Level of clinical evidence presented at the Society for Vascular Surgery Annual Meeting during a 5-year period (2012-2016)

Objective: During the past decades, there has been an increasing emphasis on the use of high-quality evidence to inform clinical decision-making. The purpose of our study was to assess trends in the level of evidence (LOE) of abstracts presented at the Vascular Annual Meeting from 2012 to 2016.

Methods: All Vascular Annual Meeting abstracts for 2012 to 2016 were obtained through the Journal of Vascular Surgery. Two reviewers independently screened abstracts for eligibility. Research with a nonclinical focus was excluded from the study. Data extracted from eligible abstracts included study type (therapeutic, prognostic, diagnostic), study size, country of academic institution of primary author, presentation type, and whether the sample was recruited or from a database. Abstracts were assigned an LOE using the 2011 Oxford Centre for Evidence-Based Medicine classification scheme based on study design (eg, case series, randomized controlled trial). A χ² test and analysis of variance test were conducted to assess nonrandom changes in LOE during the study period.

Results: Of the 1403 abstracts screened, 1147 were included. Inter-rater agreement was high (κ value for abstract screening was 0.93; κ value for data extraction was 0.89). Therapeutic studies were the most common study type (58%), followed by prognostic studies (37%), then diagnostic studies (5%). The majority of abstracts (75.0%) were submitted from North American institutions. Overall, 0.35% of the presentations were level I evidence, 3.1% level II, 52.8% level III, 38.0% level IV, and 5.7% level V. The average LOE per year fluctuated between 3.54 and 3.32, with a mean LOE of 3.45. The proportion of high-quality evidence (level I and level II) increased in the years 2015 and 2016, representing 78% of all level I and level II abstracts presented in the 5-year period. A χ² test between LOE and year yielded a P value of .0084, indicating significant nonrandom change in LOE between 2012 and 2016. The majority of high LOE research was presented in poster sessions (37.5%), plenary sessions (27.5%), and international forum sessions/talks (25%) at the meeting.

Conclusions: Overall, average LOE remained relatively consistent between 2012 and 2016, with most abstracts classified as level III or level IV. There was a gradual, albeit minor, increase in the proportion of level I and level II evidence in 2015 and 2016, potentially indicating the increasing commitment to producing and disseminating high-level research in vascular surgery. Furthermore, a lack of a classification tool specific to vascular surgery research occasionally presented a challenge in assigning LOE, perhaps indicating a need for such a tool in this specialty.