Tuesday, August 5, 2008

Writing an Abstract

Not all that long, the abstract for a scientific paper was a prose paragraph that attempted to provide the 5 W’s of a piece of scientific research: who, where, why, when and what. That is, it attempted to give the reader a general sense of what question was being asked, how the researchers attempted to answer it, what they found and what they felt it meant. But studies done in the late 1980s and early 1990s indicated that the abstract was failing in is goal. Squires (1) found that 56% of abstracts did not report the technical design of a study, 79% gave no information on how subjects were selected, 86% made no mention of limitations and 93% made no recommendations for further study. To resolve this failure, the Ad Hoc Working Group for the Critical Appraisal of the Literature (2) proposed the structured abstract, and that is now the professional norm.

A structured abstract is simply that; an abstract that has a specific structure. In the case of an original data report, the structure can be as simple as: Objective; Methods; Results; Conclusion. It may be slightly expanded, so that instead of a Methods section, there are sections on Subject Selection; Site, Main Outcome Measures, etc. Some journals ask that the author prepare a Background section as well, to precede the Objective section. It is important to keep mindful that there is usually a limit of 250 words to the abstract, so brevity is important. The Background section provides context; Objective provides the intent of the paper; Methods tells what was done and how; Results provide aggregate important findings; and Conclusion briefly tells what was discovered.

Abstracts for literature reviews and case reports differ. For a literature review, the main headings should include: Objective; Study Selection; Data Extraction; Data Synthesis; Results; Conclusion. In this form of abstract, the key is in describing what search terms and databases were used, what information was taken from the yield of papers, and how that information was analyzed. In a case report, the main headings include: Objective; Clinical Features; Intervention and Outcome; Conclusion. Here, clinical features refers to important findings leading to diagnosis; intervention and outcome refers to what was done and how the patient responded.

I should also note that different journals may deviate slightly from these models and as always it is important to read Instructions to Authors. And keep mindful that your abstract is the single most important key to whether people decide to read your paper, so spending a little time to get it right serves you very well indeed.


REFERENCES
1. Squires BP. Abstracts: the need for improvement. Can Med Assoc J 1991;144:421
2. Ad Hoc Working Group for Critical Appraisal of the Medical Literature. A proposal for more informative abstracts of clinical articles. Ann Intern Med 1987;106:598-604

No comments: