The Systematic Review
Posted on 8th September 2014 by Danny Minkow
Key message/Bottom line:Â The Systematic Review is considered the highest level of research design and brings together all of the available evidence to find an answer to a research question.
So what is a Systematic Review?
There are many different study designs, but a Systematic Review is unique. Basically, itâ€™s a study of studies about an intervention. The benefit of the systematic review is that it is a one-stop shop summery of the evidence about a research question. In the Pyramid of Evidence Based Medicine, a Systematic Review of Randomized Control Trials is located at the top; because so many studies are used, it greatly reduces bias.
One of the first steps researchers take is to conduct an organized search to find and collect all of the relevant studies. This part is key. It is called a â€œSystematicâ€ Review because the method of searching and selecting studies that will be included in the review is done systematically. This selection process is spelled out in the Methods section. This way, the reader knows why and how certain studies were selected or not selected, and that the studies are not just â€˜cherry-pickedâ€™ based on their outcomes.
The researchers then decide which studies are of high enough quality to include in the review. Again, this criteria should be clearly spelled out in the methods. An assessment of â€œrisk of biasâ€ for each of the studies is then included. Finally, the researchers evaluate all of the studies together to come up with the results.
Frequently, a Meta-analysis is also included into the Systematic Review results. Â A Meta-analysis is a statistical comparing and combining of the results of each study included in the review. The results are frequently summarized using a diagram called a Forest Plot.
Forest Plots basically illustrate the effects of a treatment. Â I like to think of the Forest Plot diagram as helping to see the â€œforest for the treesâ€.Â With all the studies together, side by side, this gives the reader the big picture on how the studies compare with each other and the combined results show the overall pooled effect of the intervention.
The Cochrane CollaborationÂ takes systematic reviews to the next level.Â They are the experts in the area of systematic reviews. The researchers at the Cochrane Collaboration have an added level of rigor by developing special techniques to identify bias and they are a refreshingly independent voice of medical research. Cochrane Systematic Reviews are also different because they may include a Plain Language summary. This is a short separate section that clearly explains the study and results, using everyday language that the average healthcare consumer could understand.
What are the steps to conduct a Systematic Review?
TheÂ Cochrane HandbookÂ outlines eight general steps for preparing a systematic review:
- Defining the review question(s) and developing criteria for including studies
- Searching for studies
- Selecting studies and collecting data
- Assessing risk of bias in included studies
- Analyzing data and undertaking meta-analyses
- Addressing reporting biases
- Presenting results and “summary of findings” tables
- Interpreting results and drawing conclusions
So how is a Systematic Review structured?
A basic Systematic Review contains the following sections:
- Background: In this section, the researchers basically review what we know and donâ€™t know about the topic.
- Objectives: This section should clearly state the nature of the question the researchers are trying to answer.
- Methods: Here the researchers should transparently explain how the study was done. This way, anyone could reproduce the study following the exact methods.Â This section also lays out how the various studies were found (which databases and search words used) and why the studies were selected for inclusion in the review.
- Results: This section details what the researchers found in the review. Â This is where results are frequently summarized using a table or Forest Plot.
- Discussion: Finally, strengths and limitations of the review are listed in this section and the researchers talk about how the findings affect real world problems
How can healthcare providers use a Systematic Review?
Since we are all busy, itâ€™s great to have one study, a Systematic Review, to summarize all the best evidence on a topic. While a Systematic Review may be the strongest form of medical evidence, itâ€™s important to keep a few things in mind before acting. First, ensure the review is up to date with the latest studies. Also, consider that when making a medical decision, the best available evidence is, of course just part of the Evidence Based Medicine formula. A healthcare provider should use their clinical judgement to evaluate the evidence of potential harms, benefits, costs, and work with the patient to make a shared decision about what to do in the situation.
Conclusions & comment:
Systematic Reviews can be really useful researcher tools. Â By taking a look at all the evidence together in one study, we can learn a lot more than just looking at individual studies on their own. This could also serve to avoid unnecessary research and help find answers sooner and help more people.
Akobeng, A. K. â€œUnderstanding Systematic Reviews and Meta-Analysis.â€ Archives of Disease in Childhood 90, no. 8 (August 1, 2005): 845â€“48. doi:10.1136/adc.2004.058230.
Higgins JPT, Green S (editors).Â Cochrane handbook for systematic reviews of interventions, version 5.1.0 (updated March 2011).Â The Cochrane Collaboration, 2011. Retrieved 2014-09-06
Khan, Khalid S, Regina Kunz, Jos Kleijnen, and Gerd Antes. â€œFive Steps to Conducting a Systematic Review.â€ Journal of the Royal Society of Medicine 96, no. 3 (March 2003): 118â€“21.
Â Mulrow, C. D. â€œSystematic Reviews: Rationale for Systematic Reviews.â€ BMJ 309, no. 6954 (September 3, 1994): 597â€“99. doi:10.1136/bmj.309.6954.597.
Â Murad M, Montori VM, Ioannidis JA, and et al. â€œHow to Read a Systematic Review and Meta-Analysis and Apply the Results to Patient Care: Usersâ€™ Guides to the Medical Literature.â€ JAMA 312, no. 2 (July 9, 2014): 171â€“79. doi:10.1001/jama.2014.5559.