6.6- Interview with Dr. Gary Friedman

So welcome, in this next module I’m going to be talking with Dr. Gary Friedman. I’m thrilled to have him here today. I’ve mentioned his name a few times in this course. He provided a number of the examples that I’ve shared with you and he’s been a long time champion of good writing in the medical literature. So it’s great to have him here today. He has been an editor at the American Journal of Epidemiology for over two decades. And was the director of the division of research at Kaiser Permanente, Northern California for almost a decade. And also he’s on the editorial board of a number of other journals and currently a consulting professor at Stanford. So I really appreciate you being here today, Gary.

It’s a pleasure.

So I’m just going to start by asking, obviously we need to do good science, but besides good science, what are some of the key elements that journal editors are looking for when they get a paper?

Well, one is novelty. We will often reject a paper without even sending it out for peer review, if it’s just a repeat of findings that are already well known. I mean if you’re writing the 30th paper on alcohol and how alcohol is related to high blood pressure, or obesity is related to mortality, don’t expect people to be very enthusiastic about it. So we will often reject a paper right away. Now let me just tell you the process that we go through with the American Journal of Epidemiology. First it’s looked at by the editor in chief, and then he assigns it to one of the other editors, who really makes the final decision. And I happen play that role as a general epidemiology editor. There are other more specialized editors that work just in cancer or just in cardiovascular disease or just in infectious disease, but I tend to be more general. And so, if the two of us think that the paper is not going to be of sufficient priority to be accepted, even if it’s good science and well-reviewed. Then as a favor to the authors, we won’t delay a rejection of it. Besides novelty, another criteria that we look at is interest to the readers of this particular journal. For example, sometimes we’ll get a paper that really is not so much an epidemiologic study but a way to improve public health. And we will decide, this is really more suitable to a public health journal than to the American Journal of Epidemiology. We also look for good writing. And this can be especially a problem with a non English speaking author. Usually when a paper comes from a foreign country, often the English is really bad. And we know it’s going to need a lot of work. And if we think it’s a good study we will often send it back and say, please have a native English speaker go through this and edit it. And occasionally, this comes when a foreign person whose native language is not English works in an American institution and submits a paper. Usually that is not a problem, but it can be. Another factor is that, don’t make the paper too long, don’t put in too much excess material that’s not needed. And another important problem is what we call slicing the salami too think.

Yeah, yeah, that would be great if you could say a bit about that.

If you have a study that involves both men and women and want to try to get two papers out of it. Say, here’s the effect of this risk factor in men and then another paper, here’s the the effect of this risk factor in women. That is really frowned upon by journals. And my own personal experience is that I had to deal with this as an author. A colleague and I each wrote a paper about the health effects of cigar smoking. This was at Kaiser Permanente. This was on cardiovascular disease, mine was on cancer. I submitted mine to the Annals of Epidemiology, it was accepted. He submitted his to the New England Journal, it was accepted but they said, do you have something on cancer? So naturally he was younger and needed publications more than I did. So I sort of fell on my sword, withdrew my paper from the annals of epidemiology, and combined it with his, and the New England Journal took it. So even I was faced with a salami problem as an author, so those are the main things.

Yeah, sorting those things out early can be very helpful. And what do you think is, if you had to point to one thing, the number one mistake that scientists make when submitting their paper for publication?

I think it’s overconfidence in how important and how good their study is.

You’ve worked on it really hard-

Yeah.

And you’ve done your best to write it up. And this was a topic or a question that was really important to you, and you think, gosh, this is really an important finding but others may not value it so high. So if it gets rejected due to some factor like that, I would persist. I think it’s important to persist and submit it elsewhere, fix it and submit it elsewhere.

Yeah, finding the right journal for it might be the key, yeah.

Right, because you’ve devoted a year or two to this project and you think it’s really important, don’t assume that other people will place it as high priority as you do.

Yeah, that’s a good point.

And I have to, of course, ask you to give some advice to authors about writing style because you’ve written a number of editorials in the American Journal of Epidemiology dealing with writing. So what advice do you have in terms of the writing?

Well I guess clarity and conciseness are what I would put as the highest priority characteristics of writing. Avoid repetition. Sometimes people will say something in the introduction and then repeat it in the discussion. Avoid that, or they’ll say something in the methods section, which gets repeated in results. So avoid repetition, avoid excess verbiage. And I’ve written editorials about that, which I think you’ve mentioned.

I have, yes.

And avoid, sometimes people will write a rather long discussion in the introduction of the paper. Whereas really, the introduction should be brief. Say why you did this study and basically what you tried to accomplish. Leave all the discussion, the review of the literature, for the discussion section. Don’t try to put too much in the introduction. Some things, certainly you need some background, as to why you did the study, which may involve citing some previous literature, but don’t go into it in detail, that’s in the introduction. And don’t repeat numerical data in the text. You’ll have a table, which has nice numerical data with and ratios, and so on, and then in the results, Table one shows the blah, blah, blah, and the odds ratio was this and that. So avoid repeating. We will often ask authors just describe the finding of the table in words, and the reader can see the numerical examples.

