A reduction in the number of investigators willing to recruit patients into a trial or supervise its conduct in a country will compromise the likelihood of successful completion of large trials that address important questions.
This may be a legitimate concern, but settling for long data embargoes instead of seeking alternative solutions seems a terrible mistake.
we believe there are also risks (e.g., misleading or inaccurate analyses and analyses aimed at unfairly discrediting or undermining the original publication)
These are red herrings. The risk of misleading subsequent analysis is minimal compared to the risk of misleading original conclusions allowed to stand because the data are unavailable for reanalysis.
Data Sharing
This article could have been written in a way to raise similar concerns surrounding data sharing without coming off as nearly so reactionary. Consider:
What Longo and Drazen should have written
Carl T. Bergstrom
At present the allocation of research effort within science seems to exhibit a reasonable balance between data production and data analysis. Traditionally, researchers who generate primary data have been permitted to hoard those data for extended periods of time, to the private advantage of their own research groups. Today, with the frequent calls for data sharing and occasional policies for data deposition, researchers are finding it more difficult to hoard data.
The consequence of this trend is a reduction in the cost of obtaining others’ data, coupled with a reduction in the benefits of producing one’s own. One doesn’t need a PhD in economics to see what the outcome will be: researchers will decrease effort allocated toward generating data, and increase effort toward analyzing it.
We don’t want to allow data hoarding, so we’re stuck with the decreasing cost of data acquisition. If we want to maintain the previous balance of the two activities, we need to find some of way of increasing the rewards associated with data creation. More formal mechanisms of acknowledging and rewarding data use—perhaps something intermediate between citation and authorship—would be a good start.
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.
So what are you going to do about meta-analyses, NEJM?
Will you require that all authors of all original papers be authors of the meta-analysis? The protocol advocated here seems to demand as much.
There is concern among some front-line researchers that the system will be taken over by what some researchers have characterized as “research parasites.”
What elegant rhetoric, the juxtaposition of noble "front-line researchers" against despicable "research parasites".
even use the data to try to disprove what the original investigators had posited
...as is the whole point of doing replicable science.
people who had nothing to do with the design and execution of the study but use another group’s data for their own ends,
This paragraph represents a staggering misunderstanding of how science works. In science, the whole point of publishing data instead of just conclusions is to allow others to verify and replicate one's findings -- whether this supports the original authors "own ends" or not.
This is the first appearance that I know of in Zahavi's writing of the modern sense of a costly signal in which the organism evolves a strategic norm of reaction associating the trait value with the underlying "quality". This is what Grafen (1990) J. Theor. Biol 144:517-546 actually models, and what he terms a "strategic handicap".
A fight for a larger territory must be more difficult than a fight for a small one; hence, males which succeed in occupying large territories should be of a better quality than males which occupy small territories.
Once again, this refers to a revealing handicap rather than a Zahavi's handicap.
Such fights, however small, may further evolve through mate preference to serve as an index of quality.
Now Zahavi has switched -- without any change in terminology -- to discussing what Grafen (1990) calls a revealing handicap.
I suggest that a mature, colourful male has already proved itself to be of a better quality (than one with cryptic plumage) since it has already withstood the extra predation risk involved in its plumage.
Here again, Zahavi is referring to what Grafen (1990) terms a "Zahavi's handicap".
uality in
I cannot see how to interpret this figure as anything but a "Zahavi's handicap" in Grafen's (1990) sense. If so, the figure fails to capture the stochastic nature of selection against lower-quality individuals with larger handicaps.
dividual. An individual with a well developed sexually selected character, is an individual which has survived a test. A female which could discriminate a male possess- ing a sexually selected character, from one without it, can discriminate between a male which has passed a test and one which has not been tested. The more developed the character the more severe was the test. Females
Grafen (1990) J. Theor. Biol. 144:517-546 helpfully distinguishes among several possible interpretations of the handicap idea. Here Zahavi is referring to what Grafen calls "Zahavi's handicap", the idea that selection in real time weeds out handicap-displaying individuals of lower quality.