematic Reviews 2014;3:74. http://www.systematicreviewsjournal.com/content/3/1/74