Good, yeah.

So those are my main suggestions.

Those are good tips, yeah. And one of the fears I think that comes up often when I am teaching a course like this and has already come up in this class is a lot of times they will say to me, I’ve been teaching them to write in simple and concise language. And they’ll say well if I Simplify it too much, I’m going to dumb down the science, it’s going to lose precision. And so there’s this kind of fear that a lot of scientists have, so I was hoping you could comment on that .

Well, I don’t think you’ll necessarily dumb down the science. Of course, you do want precision and clarity, but I think you should avoid jargon and things that are really hard to understand. George Comstock, who’s no longer alive, but he used to be the editor and chief of the American Journal of Epidemiology, he used to say that, papers should be understandable by someone with a good college education or someone who reads the scientific American. It does not have to be full of jargon that’s only understood by your peers in your particular specialty.

Yeah.

So we welcome things that are very readable, and easy to read. I would not have that fear at all.

Yeah, and in terms of getting published, when you get two papers in, let’s say, they were the same science and one was very jargony and sounds sort of the way. Scientists think they have to write, and you’d have one that’s clear. Can you comment on your chance of publication?

Well, if both of them get peer reviewed and very highly rated, and they seem important to the editors.

Yeah.

We might ask, to please explain what you mean by this word or this sentence.

Yeah, yeah.

So, it will take an extra step if it’s really hard to understand and not clear to most readers.

Yeah. Yeah, and to have, what advice can you give specifically to first-time authors? I imagine a lot of the students in this course are early in their career and haven’t had a chance to get published yet or in the process of submitting their first papers.

What bothers me most is when I get a paper, it’ll often be from a really good institution and the senior author at the end of the list is someone who’s well known, but it’s clear that it’s been written by a student.

And it sounds like a thesis, it sounds like a PhD thesis or master’s thesis.

Yes. Which, and it contains every conceivable strength of the study, every conceivable weakness.

And so to prove in writing the thesis that they learned what they were supposed to learn in the school public health or in their graduate program. So, Write as if you were a more established senior scientist who knows what other scientists understand and you don’t have to explain something to everything. I just happened to see a paper this morning, a latest issue of the American Journal of Epidemiology that was going over the strengths and limitations of their study. And it was a study about mortality in Whites and African Americans related to a certain factor; and it was saying one of the strengths of our study is that we had a large African American population as part of the study group. Well, that’s sort of obvious, you wouldn’t-

Right, right.

You wouldn’t be writing about mortality and so you don’t have to say that, I mean. Things that are sort of obvious, don’t say, just say the important limitations of your study, the important strengths. And you don’t have to label them as strengths and limit. It’s there’s sort of a jargon now, every academiologic paper you have to find the word limitation.

Yes.

Just state what they are. So, I guess just avoid making your paper seem like a thesis.

Yeah, I’ve encountered that as well in reviewing papers.

Yeah

Where you could tell that somebody had taken a thesis and just tried to turn it into a paper, it’s a very different product.

Right.

Paying attention to actually writing it like a paper or a thesis, is a great tip And let’s talk a little bit about resubmission. So let’s say that you are offered the opportunity, the paper’s rejected but you’re offered the opportunity to resubmit with a lot of comments from reviewers. What kinds of tips can you give people who are in that stage of the publication process?

Well, first of all, although acceptance of the revision is not guaranteed, and that will say so in the letter, I think you should take that as encouraging. The fact that it got passed that first reviewing stage is really a good sign, and you should work hard to improve the paper so that it will get accepted. And the, first of all, whatever the, first of all list each comment that the reviewers make and respond to that. It makes it much easier as an editor to be able to see that the author took into account every comment. That both reviewers or as many reviewers who are involved made, and the editor will often make comments too and list them and respond to each one separately. And either, whatever the comments says, either fix it, fix what the problem is or explain why you don’t think that that criticism is valid or important. Explain why you don’t want to fix it, and give a good argument for that. Show, and also please show the editor where you made the changes in the manuscript. If someone says well, I just changed the method section to reflect this, that’s sort of hard for us to find what exactly you did. So, list the changes when you respond to the comments. Either highlight them in the manuscript, see where the occur, highlight them or use track changes. Apply in addition to the version where you get rid of the track changes. Also provide the version that has track changes so that the editor can see exactly what you did.

Yeah.

Or sometimes the paper will be sent out for further peer review, sometimes the peer reviewers have really strong concerns. And I as an editor will send it back to them and say did the author really satisfy your concerns about this? And be polite.

You know, this reviewer’s stupid, you know. He didn’t really understand, you say, we thank the reviewers for their constructive criticisms, I’m sure the paper’s improved as a result of this, so that does not hurt.

