Approach to reviewing a review article

[00:00:45] Dr. Christine Klein: Hello and good afternoon, Dr. Prasad. It's a great pleasure to be here today. I would like to thank you very much for the invitation and I look forward to our discussion.
 

[00:00:53] Dr. Shweta Prasad: So today's podcast aims to discuss the approach to a review article, which is perhaps a rather intimidating assignment for a young reviewer. And one of the core requirements of accepting review assignment is a certain baseline level of awareness of a specific topic.

Given that review articles are intended to be thorough and detailed, young reviewers often lack the confidence to take up such an assignment. So how important is preexisting knowledge when it comes to critiquing or review article?
 

[00:01:21] Dr. Christine Klein: Thank you for this important question. So I think first of all, more generally speaking, I think you should only accept such an assignment when you have written a review article yourself.

A review article is fundamentally different from other types of articles, such as original articles, or even viewpoints. And therefore, I think it is good to have own experience, no matter what the topic is, but to have put something together as a review. I think that is an important experience that you actually have to have.

Otherwise, I think it's not only very helpful, but probably a prerequisite, really, should be, that you have a relatively good knowledge of the topic.

 Of course you don't have to know everything in detail, but if, for example, you are lacking a certain part of what is important for the review? Then maybe you won't be able to judge it properly.

However, I think there's one important other remark that I would like to make, and that is, of course, nobody knows everything. And often review articles are not written by a single person, but several people really gathering their expertise. So for example, let's say if it's a review article on neurophysiology in genetic forms of Parkinsonism. You, for example, are an electrophysiologist; you feel very confident on one part of the review, but not so much on the genetic part, perhaps. Then you can just state that in your message to the editor and can say, " I don't feel confident in this part. It will be good if another reviewer could take that on, but I focused on this part," so that's perfectly fine.
 

[00:02:37] Dr. Shweta Prasad: How do you actually judge the quality of the structure of a review article, whether the figures or the tables are actually adequate or not?
 

[00:02:45] Dr. Christine Klein: Right. So again, I would probably like to take one step back. I think most important thing when reviewing a review article is understanding the methods. I think this is extremely important so that, you know, how was this data put together? Is it a narrative review? To put a bit bluntly, to know, did somebody really just write down what he or she thinks, where the field stands and the papers that he or she contributed, or maybe the one or two other papers that they recently read. That's of course something that you will have to check, whether that's the way it was done, or whether there was at least some sort of a systematic review and approach to it.

Which doesn't mean that every review or every good review has to be either meta analysis, which is something different anyway, or a systematic review in the sense that the entire literature of the world in all languages was covered, which is typically not possible. But some approach that you can follow, and that I think then will answer your question. If they took this approach, then the tables and the figures should follow that naturally.

And at the same time, what I find important, is obviously we're seeing lots of reviews. And as we know, a lot of journals, they love reviews because that helps the impact factor. And so we may see more reviews than we actually need. And therefore it's sometimes a little bit difficult to really come up with something new and novel. So therefore it would also be good if you are then also looking at least in the last year or two, what else has been published in that field? And is it really an advance over what has just been published in maybe another journal? So I think this is also important, or are things actually quite repetitive? Which obviously it should be the case. At the same time, to be fair, a review article can not reinvent the wheel. So there will always be some amount of duplication, but that should be a good balance.
 

[00:04:21] Dr. Shweta Prasad: Well, as you mentioned, there are numerous variants of review articles. You have the meta analysis, you have the systematic review, you have the narrative review. Now, especially for the first two, how deep should a reviewer go to actually verify if everything has been accounted for by the authors of the article?
 

[00:04:39] Dr. Christine Klein: Right. Very, very important question. So, first of all, a review can be a lot of work to put together. It typically is, actually. And unless it's a very small, just emerging field, then it's easier.

But sometimes it can take years to really put this together carefully and depending on the methods. For example, the MDs gene review series that we working on, on phenotype genotype correlations in movement disorders, that typically takes up to two years to complete such a review. Because it's just so much work. There's so much literature and it all has to be covered carefully. Of course, it cannot be asked from the reviewer to do exactly this work again, obviously that's not the case.

But what I typically do is, and this tends to be quite a good guide and, gives you a feel for the quality of the review, I'll cross- reference some numbers . And then I dig a little bit deeper, but I'm not redoing the review obviously, which I don't think should be done. So you have to trust to some extent, the authors of the review. But I would definitely cross check a few things. And that, that typically is very informative already.

There's another thing, perhaps, in this context, which is the knowledge. as we discussed, It's sometimes difficult to really cover a topic completely because there also tend to be more and more methods involved. There's more interdisciplinary research in almost any area, and therefore it can sometimes be difficult, although the authors make a very good attempt.

