Structure formats of randomised controlled trial abstracts: a cross-sectional analysis of their current usage and association with methodology reporting

Fang Hua, Tanya Walsh, Anne-Marie Glenny, Helen Worthington

The reporting of randomised controlled trial (RCT) abstracts is of vital importance. The primary objective of this study was to investigate the association between structure format and RCT abstracts' quality of methodology reporting, informed by the current requirement and usage of structure formats by leading general medical/internal medicine journals (secondary objective). A two-part cross-sectional study. First, through hand searches, we identified all RCTs published in the top-50 high-impact general medical/internal medicine journals during July-December 2015 (n = 370), and retrieved the 'instructions to authors' of these journals. From these, we extracted the actual usage of structure formats and headings, as well as relevant journal policies. Then, after a pilot study and sample size calculation, we assessed the methodology reporting quality of 176 IMRaD (Introduction, Methods, Results, and Discussion) and 165 HS (Highly Structured) RCT abstracts sampled from 33 of the 50 selected journals, using a 9-item checklist developed based on the CONSORT for Abstracts guidelines (primary outcome: overall quality score, OQS; score range 0 to 9). 88% (324/370) of all identified RCT abstracts were structured, among which 66% (215/324) used the IMRaD format and 34% (109/324) used HS. According to journals' 'instructions to authors', 48% (24/50) journals required IMRaD, 32% (16/50) required HS, 8% (4/50) required unstructured, while the rest did not state any requirement on structure format. According to generalised estimation equation analysis adjusting for potential confounders and clustering effects, the OQS of HS abstracts was 0.5 (95% CI 0.1 to 1.0, p = 0.028) higher than IMRaD abstracts. More HS abstracts reported study setting (adjusted odds ratio, 4.2; 95% CI: 1.7 to 10.0; p = 0.001), definition of the main outcome measure (2.5; 1.3 to 4.9; p = 0.006) and the time point for main outcome assessment (3.0; 1.5 to 6.2; p = 0.002), whereas more IMRaD abstracts described the unit of randomisation (0.4; 0.3 to 0.8; p = 0.004). For RCT abstracts, the IMRaD format is more frequently used and required by leading general medical/internal medicine journals than the HS format. Abstracts in the HS format report trial methodology more completely than those in the IMRaD format.