It’s everybody’s instinct I think to at first be a little defensive when you get your review back. So you have to go back and tweak the language a little bit to make sure you’re being polite, that’s a good tip, yeah. And, can you give us some words of encouragement for young scientists who might have submitted their first paper and actually got an outright rejection. What are some words for encouragement?

Well, I would just say fix it, and if it’s totally rejected by that journal, they’re not asking for re-submission, fix it, try to respond to the reviewer’s comments that you feel are justified, and submit it elsewhere, and persist. As an author, I’ve had to submit papers to as many as four journals until it finally got accepted. So I think persistence is important, don’t take the rejection personally. It’s hard not to and I sometimes do. But you realize that some reviewers May not be very competent, may not really understand what you’re doing. This will often happen if you submit an epidemiologic paper, and it gets reviewed by a clinician who does not really understand principles of epidemiology. They’ll feel that if you write a paper about stroke, if every case was not reviewed by a neurologist, then it’s not a valid paper.

Right, right.

So, or maybe they’re competent, but they just had a bad day.

And didn’t really pay close attention to what you wrote, and missed some passages that really would’ve answered their question. So, just realize that reviewers are human, just like authors, and that’s just the way the system works.

Yeah, yeah.

It’s not perfect.

And then talking about the system a little bit, what changes do you anticipate are going to occur in the publication process? There’s kind of a lot going on now with the online and open access.

Right.

Say a little bit about what’s going to happen for in the future.

Well, I think you’ve described it.

Less paper, more electronic publications. It’ll appear faster, like EPUB before it appears in the paper version, there’ll be more new journals. I keep getting emails from journals that just started up, and they’ll make it sound like it’s your field. And there’s probably not such great peer review with them, they want to get some known authors and publish. But I’ve had a paper which was submitted to a journal that’s supposed to be peer reviewed, and the acceptance came back in about five days.

So I don’t think it was really peer reviewed. Of course, I’m biased, but I thought it was a decent paper. But I don’t think it got the kind of peer review that papers normally get from good journals.

Do you think peer reviews are going to change then somewhat in the future because of this proliferation of online journals?

Say that again, sorry.

Do you think that the peer review process itself is going to change somewhat because of the proliferation of online journals?

Well, it shouldn’t, but it might, I’m not sure how.

Yeah.

Yeah, if they’re really in a hurry to publish, they might say, get your review back in a week, or.

Yeah.

A lot of people just can’t do that.

They can’t do it that fast, yeah.

There was another point I wanted to make that, now NIH has a policy that if your paper is supported by them, as part of a grant or contract, even if the journal wants to charge money for downloading the paper, after a year, NIH will put it on PubMed as a finished manuscript. Which may not have had the kind of copy editing that will finally appear in the journal, but it’ll at least be your final version of the manuscript. And so, you’ll be able to read papers that have been out for a year, for no cost, even if the journal is a kind that wants to charge money for it.

That’s great to know. And if you could change one thing about the publication process, what would you change?

Well, when you asked me that question before, I got two things.

Two is fine, .

First of all, I think there should be greater valuation of negative findings by both researchers and by the journals. When you do studies, you’re going to have this great idea that you think something is a cause or a prevention of some disease, and you’re going to do a study, and it’s not going to pan out. That is the most common experience that we epidemiologists have. In fact, and I’ve now got a graduate student here who is doing a study of Metformin in relation to risk of breast cancer, because there was some thought that this might be preventive agent for reducing the risk of certain kinds of breast cancer. And this is being done at Kaiser Permanente in our division of research. But it’s looking like, and this is not a final result, but it’s looking like that it’s probably not going to turn up what was hoped for, or what was expected. And naturally, the researcher’s very disappointed in that, but this is important to know, too, that there’s a negative finding. So don’t be discouraged if you have a negative finding, it’s important that that get published. And I wish that journals would be more interested in publishing these, sometimes a negative finding, if you want to get it out there, could be written in the form a brief report, which will take less journal space, it will get higher priority. So I think that is my main answer to your question. I also wish that there was greater ease in getting peer reviewers to review a paper. Sometimes, we have a list at the American Journal of Epidemiology, and I send a particular paper out to people who said this was their specialty, or special interest. And out of 10 people who we send it to, only one agreed to review, and we try to get two, and I was asked, can you make a decision on just the one review? because we can’t keep trying to get reviewers. So, as a scientist and an author, please accept the responsibility of being a reviewer, as well as an author. And we’ve had some cases where we’ve published papers by a certain author, and this author consistently refused to review papers, and naturally they’re not well regarded by the editorial staff. So, those would be my two concerns.

Yeah that’s a good point, a good tip. And the opportunity to review is a good way to learn also as an author.

Absolutely.

So, great, is there any parting thoughts or tips that you want to offer to the class?

No, I think we’ve covered it.

Great, okay.

Your questions are so good that you’ve covered everything that I might want to say.

Thank you so much for taking the time to speak with us.

You’re very welcome.