[MUSIC] Hi, this is Kay Dickersin, and I'm going to talk today about Qualitative Synthesis and Interpreting Results. Our objectives in this lecture are first to identify the main components of a systematic review. And by now, you really know what those components are, but let's look at them in this lecture, in terms of where does the qualitative synthesis fit in. To try to understand what it is. Second, then we'll review the importance of the qualitative aspects of systematic reviews, and some examples of a qualitative synthesis. Before I begin, I will say that this is absolutely the hardest part of your systematic review. It's also the most important. We're talking about systematic reviews here, and sometimes you do a quantitative synthesis that is a meta analysis, bringing together the results from all the studies. But, the qualitative synthesis is bringing together the results of all the studies with out the numbers. So how do you actually do that? And it's hard to do. I'll try to give you my interpretation of how to do it. But in the end it's going to take looking at examples, talking amongst one another, and really trying to figure out what it is as you go along. But remember you are the one who knows the most about the topic that you are doing the systematic review on. You know more then me, you know more then any of your readers because you have been immersed in the topic doing your systematic review. In section A, we're going to talk about why do we need a qualitative synthesis, and then we're going to move onto and what's it all about. So let's look at a definition first about systematic reviews. Now, in the old days, systematic reviews were called meta-analysis. We now divide the systematic review into two parts. The idea of doing the review systematically and then the quantitative synthesis part, which may or may not be there. And that's called meta-analysis. But in the old days, which is 20 years ago, 1995, meta-analysis was considered the whole thing, so let's just think of that when we look at what's being said here. I'm going to read it. Meta-analysis was developed to replace the artistic narrative review with a scientific and systematic method. So far so good. Yet in fear of allowing bias to creep in, meta-analysis is typically mechanistic, driven more by concerns about reliability and replicability than about adding to understanding the phenomena of interest. And that's what's important here, is we're not just trying to do a mechanist synthesis of all the studies that are out there on a particular topic. We're trying to also bring in the part about knowledge, the part about understanding the topic, about interpretation, because as I said, you know more than anybody else by the time you finish your systematic review. And what I've said before, meta-analysis is not the same as systematic review. A meta-analysis always requires a systematic review but not vice versa. That is, a systematic review does not have to have a meta-analysis. So what is a systematic review if it's not a meta-analysis? We've talked about the structure of systematic reviews, that there has to be a background and a description of why you need a systematic review and meta-analysis. You need the methods section, which describes exactly what you did in your systematic review, the search strategy, the eligibility criteria. How you do data abstraction and data management, your assessment of the quality and the risk of bias of the individual studies you include and your statistical methods. And then when you come to the results section. You'll talk about what you found when you did your search. And that will typically involve a flow chart of some sort. And then both a qualitative synthesis and a quantitative synthesis. And sometimes, as we said, the quantitative synthesis won't be warranted for various reasons. We'll talk about that in another lecture. But the qualitative synthesis. A part of the results. Now, you will see sometimes in some systematic reviews, and this has been true sometimes for Cochrane systematic reviews, that this qualitative synthesis is in the discussion section. People see the qualitative synthesis as putting the findings in context. But we're not interested in context here so much as what people found qualitatively, not quantitatively. So that may involve the discordance among the different studies that you're including in your systematic review. When, for example, people have looked at different systematic reviews themselves. How do they differ if they're on the same topic? Well they might have asked a slightly different research question. That is, the population of patients may be different. The interventions might be slightly different. The outcome measurements might be slightly different in the settings. And we've talked about this before. Difference among the individual studies you'll combine as well as differences in the final systematic review from another systematic review that might be on a similar topic. Well what is that then? That's the qualitative nature of the systematic review itself. A systematic review might differ from another systematic review because of how the author's went about selecting the studies to be included or how the data were extracted or how the study quality was assessed or the risk of bias assessed. And then whether the studies could be combined or not. Some of you may have already come across in your own systematic review that you don't think maybe these studies should be combined quantitatively. While in the index review that we've given you to look at on the same topic, they did combine the studies to get a meta-analysis. You might disagree with them. Well, why do you disagree with them? What is different about your interpretation of the studies you found than theirs that you don't think a quantitative synthesis is warranted? So there can be difference among systematic reviews just as there can be differences among the studies that go into a systematic review. And that perhaps would be best explained if you look at the data qualitatively and say. What are they like? How do we interpret them? Why do we believe they belong under this question? In the next section, we're going to discuss what goes into a qualitative synthesis, and this is where things start getting a little bit fuzzy. But more soon.
