In case you wondered what’s wrong with biomedical research, just read this editorial on data sharing by Longo and Drazen in the New England Journal of Medicine, a leading journal in the field. What you will find is a desperate attempt to take data hostage and to enforce co-authorships for people who didn’t make any intellectual contributions.
But let’s take it one step at a time. What did Longo and Drazen actually say? They think there are major problems with sharing data fully, timely and openly.
The first concern is that someone not involved in the generation and collection of the data may not understand the choices made in defining the parameters.
I am an experienced biomedical researcher. If I completely misunderstand the choices you made in designing your study, then you failed to explain them well in your paper.
One more reason you should make all your data and patient information available, so people not involved in your research can unbiasedly assess the choices you have made.
Special problems arise if data are to be combined from independent studies and considered comparable. How heterogeneous were the study populations? Were the eligibility criteria the same? Can it be assumed that the differences in study populations, data collection and analysis, and treatments, both protocol-specified and unspecified, can be ignored?
These are very good questions and they are exactly what people like me worry about.
We use statistics together with decades of experience in data analysis. And this is why you need to share your data and all information you have about the patients – because else no one can answer the questions you just asked.
Many journals also have this thing called “peer review” where scientists, who often were not involved in the original study nor in my analyses, comment on my work and help me to fix potential flaws. I am sure the NEJM has something similar …
The parasites are here already
So far, so good. They haven’t insulted anybody yet .. Let’s see how it goes on …
A second concern held by some is that a new class of research person will emerge — people who had nothing to do with the design and execution of the study but use another group’s data for their own ends, possibly stealing from the research productivity planned by the data gatherers, or even use the data to try to disprove what the original investigators had posited.
What do you mean by “will emerge”? No, these people are already here, I am one of them!
I use other groups’ data to my own ends. I often address questions the data gatherers might not even have thought about.
And yes, if the original claims are shaky, I will do my best to use their own data to disprove them. This is called science.
And I am not stealing anything. Your data have been publicly funded. I am part of the public. I take what is my own. If you are not productive, don’t blame it on me.
There is concern among some front-line researchers that the system will be taken over by what some researchers have characterized as “research parasites.”
“Some researchers”, again. How very specific.
I am a top-notch, cutting-edge, front-line biomedical researcher myself and I hope to run into them one day and give them a tour of my parasitic research program.
Parasites versus symbiotes
Do Longo and Drazen have anything positive to say? Yes, they do. They outline how they think data sharing would work best:
We think it should happen symbiotically, not parasitically.
Start with a novel idea, one that is not an obvious extension of the reported work.
Second, identify potential collaborators whose collected data may be useful in assessing the hypothesis and propose a collaboration.
Third, work together to test the new hypothesis.
Fourth, report the new findings with relevant coauthorship to acknowledge both the group that proposed the new idea and the investigative group that accrued the data that allowed it to be tested. [my emphasis]
There is nothing wrong with the idea of symbiotic research. I do it all the time. Go ask Bruce Ponder, Carlos Caldas or James Brenton. They are PIs in my institute, we share postdocs and students, and our groups work symbiotically every single day. And in the end we write papers together as co-authors, because we all contributed intellectually to the study, not because one of use owns the data.
Let me speculate a bit. I believe the real reason the NEJM editorial has been written is hidden in the fourth point of the quote above: It is to argue that the initial data gatherers should be co-authors on all follow-up studies. Implicit in their proposal is the assumption that the data are not available to me unless I collaborate with the data gatherers, who retain full control over it. I can “buy into” their data only by making them coauthors. Longo and Drazen argue for a system in which data are held hostage to enforce co-authorships.
When I use published data, I acknowledge the efforts of the initial data gatherers by citing the original paper. And that is exactly how it should be done.
If you call my approach ‘parasitic research’ I am proud to be a parasite.
16 thoughts on “I am a research parasite. Got a problem with that?”
Good article. Could you please give opinion on possibility for self publishing in science? Thank you!
I have no experience with self publishing. There are many honorable journals out there. You don’t have to publish in NEJM. Try PLOS.
It just occurred to me that the editorial is not very clear about what type of data they talk about (another weakness).
My argument using public funding works for most types of data (and certainly all my lab is working with). But maybe the two MDs writing for the NEJM were thinking of drug trials completely funded by pharma (i.e. not at all through grants and other public funding mechanisms).
However, even then there is a strong argument for making all data public. The greater the commercial interests are, the greater scrutiny has to be. Pharma companies are just as likely to lie as car companies.
