Thursday, February 18, 2016

Can qualitative research help us to avoid hitting the wall in biomedical research?

Like many people I assume, I have been looking the recent exchanges between Prof. Trisha Greenhalg (Oxford, UK, see here her letter ) and the editors of BMJ (their answer can be found here ).

Obviously, this debate appeals greatly to me, as I am both a researcher heavily relying on mixed (quantitative/qualitative) approaches, and an Associate Editor for a journal ( Computers in Human Behavior ) which covers a field in which both quantitative and qualitative studies are to be found (as a disclaimer, the opinions stated below are mine and not those of the whole editorial board of the journal – although I am quite sure that most would share my views on this matter). We actually had quite a few discussions about this exact topic in the last few months with the students in the lab.

Anyway, I think if we take biomedical research as a whole, we all would agree that we are currently “hitting a wall”. Despite a sheer number of research groups which has probably never been so high, most of the research currently published does sadly not translate into major advances to patients’ bedside. We are accumulating studies after studies without being able to truly see a measurable impact on treatment outcomes, on efficiency of public health systems, and, ultimately, on patients’ health. Not that we are not progressing, rather, the progresses are slower than what we should be expecting.

In order not to hit this wall, we might need a paradigm shift. Could this paradigm shift be related to qualitative methods? Indeed, there are increasing numbers of researchers who turn back to qualitative methods to document various health-related phenomena. Qualitative research has always been around, and sadly, it always has been opposed to quantitative research. Who is guilty? I am not actually sure. Of course, “hardcore” scientists will all claim that qualitative research has no value, but I have heard quite a few times qualitative researchers saying the exact same about quantitative research. In both cases, such statements are (sorry to say, but it is true) stupid. What makes the value of a methodology, is its adequacy to answer the question we ask. If answering a question needs quantitative analyses, fine. If it would require another perspective – one that could be brought using qualitative methods – then so should it be.

As I just mentioned, I think that since a few years, more and more people are seeing that we were missing something by focusing only on large-scale measurable data. Nonetheless, the perception of qualitative research by a lot of journal editors (not to mention funding agencies) is still rather negative.

Given that as a researcher I am regularly doing research using mixed approaches, as an Associate Editor, I am obviously highly sympathetic to studies relying on such methodologies. In other words, I am not immediately turning down a study because it is solely a qualitative study. Does that mean that I will consider that any qualitative study is inherently good? Certainly not. Sadly, here come some of the issues that we are facing with qualitative research. Some people believe that doing qualitative research is a good way to do research when one doesn’t want to deal with statistics or complex tools. But in qualitative research, there is the word “research”. And indeed, qualitative research is a form of research, with its own methods, limits, potential biases, and so on. Furthermore, it is a field which is – like any given field – evolving. Unfortunately, not all qualitative researchers understand that: a common response I get from authors once their papers have been rejected by reviewers for methodological concerns (and here, I want to state again that I am familiar with such methods, so I make a lot of efforts to try to get reviewers which are competent in qualitative methods when such a paper is submitted), is typically going like that “in his paper, X said in 1980 that this approach is appropriate”. Sure, but we are in 2016 … let me count, 1980 was (my, I am getting old) 36 years ago. Your field is so numb that it did not evolve at all during 36 years? Probably not, update your methods. Second point, not all qualitative approaches are appropriate for all situations. Sometimes, a qualitative strategy was appropriate for a given situation, population, whatever, but is not for another one. Another point to consider regarding methods (and the necessary updates in qualitative methods). Back in the years, when I was younger, we were able to publish papers in mice behavior with just one behavioral test, and one experimental drug. And we were able to publish that in high impact factor journals at these times. Now, just to get in a moderate impact factor journal, you would need a dozen of behavioral tests, the same amount of drugs, and a bit of molecular biology on the top to get some “mechanistic explanation”. Well, the methodological quality requirements of quantitative research have jumped up (I know it is not the case in such magnitude for all fields, but still). So, it is legitimate to assume that people are expecting qualitative research methodology to be updated too. And some qualitative research do just that. Those papers should indeed be published, and receive the attention they deserve. Which leads me to the next point.

Not all research is good. Of course, there are a lot of bad quantitative research too. But in the case of bad quantitative research, it might create a (transitory) illusion: bad quantitative research does not necessarily look instantaneously bad. If you read it quickly, you might think it is plausible research. This is not the case for qualitative research: badly done qualitative research usually looks bad. So, from an editor point-of-view, when a paper goes to an editor, a “bad quantitative research” might pass the first quick screening and be sent to review, while a bad qualitative research will probably immediately get blocked by the editor – and not sent through the reviewers. That doesn’t mean that the bad quantitative research will pass (usually it doesn’t). That might give a bad feeling to some qualitative researchers. But trust me my friends, the proportion of quantitative research getting rejected is very high too ! When I say that, I am CERTAINLY NOT saying that all qualitative research papers which get rejected right by the editor are bad. Indeed, I fully agree with the statement of the letter of Prof. Greenhalg et al., that editors need to be educated in qualitative methods (if they are, then they will select appropriate reviewers). But with the working load we are all having, I understand an editor unfamiliar with qualitative research rejecting systematically qualitative papers after having had a brunch of them rejected by the two reviewers, with rough comments sent to the editor (“why the hell are you sending me that to review?”). Rather than forcing journals to accept one qualitative paper a month (quotas are not always the best way to do things: a study should be published if it is good, not because we need one per month), a way to change this perspective could be to include in the editorial boards of major journals some specialists of qualitative or mixed methods - so that GOOD qualitative research papers are given a fair chance and can undergo a fair peer-review process.

I have seen the argument that “qualitative papers don’t get cited”. Well, that is obviously not true. GOOD qualitative papers get cited. Like any paper, it is impossible to predict the fate of a paper in terms of citations just by defining its type. If decades ago, a review was sure to attract a fair amount of citation, with the considerable increase of the number of papers which get published on a daily basis, reviews are not necessarily more cited than experimental papers nowadays. While of course, I understand that some editors might consider the potential of citation in the decision process, it is and should not be the main point to consider. The main point is the quality of the study, its value for the field. And with that in mind, qualitative studies, when well done, can be of considerable value.

Finally, a question that I am asking myself quite regularly: is this qualitative/quantitative debate a real debate? I am actually not sure. I believe that we should be able to go beyond those old ghosts, and all (editors, reviewers, but also authors) become mature enough to understand that if we want to make real changes in patients’ life, the “truth” need to be approached by different, yet complementary angles. Depending on what we want to explore, we might need time to time quantitative tools, time to time qualitative tools, and I believe more often, both at the same time. The insights of qualitative approaches are incredibly valuable. But so are the data obtained through quantitative analyses. In my opinion, only combining both will allow us to crush this damn wall, and to go further.