Inconsistencies Between Abstracts and Manuscripts in Published Studies

Members do not see these Ads. Sign Up.

Those who follow my running blog will have noticed how often I comment about the abstract not matching the data in the actual study or the study not supporting the conclusions made in the abstract.

I picked up this abstract from a tweet by Adam Meakins that struck a chord!

Inconsistencies Between Abstracts and Manuscripts in Published Studies About Lumbar Spine Surgery
Lehmen, Jeff A. MD; Deering, Rachel M. MPH; Simpson, Andrew K. MD; Carrier, Charles S. BA; Bono, Christopher M. MD
Spine Feb 2014

 

I wonder if there is a direction to the discrepancies, ie do the abstract generally suggest a more significant finding than the manuscript?

My thinking is that the abstract frequently represents a marketing tool either to promote the manuscript or the journal in which it appears. As an advert it is almost an obligatory requirement that it stretches and distorts, without of course being frankly dishonest, the contents, findings, etc in a direction that will increase sales or revenue at some other future point.

Bill

 

The biggest error I see is an overstatement of the strength of the conclusion or the concluding statements are not supported by the actual results.

 

When I review papers submitted for publication to scientific journals (I currently read papers for about four journals), I pay careful attention to the abstract to make sure it reflects what was found in the manuscript and does not overstate or embellish the findings of the research. I have often had to tell the authors that they must change their abstracts in order to get the paper published due to inappropriate comments made in the abstract. Since the abstract will be the most widely read part of the whole paper, I believe that every reviewer of journals should also be so concerned with the accuracy of the paper's abstract and not simply skim through it. I see the problem as more of a lack of good and thorough peer review and not keeping the authors of papers totally honest in regard to their scientific conclusions contained within the abstract. It is the journal's responsibility to see that this type of "abstract embellishment" does not occur.

 

As a child I remember going the cinema each Saturday to see the follow up of the latest cliff hanger serial. At the end of each episode the hero would in in a 'cliff hanger' situation. In some cases he actually went over the cliff at the end of the episode but at the beginning of the next episode he was miraculously still moving towards the edge of the cliff and was always saved before he went over the edge. Although I hated this deception it brought me back week after week.

Are we seeing the same thing with respect to abstracts and manuscripts?

I see the problem as more of a lack of good and thorough peer review and not keeping the authors of papers totally honest in regard to their scientific conclusions. It is the journal's responsibility to see that this type of thing does not occur.

Why do you think that the authors are choosing not to be totally honest re-their scientific conclusions?

When I review papers submitted for publication to scientific journals (I currently read papers for about four journals), I pay careful attention to the abstract to make sure it reflects what was found in the manuscript and does not overstate or embellish the findings of the research.

I would think that the above is the essential function of the reviewer and the same question arises. Why is the reviewer is not ensuring that the abstract is an accurate relection of the manuscript?

Bill

 

It's OK to focus on "abstract embellishment" and to assume that the manuscript represents honesty but when the abstract doesn't match with the manuscript how much more doubt should be cast on the veracity of the manuscript, eg how much have the recorded results been massaged?

In a sense by ensuring that the abstract corresponds with the manuscript the reviwer may in a genuine attempt to ensure honesty have become complicit in fraud.

If the abstract doesn't correspond with the manuscript this should be noted by the reviewer and if the manuscript is published the discord between abstract and manuscrpit should be published with it as a note.

Back to the original question what does the author, publisher, etc gain when the abstract offers more than the manuscript delivers?

Bill

 

Although in the 20% percent or so of cases where the abstract and manuscript are not consistent the majority of inconsistencies are 'trivial' they are present as are the more serious inconsistencies.

As I pointed out in the previous post correspondence between the abstract and original manuscript does not necessarily represent honesty although it might suggest honesty. Conversely when the abstract and manuscript do not correspond it suggests dishonesty, carelessness, oversight,etc. However, whatever the cause of the discrepancy/ies, is it reasonable to assume that, in these cases at least, the same characteristics do not apply to the original manuscript.

I would think that replication of the original research is one of the few comprehensive methods of identifying inaccuracy in the original research and I wondered what percentage of studies were replicated?

Bill