How Do I Get a Paper Accepted?—Part 2
this Page, I will present 10 specific recommendations to implement (stratagems) or avoid
(pitfalls) in preparing a paper.
1. Pose a hypothesis. The consensus of the editors is that the best manuscripts
attempt to answer a specific question or achieve a specific goal. If possible, avoid purely
observational or descriptive reports, sometimes referred to as “phenomenology.” Studies
that are primarily descriptive achieve higher priority for publication if they identify a
mechanism or generate a hypothesis.
2. Document novelty. Given the importance of novelty, it follows that good
manuscripts state that they are the first on the topic or emphasize what is new in their
work. If previous studies have been published on the same topic, the editors felt it was
of value to distinguish the results of the present investigation.
3. Describe methodology in detail. The editors agreed that a detailed description
of methodology enhanced the priority for publication. The accuracy of the methods
ought to be validated. Patient acquisition should be addressed in depth, and the
appropriateness of the study group established. The best studies specify how
ascertainment bias was avoided. Control groups are of critical importance. The editors
identified failure to include a control group as one of the most common serious
methodological defects observed in manuscripts. The use of outdated and limited
administrative databases was also recognized as a common flaw.
4. Provide power calculations. All agreed with the importance of providing an
in-depth explanation of how the sample size for a study was calculated. The best
explanations include a specification of the data or assumptions upon which power
calculations are based.
5. Don’t slice the data too thin. The editors emphasized that one comprehensive
paper is much stronger than several smaller ones. Although being too broad and
unfocused results in a lower priority for publication, “salami science,” or the division of
one project into as many “minimal publishable units” as possible is the much more
common error. Given the increasing number of randomized multicenter clinical trials,
this becomes a frequent issue for substudies. In such instances, the best papers make a
strong argument for the importance of the individual subgroups they are reporting.
6. Perform careful analysis. Having asked a novel question and applied excellent
methodology, some manuscripts then suffer from faulty analysis. One of the frequently
encountered problems is the failure to distinguish causality from association. It is
important not to assert causality when only an association has been demonstrated. A
similar pitfall is to fail to recognize and acknowledge the limitations of surrogate end
points, or to assert that a surrogate represents a final end point. Premature ventricular
contractions failed miserably as a surrogate for mortality when their therapy was tested
his is a continuation of last month’s Editor’s Page presenting the consensus
opinions of the Journal of the American College of Cardiology (JACC) editors
regarding how to prepare a manuscript to optimize chances of acceptance. In
Anthony N. DeMaria,
Journal of the American
College of Cardiology
Journal of the American College of Cardiology
© 2007 by the American College of Cardiology Foundation
Published by Elsevier Inc.
Vol. 49, No. 19, 2007