Penney Upton on Systematic and Narrative Reviews

Well I mean I think the first thing, the sort of starting point is to recognise that there are different kinds of review that you can do. And there's a lot now talked about having either a systematic review or a narrative review and we're even sort of bringing in new elements of doing reviewing from other areas such as policy making. And I think that's quite exciting and quite important because it means before you make the decision you have to look at the kind of review you think is going to be most beneficial for the particular area you're looking at and there are some very good guidelines out there that people might like to look at. So traditionally Cochrane reviews that I'm sure everybody knows because of evidence-based practice they usually follow a sort of prism approach but there are new guidelines out there as well that people can follow new frameworks and one of a good sort of starting point would be together as an open journal open access journal run by BioMed Central called BMC Medicine and that has an overview of the kinds of new approaches that are available. So I think that's a good starting point, choose your framework. With a systematic review what you're doing is you're doing a very defined search to make sure that you get every single piece of literature that's out there on a particular topic. You choose a framework, you decide what it is you're looking at specifically. So for example I've done systematic reviews where you've looked at things like the reliability and validity of measures that have been used to assess quality of life. So with that, you have a very structured approach, you're very careful about how you're defining reliability and validity, what kinds of measures you're looking for in each of the papers that you choose. And narrative review has a much more flexible approach, there's in a sense almost a more qualitative element. So it's about looking at other, sort of more, less well-defined rigorous terms. Does that make sense? So it's much broader. So when you're doing an initial degree, when you do your sort of pre-reg nursing, undoubtedly, you know, you have to write essays. You may even write a literature review when you undertake a piece of research. Those kinds of reviews are selective reviews, aren't they? And they're much more fluid and flexible than a systematic review, which has very specific criteria. So a narrative review is marrying those two things together. Once you've got your framework and you know what the question is that you want to answer, you need to start looking at the literature. And I think there are some very important do's and don'ts around that. So I'll start with the don'ts. I'd say that one of the big don'ts is don't just rely on the internet. It's very easy now to find lots of literature out there. Google Scholar is absolutely brilliant and very helpful, but at the same time there's also lots of information out there that's not necessarily of a good source. So people can self-publish. Sometimes you come across a piece of information and think it might be useful and helpful, but actually when you trace back you find out it's not peer-reviewed and obviously as healthcare professionals it's really important that you find something got that evidence base and has been reviewed by peers. So that's a good don't. In terms of a do, do go to those good traditional databases, those search engines, PubMed, CINAHL, all of those. I'd also say don't just pick one, use all of them. You can usually, through sort of library search engines, you can put them all together and do a multi-search. The The reason for doing that is that sometimes you miss a piece of literature by omitting one of those engines. And just to give you what I think is quite an interesting little anecdote is a few years ago, I did a systematic review with a colleague in America. We both thought we were using the same search engines, but when we came to the end of it, there were some papers I'd been able to access on my list that weren't on hers. So it's about that triangulation. The other thing, of course, is when you're doing a systematic review, it's good practice to have two reviewers and to do that. So independently look for papers, independently do the rating of papers, and then come together and discuss and so on and take that forward. The other important thing is you have to define your terms very clearly when you do your search. And obviously, if you use different definitions, you might get slightly different set of papers to someone else. So it's important to be very clear on that. But then when you've got, usually you do a very broad search and you've probably got more papers than you actually need or that are useful even though some of them may have a keyword that links into your particular area. So you then have to go through each paper and review briefly by looking at the abstract whether or not it actually fits in so you have a set of very clearly defined inclusion and exclusion criteria. So that's your first way of reviewing and rating. And then from that, you'll have your framework for your review. So for example, if you're looking at psychometric measures and you're looking at how well those measures have been applied, you might be looking at things like what's the reliability and the validity of the measures. So you'll then be rating each paper on those kinds of criteria as well. First of all, it's important that you have inclusion and exclusion criteria. So there might be papers that then are excluded from your final review. But also the other thing that you would do is once you've got your final set of, or what you think is your final set of papers from that sort of literature search, is then going through and actually doing what we call snowballing, which means that sometimes there might be things that are quoted that you want to go and look up, so it's checking that reference list and going and following up secondary sources, so they then become part of your primary sources.