Only one-third of cardiac meeting abstracts get published

Twitter icon
Facebook icon
LinkedIn icon
e-mail icon
Google icon
 - laptop, literature, data

Only about one-third of abstracts presented at the major cardiovascular meetings get published as papers within two years, according to an analysis published online Nov. 2 in Circulation. “The whole peer-review and publication system is getting outdated,” lead author Emil L. Fosbol, MD, PhD, told Cardiovascular Business. “It is too slow for how we share knowledge now.”   

For the study, Fosbol, of the Duke Clinical Research Institute in Durham, N.C., and the Heart Centre at the University Hospital of Copenhagen in Denmark, and colleagues focused on medical meetings by three major associations: the American Heart Association (AHA), the American College of Cardiology (ACC) and the European Society of Cardiology (ESC).

They described the three scientific sessions as the largest and most influential in cardiology. Abstracts at these meetings represent the latest scientific knowledge, but to have an impact on practice, the science typically must percolate into peer-reviewed journals, Fosbol and colleagues noted. But how successfully and quickly that happens is unknown.

The researchers used an automated computer algorithm to search ISI Web Science to find publication of abstracts presented at the three organizations’ annual scientific sessions between 2006 and 2008. They then compared abstract publication rates and journal impact factors.

The AHA presented the most abstracts, at 11,365, they found. ACC had 5,005 abstracts and ESC had 10,838. AHA also had the highest publication rate, with 34.5 percent of its conference abstracts being translated into publications within two years. The rates for ACC and ESC abstracts were 29.5 percent and 27 percent, respectively. Five-year rates for the 2006 abstracts inched up to 49.7 percent for AHA, 42.6 percent for ACC and 37.6 percent for ESC.

“It is not optimal that it takes so long to get published,” Fosbol said. “It loses its relevance.”

Authors of the abstracts at the AHA scientific sessions had 24 percent higher odds of being published than their peers at the other meetings. AHA also accrued the highest mean impact factor, at 4.8 compared with 4 for ACC and 3.9 for ESC.

Fosbol and colleagues calculated that 23.4 percent of publications appeared in journals with impact factors of greater than 10 and the high-impact publications trended toward publication closer to the conference date compared with lower impact publications. The average impact factor declined with time.

The authors cited a number of possible reasons for the lag between abstract presentation and publication: the impact of social media; a slow or over-stringent peer-review system; journals vetting less worthy abstracts; authors trying top-tier journals first and resubmitting elsewhere if rejected; and abstract submissions motivated by a desire to attend a conference rather than to promote science.

Fosbol also pointed to dissemination of results through other media as a possible factor. “More and more, some of these abstracts hit the media,” he said. He speculated that as results get communicated through news media, on websites, via Twitter and other outlets, some researchers may view the findings as now in the public domain. “In a way, they are published.”

Among solutions, the researchers cited a movement within meetings to require authors to send a full manuscript if an abstract is accepted. “That is a good way to get more serious science and also to get the researcher to write the paper and potentially publish it at the same time it is presented,” Fosbol said.

Efforts to publish online first also help to get findings out in a timely manner, as was the experience of Fosbol and colleagues with this particular paper. The authors cited open-access journals as another avenue for swiftly making data and results publicly available.

Fosbol emphasized that not all abstracts presented at scientific meetings ultimately are worthy of publication. “It is not ideal to publish everything,” he said. “But it is not ideal to publish one-third of the abstract information as well.”