Do abstracts in otolaryngology journals report study findings accurately?

OBJECTIVES

Although published abstracts have the advantage of being prominent and concise, their quality in the otolaryngology literature has not been studied. We sought to understand the limitations of relying on abstracts as the sole source of information about published research. In addition, we sought to identify specific opportunities for improving the quality of published articles and their associated abstracts.

METHODS

Studies of original research published in four major otolaryngology journals from January 2008 to June 2008 were included. Each study article and abstract was reviewed separately for 12 domains relating to study design, results, and conclusions. Good inter-rater reliability was established as part of the validation process.

RESULTS

Four hundred eighteen articles were identified for study, which included 75 percent clinical research and 25 percent basic science. The most common omissions in the abstract when compared with the complete article were study limitations (91% left out of abstract), geographic location (79%), confidence intervals (75%), dropouts or losses (62%), and harms and adverse events (44%). Conversely, the abstract often included information about research design (99%), sample size (92%), source of the data (81%), and quantitative results (67%). These results did not differ significantly with regard to article type, journal, or level of evidence.

CONCLUSIONS

Readers of otolaryngology journals may form biased or inappropriate conclusions if they read only the abstract of a study, particularly with regard to study limitations, adverse events, and subject dropouts or losses. These results highlight the perils of using the abstract as a sole source of information.