ABSTRACTS: ACCEPTED: PRACTICE MANAGEMENT: Presidential Poster Award
Within the GRADE (Grading of Recommendations Assessment, Development, and Evaluation) framework, recommendations are accompanied by a strength of recommendation (strong/conditional) along with an assessment of the certainty of evidence (high/moderate/low/very low). Strong recommendations based on lower certainty evidence are termed “discordant recommendations.” While generally not considered appropriate, within GRADE there are 5 situations in which such recommendations may be warranted: Strong recommendations FOR an intervention can be made when there is 1) only low/very low certainty of benefit but you have a life-threatening problem, 2) low/very low certainty of equivalence in benefit of two interventions but one is associated with less harm. Strong recommendations AGAINST the intervention can be made when there is 1) low/very low certainty of harm but the harm is catastrophic and 2) high certainty of similar benefits of two alternatives but one is less harmful, and 3)low/very low certainty of benefit but one is more harmful. We strived to assess the prevalence of GRADE recommendations and temporal trends of discordant recommendations (DR).
Guidelines produced by AASLD, ACG, AGA, and ASGE from 2013 to October 15, 2018 that used GRADE, were reviewed. Across all guidelines, we identified the prevalence of and temporal trends of DRs.
A total of 87 guidelines were identified and of these 78 (89.6%) were based on the GRADE framework. Across these 78 guidelines a total of 1,443 recommendation statements were generated. The number of DRs that did not fit the 5 paradigmatic scenarios above by year were as follows: 2013 = 12.7 % (45/353), 2014 = 15.9 % (50/313), 2015 = 19.5 % (60/307), 2016 = 29.3% (54/184), 2017 = 17.9% (32/178), 2018 = 17.5% (19/108) [Figure 1].
Discordant recommendations are discouraged because when either the benefits or harms of an intervention are uncertain, one cannot be confident that an intervention does more good than harm. Over the last 5 years, the number of DRs in gastroenterology guidelines increased until 2016 and then started to downtrend. Further investigations should explore the etiology of why discordant recommendations are made as they may contribute to patients receiving suboptimal care.