By Jenny Mc Sharry
For no particular reason I thought that a 5 tips on… post might be a nice way to give a brief intro to conducting a systematic review. This is completely my own idea and definitely not stolen from Lisa’s lovely post on 5 tips for taking on qualitative research. Maybe a tiny bit stolen.
Systematic reviews are different from traditional literature reviews, and different from introduction sections to a dissertation or paper, because they:
- Address a specific research question
- Use explicit and replicable methods
- Attempt to identify all relevant studies in an area
Systematic reviews synthesise large bodies of research, are very useful to other researchers, and can inform policy making and clinical decisions. Conducting a systematic review is also a requirement of the British Psychological Society Qualification in Health Psychology Stage 2 so is seen as a key skill for Health Psychologists. Having completed a few over the last couple of years, I’ve come up with the following 5 tips which might help you conduct a systematic review while staying (relatively) sane.
The best skill you can have when conducting a systematic review is to be organised and keep good records. If you are thinking of publishing your review, most journals ask for a PRISMA flow diagram so you will need to keep track of the papers included and the reasons for exclusion at different stages. I did my first systematic review during my PhD and thought I was being really organised. I wasn’t.
Since then, reference management software has become my best friend and I use folders and sub-folders to sort through references and keep track of what’s included and excluded at different stages.
These days, I also write the methods section for the systematic review as I go to make sure I record all the relevant information. This makes things a lot easier than waiting until you write the final report or paper and trying to frantically remember what you actually did six months earlier. Writing a protocol before you start is another good idea to help you plan exactly what you will be doing in advance.
Systematic reviews aim to address a research question, and the more specific you are, the more likely it is that you will make a useful contribution to the area. Having a specific research question makes it easier to develop narrow search terms that only identify relevant studies and will reduce the time you spend screening irrelevant abstracts. Being specific will also allow you to develop really clear inclusion and exclusion criteria and increase the chance that you and other members of your research team will agree on studies to include and exclude at different stages. The PICO criteria (or SPIDER for qualitative studies) can help specify your question.
Know the literature
It might seem a bit strange when the aim is to review a particular area, but the more familiar you are with the literature before you start the better. Knowing the literature in advance will help pick a useful review question that, importantly, is feasible with the resources you have. If you are planning a review in an area you are not familiar with, carrying out scoping searches is invaluable to get a feel for the amount and type of studies relevant to your review.
Form a team
Working as part of a team is recommended and is a requirement when conducting a Cochrane review, the gold standard in reviewing for evidence-based health care. Screening for inclusion, data extraction and quality appraisal should ideally be conducted by at least two people. If that isn’t possible, having a second reviewer for a percentage of all studies is still worthwhile. You can then calculate agreement between you and other reviewers to check if what you are doing could be replicated by others. Having other people working on the same review can also make the process a bit less tedious and provide much needed motivation when faced with an afternoon of screening references.
Make use of available expertise
Working with someone who has conducted a systematic review before can make the process a lot easier. Depending on the way you plan to bring together the results of the papers you have identified (e.g. meta-analysis, qualitative meta-synthesis or narrative synthesis) having a statistical or qualitative expert you can ask for advice can also really help. If you are studying or working at a university, most libraries can offer support and give advice on conducting a review. Talking to an academic librarian about developing search strategies, identifying suitable databases and working with reference management software can make the whole process a lot less mysterious. There is also lots of great advice available online; the Cochrane Collaboration handbook is a good place to start.
Finally, for NUIG MSc Health Psych students, there is a systematic review session coming your way on November 5th (lucky you!). For anyone else at NUIG interested in conducting a systematic review, we’ll be giving a session in the library on November 26th, you can register here.