Assessing the Quality of Evidence Presented at the Annual Conferences of Diabetes Canada

OBJECTIVES: Annual general meetings (AGMs) are often seen and promoted as great sources of contemporary information for the modern health-care professional. The quality of research evidence presented at these conferences, however, remains unclear. This paper evaluates the level of evidence (LoE) of research presented at the 2015 to 2019 AGMs of Diabetes Canada (DC). METHODS: Using the framework we first published to evaluate the LoE presented at the Canadian Society of Nephrology (CSN) AGMs, 2 authors independently assigned eligible abstracts a study type and grade. Research assistants separately recorded corresponding author and city affiliations. RESULTS: Of 832 published abstracts, 68% (N=568) met the inclusion criteria. These abstracts were classified as follows: 12% Level I (highest quality); 30% Level II; 36% Level III; 14% Level IV and 8% Level V (lowest quality). The LoE remained consistent over the 5-year study period, as observed by the Jonckheere-Terpstra test (p=0.754). The LoE of DC AGMs was similar to that of CSN AGMs (p=0.125). The number of authors was positively associated with improved LoE (p<0.001). CONCLUSIONS: Based on our novel auditing methodology, the LoE of DC AGMs is on par with other analyzed national AGMs. Adoption of this metric for all AGMs would aid in identifying unfavourable trends and alert planning and abstract selection committees of trends in LoE. Such studies will increase transparency for stakeholders and facilitate quality improvement initiatives.