High marks: CT angiography research improving STARD compliance

Journals that have adopted the Standards for Reporting of Diagnostic Accuracy (STARD) feature CT coronary angiography studies that have been largely successful at complying with manuscript guidelines, according to a study published in the January issue of Radiology.

“[O]ur study results show that the overall compliance with reporting guidelines of studies addressing diagnostic accuracy of CT coronary angiography has improved over time and is moderate to good, with a higher total score for STARD-adopting than nonadopting journals,” wrote Stefan Walther, MD, of Charité Medical School in Berlin, and colleagues.

Because methodological deficiencies can affect diagnostic accuracy estimates, understanding the internal and external validity of a study is essential, explained the authors. In an effort to improve reporting quality, the STARD statement was published in 2003. It contained a checklist of 25 items along with a flowchart to help authors report essential study elements. Examples of checklist items include a thorough description of study population and participant sampling as well as descriptions of how indeterminate results were handled. The full checklist can be viewed at www.stard-statement.org.

The checklist has become a quality assessment tool for reporting and while it can be applied to a broad variety of diagnostic accuracy studies, Walther and colleagues focused on prospective diagnostic accuracy studies of CT versus conventional coronary angiography for the evaluation of coronary arteries.

A total of 130 studies from 44 scientific journals were included in the analysis. Literature review included a search of the MEDLINE, EMBASE and Institute for Scientific Information Web of Science databases.

A modified version of the STARD checklist, containing only 21 of the original 25 items, was used in the study to accommodate the inclusion criteria used by the authors.

Results showed that the mean STARD score was 14.4 on the modified 21-point checklist. Overall, 92 percent of the studies reported on more than half of the STARD items, with individual item compliance varying between 17 percent and 97 percent. From 2003 to 2011, the average STARD score of studies increased 0.3 points per year.

As would be expected, STARD-adopting journals, such as Radiology, had a statistically significantly higher STARD scores than nonadopting journals, at 15.4 versus 14.1, respectively.

Walther and colleagues noted that there was still room for improvement in applying the STARD guidelines to studies of noninvasive coronary angiography. Only a few authors explicitly stated they prepared their manuscripts according to STARD criteria and only 16 percent of the journals explicitly asked authors to use STARD.

“Journals that have not adopted the statement yet should include the STARD statement in their instructions to authors and expect them to use the checklist,” wrote the Walther and colleagues.