Elicitation of expert prior opinion to design the BARJDM trial in juvenile dermatomyositis.

Abstract

To elicit and quantify expert opinion concerning the relative merits of two treatments for a rare inflammatory disease: Juvenile dermatomyositis (JDM). The formal expression of expert opinion reported in this paper will be used in a Bayesian analysis of a forthcoming randomised controlled trial known as BARJDM (baricitinib for juvenile dermatomyositis). A Bayesian prior elicitation meeting was convened, following a previously described methodological template. Opinion was sought on the probability that a patient in the BARJDM trial would achieve clinically inactive disease, off glucocorticoids (GC) within a 12-month period with either methotrexate (standard of care); or baricitinib (a Janus kinase inhibitor, JAKi), with GC schedules identical in both arms of the trial. Experts' views were discussed and refined following presentation and further discussion of summated published data regarding efficacy of methotrexate or JAKi for JDM. Ten UK paediatric rheumatology consultants (including one adolescent paediatric rheumatologist) participated in the elicitation meeting. All had expertise in JDM, leading active National Health Service clinics for this disease. Consensus expert prior opinion was that the most likely probability of clinically inactive disease off GC within 12 months was 0.55 on baricitinib and 0.23 on methotrexate, with a greater degree of uncertainty for baricitinib. Experts currently think that baricitinib is superior to MTX for the treatment of JDM, although there is uncertainty around this. BARJDM will therefore integrate randomised trial data with this expert prior opinion to derive a posterior distribution for the relative efficacy of baricitinib compared with MTX.