Finding the best available evidence in pancreatic surgery – the EVIglance randomised controlled trial. (1st June 2022)

• Record Type: Journal Article
• Authors: Probst, P; Merz, D; Joos, M; Klotz, R
• Is Part Of: British journal of surgery. Volume 109:(2022) Supplement 3
• Publication Date: 2022-06-01
• Subjects: Surgery -- Periodicals; 617.005
• DOI: 10.1093/bjs/znac178.007
• Languages: English
• ISSNs: 0007-1323

Abstract

Abstract: Objective: Profound and thorough literature search is a vital element of evidence-based medicine. However, increasing number of publications and limited time make it hard to find the best available evidence. The objective of this randomised controlled trial (RCT) was to demonstrate the superiority of the ISGPS Evidence Map of Pancreatic Surgery (via EVIglance on www.evidencemap.surgery ) compared to a conventional literature search via PubMed for answering clinically relevant questions in pancreatic surgery. Methods: A single-centre, blinded, cross-over RCT including medical students, residents and consultants as participants was performed. Participants conducted literature searches with two predefined PICO (Patient, Intervention, Control, Outcome) questions, one with PubMed and the other with EVIglance. The order of the search tools and the PICO questions were assigned by randomisation. Primary endpoint was time in minutes until a synopsis was made regarding the PICO question. The synopsis was characterised by the direction of the effect and the certainty of evidence. Three times 28 participants were needed to show a minimal difference of 3 minutes at a level of significance of alpha = 1.67%. Furthermore, the correct number of RCTs found by participants that were relevant to answer the PICO question was analysed. Results: Each 28 medical students, residents and consultants were randomised and analysed. A synopsis for the PICO question was found with PubMed after 10.8 minutes and with EVIglance after 1.7 minutes (95%-CI for difference: 9.9 to 8.3 minutes; p<0.001). EVIglance answered the PICO significantly faster in all groups (medical students, residents and consultants) even after Bonferroni correction. Participants were able to guess both the direction of the effect (95% vs 48%; p<0.001) and the certainty of evidence (99% vs 30%; p<0.001) better with EVIglance than with PubMed. The correct number of relevant RCTs to answer the PICO question was found more often with EVIglance compared to PubMed (99% vs 15%; p<0.001). Conclusion: Pancreatic surgeons find best available evidence faster via EVIglance on www.evidencemap.surgery . Furthermore, a synopsis made from EVIglance is more concise regarding direction of effect and certainty of evidence. Given the advantages of EVIglance it may be considered the new gold standard for finding best available evidence in pancreatic surgery.