To identify the prevalence of errors in figure legends in leading ENT journals with regard to the discussion of cross sectional imaging technique as well as the accuracy of the reported radiological findings.

All articles published in 2001 in five ENT journals with the highest impact factors were reviewed for the presence of cross sectional imaging figures. Of the articles containing cross sectional imaging figures the description of the type of modality and the radiological findings were reviewed for their accuracy.

Of the 984 articles reviewed, 185 had cross sectional imaging figures. Thirty-two articles had significant inaccuracies in either the description of the type of modality employed or in the reported radiological findings. Seventeen of the total 509 cross sectional images had an incorrect description of the type of the modality. Incorrect radiological findings were stated in 29 of 509 cross sectional images. These findings identify that 17.3% of the articles had significant errors present, 3.3% of the images had errors in the description of the modality and 5.6% of the images had inaccurate radiological findings. A general trend observed was that the prevalence of errors between journals was inversely related to the percentage of articles that contained cross sectional imaging. The journal with the highest frequency of errors (20%) had the lowest percentage of articles that contained cross sectional imaging figures(13.1%). Conversely, the two journals that had the highest percentage of articles with cross sectional imaging figures (35% and 25.3%) contained the lowest prevalence of errors (15.9% and 15%, respectively).

We found a high prevalence of errors present in five leading ENT journals. We also observed a general indirect relationship between the prevalence of errors and the frequency with which cross sectional imaging figures occurred in these journals.