Imaging studies with positive titles, findings cited more frequently

Studies focused on diagnostic accuracy with positive titles or findings are cited more frequently in imaging journals than those with neutral or negative titles or findings, according to new research published by the American Journal of Roentgenology.

The authors explored diagnostic accuracy-related research published by radiology, nuclear medicine and medical imaging journals from January 2005 to April 2016. Retracted studies and those with no abstract were excluded.

Overall, the authors analyzed more than 900 studies, including 94 meta-analyses. Sixty-one percent appeared in imaging journals, with RSNA’s Radiology journal being the most common source. The number of citations for each study ranged from zero to 1,292. The mean number of citations was 69.

The mean citation rate was 0.66 per month for studies with positive titles, 0.50 for studies with neutral titles and 0.06 for studies with negative titles. The mean citation rate was 0.54 per month for studies with positive conclusions, 0.42 for studies with neutral conclusions and 0.34 for studies with negative conclusions.

So what does this mean for radiology as a whole?

“Preferential citation in the imaging diagnostic accuracy literature has adverse implications for the field of diagnostic imaging and clinical practice,” wrote Lee Treanor, department of radiology at the University of Ottawa, and colleagues. “To be thoroughly informed on a topic, clinicians and radiologists require balanced exposure to all relevant evidence. Well-conducted systematic reviews of diagnostic accuracy studies will consider all relevant articles on a topic.”

In addition, the authors explained, this preferential citation can lead to certain studies being overrepresented in journals and the “perceived accuracy or usefulness of a diagnostic test may be overestimated.”

“Inflation in the perceived accuracy or usefulness of a test may cause clinicians to be overconfident in the diagnostic utility of a test and undervalue the need for a more appropriate test to be used or developed,” Treanor et al. wrote. “Ultimately, the increase in the perceived utility of a test may lead to suboptimal patient outcomes such as misdiagnosis.”

Stakeholders should be aware of these findings and keep them in mind when reading research related to diagnostic accuracy, the authors concluded, calling it a “potentially harmful phenomenon.”