 For example, I just saw a review article that put together all of the variants that were found for a specific familiar disease, a genetic disorder. And they listed very beautifully, all of the different cases that had been in published in the literature, but without any critical review, actually. And I think this should be part of the review and because they didn't have a background in genetics, they put together all of the variants of unknown significance and Medtronic changes, some real truncating material. Everything was in the same bin, whereas it should have been differentiated. But I'm not blaming them because they probably, they really relied on what they found in the literature. But a review should have this extra level that you really go a little bit beyond. And if you need this specific area of expertise, then it's good to invite another co-author that can look at, for example, in this case, it would have been good to have somebody who could have judged on the pathogenicity of these variants, because then half of the cases would have gone.

 And so these are the things where I typically take a look. Does this make sense? Does it have all the expertise that's needed?
 

[00:06:47] Dr. Shweta Prasad: Do you at any time feel that it is fair, or what was the circumstance wherein you could, possibly ask somebody who's written a narrative review to change it to a systematic review?
 

[00:06:57] Dr. Christine Klein: Honestly speaking, yes — because I've done that in the past. So I think I do. Yes, I think, I think that's fair. Again, review articles can be very different.

Let's say they review or certain biological mechanism and maybe there's only six or seven papers around, but still, even then, one would do a comprehensive review of the literature before writing this, right? So These days, it's so easy to access the literature, at least in the English language. So I think that is something you should really demand.

Unless that's a viewpoint but that's different. And then you don't have to cover everything. But for a review article to be balanced and comprehensive, and I think a review should be both balanced and comprehensive, I think you have to be systematic to some extent.
 

[00:07:34] Dr. Shweta Prasad: When we're speaking about systematic reviews, how far back, prior to submission, should the last date of check have been? Because as we discussed, it takes forever to write a review article and suppose you have collected your data. And that data collection has ended, say, 10 months prior to you submitting it. What happens in that interim? Would you expect a reviewer to say no, no, I want the author to check up everything that has gone in between their last date?
 

[00:07:59] Dr. Christine Klein: Again, thank you for this important question. We've been discussing this actually just yesterday among some colleagues of ours. So again, I think the MDS gene reviews series is a good example there, as it takes typically about two years. And so what we do now, when the author start working on it, we already tell them, you will probably need one more update before you can submit. So please, even if they already have like a little bit of an analysis in between so that they know what they're working on and what they can start writing about, you may need to update and so make sure that you have your Excel tables or whatever you work with in good shape so that you can then later, just before submission, adapt and update the numbers. And for us as a rule of thumb, there's no firm rule, obviously, but we say it's 6 months.

 I probably would say okay until a year, perhaps, but otherwise I think it's getting outdated. Somewhere between six months and a year, would be my personal feel and deadline. 
 

[00:08:50] Dr. Shweta Prasad: Finally, how does one actually ascertain the real importance and novelty of a submitted review article? I mean, everything may be well-written et cetera, but what makes it really important?
 

[00:09:01] Dr. Christine Klein: Yeah, again, a very good question and I think no, really easy answer to this. I think what can be very important is when the topic is novel, and people don't really know about it yet. Or something that has emerged, but nobody really has put it together.

So what should a review do? So I think it should do two things. First of all, it should be interesting. So that's number one. Second, it should guide two groups of people: readers that are not familiar with the topic, so that they can have a good read, and they now have reading the review, they have at least basic understanding of the topic. At the same time. I think it should be interesting, ideally, also to people working in the field. So that even they get something new out of it. That's when it comes to the supplementary tables and the interesting materials that they hopefully have worked on. That's also something I judge, is it really interesting to these two types of audiences, which I think almost any reviews should cater to.

And so first of all, if somebody really finds some new topic or sometimes topics are already there, but they need to be made accessible. So that I think is important.

 I also think it's important when something changes, or we are having the first really systematic review that may change the view of a field where we thought we know already how everything is, but then it turns out no, it's not the case. People don't respond to this drug or they don't actually have this more often than others .

I think it's important also when a deals just with an important topic. So for example, The Lancet, I think every five years they publish a review article on Parkinson's disease. And I think that's in itself a good idea, because it's an incredibly important and growing disease. And therefore, I think even that is a good argument to write a review, just to keep people updated on this very important disease.

I think those are probably the things that I find, but still I'm not against good reviews, reviewing something that has already been reviewed. So I think there's a room for those as well.
 

[00:10:42] Dr. Shweta Prasad: With that we end this episode of our special podcast series. Thank you very much, Professor Klein, for sharing all your insights on reviewing a review article.

In the next episode of the series, we'll be discussing the role of ethics in the peer reviewing process. Thank you so much.