Characteristics and Quality of Radiologic Randomized Controlled Trials: A Bibliometric Analysis Between 1995 and 2014.

Abstract

The aim of this bibliometric study was to assess the characteristics and quality of radiologic randomized controlled trials (RCTs) over the past 20 years. A PubMed search was conducted to identify radiologic RCTs (defined as RCTs in which the first author or corresponding author is affiliated with a radiology department) published between 1995 and 2014. The following information was extracted from each article: journal, radiologic subspecialty, imaging technique, number of subjects, study result, funding, number of authors, number of institutions, country of origin, and methodologic quality (assessed using the Jadad scale). A total of 358 radiologic RCTs were published between 1995 and 2014. Dramatic increases in the numbers of radiologic RCTs were found, from 43 conducted in 1995-1999 to 172 conducted in 2010-2014. One-hundred seventeen (32.7%) RCTs were concerned with the field of vascular and interventional radiology; 78 (21.8%) evaluated more than one imaging technique; 164 (45.8%) had a sample size of 50-150 subjects; 246 (68.7%) showed positive study results; 185 (51.7%) were not funded; 179 (50.0%) had four to seven authors; 263 (73.5%) were single-center trials; 88 (24.6%) had a first author or corresponding author located in the United States; and 187 (52.2%) were of low quality. In the time trend analysis, the following variables showed a significantly positive trend: abdominal imaging subspecialty, CT as the imaging technique, more than 150 subjects, more than seven authors, and high methodologic quality. The quantity and quality of radiologic RCTs have significantly increased over the past 2 decades; however, the methodologic quality remains suboptimal.