Most of public financed data are available, but there is still no transparency in actual data researchers are using. As well if you use different statistical engine result can be a bit different, which can make quite a difference in models after (if you are using raw data none of them is perfect, especially long strings over years) as well company research is only published for self marketing it is issue of objectivity there, who will publish results that are bad, you do not have approach to actual raw data as that is, in that case private property, maybe you could buy them but I have big doubt about that. Thing about self publishing is pure curiosity, and theory of avoiding parasitism 🙂 in system (if you have to add someone just because your institution would not pay in case that you are solely author), all thing about amount that some papers are charging to publish your product (should not be other way?! that researchers are paid to publish?) maybe than they would be more eager to go out with full research, as well issue of ego is one more problem, and idea sharing (what if someone makes better thing than you from your idea?, me personally I do not mind, but some could be against full transparency). As well if you self publish it could be faster and more available to many, question is would scientific community count self publishing as much as publishing in some research papers (my thought that it should in case that self publish paper has good citation and visibility and make influence on further research). I would agree that there are issues with all publishing system right now and data exchange but that is more up to researchers and those who influence what can be published (there is still censorship even if we think that all info is out there). Most of researchers are doing secondary research, and you can make good career checking and proving validity of some theories and methodologies, …., over all if one did not sit beside me and dictate sentences or was involved intellectually in paper writing giving ideas, or did not work on actual research (no matter even if it is secondary you still have lot to do) ,… should not be stated as coauthor, you can put in acknowledgment all people that influenced your work but not as coauthors
A very fast and cheap way of ‘self-publishing’ is arXiv.org and bioRxiv.org – widely visible, but not peer reviewed. I hadn’t thought of these earlier …
The problem with self-publishing is that there is no peer review which, for most researchers, makes the article worthless as its too difficult to fact check
“The problem with self-publishing is that there is no peer review which, for most researchers, makes the article worthless as its too difficult to fact check” with todays search engines do not see why? And those who are checking experiments they do not actually do them themselves you are only relying on idea that they know what you are talking about,…, that bring issue of new stuff ..best check is community itself and researchers that are doing secondary research, I just find it would be more energy in research community if ideas would be tossed and turned more, and I am anyway checking even published stuff and searching for papers that they based they conclusions and theories. Only honest and real fact checkup with experiments is to do them your self and that is secondary research based on assumptions and theories of someone’s primary research….for example how peer review knows your statements are facts? They have to do step by step checking of each single paper, like checking of thesis, a I see good compromise would be that all self published things do have after publishing peer review where research community could call you back to explain facts that are not correct or clear, with modern technology engines would be easy to vote paper up or down with positive or negative critic 🙂 just think that would give more life to research community and more possibility for interdisciplinary research and collaboration,…, you always need to be critic and self critic 🙂 and question everything and look behind, figuring out how things work, is not that simple way of being a scientist? In setup as it is now I think lots of great stuff has been lost ….
Good morning. Having slept over it, I now worry that Longo and Drazen’s fault is not saying something stupid, but being overly HONEST. What if they just openly represent the latent opinions in the clinical community, who are the core readership of NEJM? Making fun of the editorial on twitter will not help … no matter how much #researchparasites is trending.
I agree that they are probably mostly giving honest voice to the latent opinions in the clinical community.
However, this just illustrates that the clinical community has latent opinions that are dangerously anti-scientific. If enlightened, Longo and Drazen should vigorously repudiate this view. If not, they still deserve to be chastened for giving voice, uncontested, to such counterproductive and even offensive points of view.
“However, this just illustrates that the clinical community has latent opinions that are dangerously anti-scientific.”
That is an old story in biomedicine. Even after Drake used fruit juice very successfully against scurvy in his voyage round the world, and even after James Lind did the first clinical studies with citrus fruits in the 18th century, the Physicians the Admiralty listened to were those at Court. Their opinions held sway till after the Napoleonic Wars. Their opinions?
“This is a long-exploded theory that distracts from treatment of scurvy through abating bad air and providing sunshine for the afflicted. Tangerines and oranges are sweet and tart, but of no use to scurvy-afflicted seamen of his Majesty!”
Power at Court (the “Court” of the committee rooms of congressional committee chairs) still seems to be the dominant interest of far too many in far too many disciplines. As long as the federal government has a monopsony over the buying of research, this will continue. Break the monopsony! Doing that is not getting much support from academia, because they are the first beneficiaries of the monopsony, since it depends so heavily on “certification”. Only when the buying of research by the federal government in a field drops below 50%, with the rest from a wide diversity of funders, will there be lowered skepticism among the rest of us.
The one word I did *not* hear in this article is *replication*! That is the other half of freeing science from the grip of “Court” influence. Replication is the gold standard that peer-review advocates aspired to replace, and failed. The data must be gathered *at*least* twice, and the analysi done *at*least* twice for reputable scince. Yes, once the committee chair has funded a study he’ll frown on funding a replication, but that is what we need a wide range of funders for!
So many studies get published where the results are positive, but barely so. Confidence levels that hardly make you confident. This would certainly make meta-analyses much more useful because you could actually combine data at a more significant level. My only concern is that researchers will delay publishing their data till they have minded for every conceivable paper and idea because they will be afraid that others will use their data and published before them
Translation to plain English of selected portions of Longo and Drazen’s editorial on data sharing
I like this!