386. Vasculitis: Effects of Remission Maintenance Therapies on Relapse and Side Effects: Patient Preferences (VERITAS).

Background: Relapses are common in vasculitis and preventing relapses is a goal of treatment and a focus of research. However, little is known about the degree to which patients value avoiding relapses and how much risk from treatments for vasculitis patients are willing to accept to avoid a relapse. Methods: With patient partners, we designed, piloted, and administered a survey called the Vasculitis: Effects of Remission maintenance Therapies on relapse and Side effects: patient preferences (VERITAS). The survey presented scenarios in which the risk of relapse with current treatments was 30% over 5 years followed by a scenario in which additional remission maintenance therapy resulted in either a 5%, 10%, or 15% absolute reduction in the risk of relapse and either a 0% or 5% excess risk of adverse events. Participants were randomly allocated scenarios that either left the additional therapy unnamed or specified it as prednisone. The distribution of minimally important difference (MID) in relapse risk, defined as the lowest ARR at which additional therapy was accepted, was determined for each respondent for each scenario. We used multivariable logistic regression models to calculate estimates of the effect of naming prednisone on the MID for a relapse compared to an unnamed therapy. Given the bimodal response distribution, a MID >5% was used as the outcome for the logistic regression models. Results: Of 274 respondents, 207 (75.5%) had ANCA-associated vasculitis. 201 (73.3%) participants had usable data for scenarios without, and 205 (74.8%) with, excess AEs. 138 (68.7%) accepted the additional therapy at all ARRs presented (5, 10 and 15%) without excess AEs, while 115 (56.1%) with excess AEs. Conversely, 43 (21.4%) of respondents refused additional therapy for all ARRs presented without excess AEs, and 63 (30.7%) with excess AEs. Naming the additional therapy prednisone clearly increased the benefit needed to accept therapy when no excess AEs were specified (37 [38.5%] respondents had an MID >5% in prednisone scenarios as compared to 26 [24.8%] in unnamed therapy scenarios, p=0.002) but not in scenarios with excess AEs (49 [49%] respondents had an MID >5% in prednisone scenarios as compared to 41 [39.1%] in unnamed therapy scenarios, p=0.11). Similarly, presenting scenarios with excess AEs first increased the benefits required to accept additional therapy compared to when scenarios without excess AEs were presented first (41 [42.3%] respondents had an MID >5% when excess AE scenarios were presented first compared to 22 [21.2%] when no excess AE scenarios were presented first, p